AI War:6.000 Release

Statistics For The Curious

AI War 5.0 - 6.0 spanned from Feb. 15th, 2011 through Oct. 18th, 2012 and it was full of some HUGE changes! 895 distinct changes were made as part of 97 different releases over the course of 611 days. One does not simply stroll through AI War patch notes, indeed.

Even though this was during the period that A Valley Without Wind was developed and released, we managed to keep up an average of 1 release every six days (though some periods were very light with releases and others were very dense with them, if you look at the actual dates; the average makes it sound more regular than it was). Given the length of time between 5.0 and 6.0, it's essentially the same length of time that was between 1.0 and 5.0. So that's certainly a good reason to have such huge patch notes and so many changes!

A full 149 Players are thanked in this series.

This document is the abbreviated, organized version of the full release notes (here).

Highlights

Huge

• Ancient Shadows, an entire new (paid) expansion, has been added to the game! Briefly summarizing the additions:
  • Champions! Optional hero units that can be added at the beginning of the game or midway through, these have the unique ability to fly through special wormholes into the new nebulae. In those new areas, champions encounter both friendly and hostile human, zenith, neinzul, and spire splinter factions. As your champion gains experience and recovers artifacts, you can guide its progression from a super-starship to a truly awesome powerhouse.
  • 9 New Bonus Ship Types, including the hilarious Tackle Drone Launcher and the first-ever bonus starship.
  • A New Minor Faction: the Dark Spire. These only awaken when many ships die near their spawners, but once they really get going neither the AI nor player will find them easy to stop.
  • Modular Fortresses, a long requested addition to the player's heavy defensive lineup.
  • 2 New AI Types: The tough-as-nails Heroic AI that throws champions at you, and the modular-fortress-happy Fortress King AI.
  • 2 New Map Types: The long-requested "Clusters" map type and a more advanced "Microcosm Clusters" variant.
  • 3 _Nasty_ new Core Guard Post types for the AI to defend its homeworlds with.
  • Over 90 minutes of new music from Pablo Vega.

• Hacking! We finally found a solution to knowledge-raiding that's both balanceable and fun:
  • Knowledge raiding and Superterminal hacking now both increase the same "hacking antagonism" value on the AI, and as that value gets higher the AI's response to any hacking activity gets increasingly powerful and erratic.
  • On top of this, we added a new "Ship Design Hacking" mechanic whereby you can hack an AI-controlled Advanced Research Station so that when you later capture it you can select from one of three randomly-chosen bonus ship types to unlock, instead of just being given one of them. This shares in the hacking-antagonism mechanic used by knowledge hacking and superterminal hacking.

• Completely redid the AI's Special Forces to make them an extremely important defensive mechanic that adapts to the player's offensive deployments and tries to thwart them where it counts.

• Added a new AI mechanic: the Strategic Reserve. The AI gradually adds strength to this reserve and deploys ships from it to critical defense actions, especially homeworld defenses. It also uses the reserve to contribute to Cross Planet Attacks, taking some of the burden off local defense forces.

• Added another new AI mechanic: the Threat Fleet. "Freed" AI ships will still do their normal stalk and attack behavior for about half an hour, but after that the AI figures that they've probably been stymied by strong human defenses and pulls them into a cohesive offensive fleet that waits for an opportunity to strike the human where it hurts.

• Massively reworked the energy system to cut out the "trivial to compute, time-consuming to do" micromanagement that the previous system encouraged:
  • There's no such thing as an inefficient reactor anymore: you can have one free-to-run energy-collector on each planet, and as many high-but-constant-cost matter-converters as you want anywhere.
  • Putting units in low-power-mode no longer reduces their energy cost (or ongoing m+c cost, for the energy-producers), so there's no motive to micro them up and down to pay the minimum m+c for the minimum energy necessary.
  • The free-energy collectors scale up with the number of homeworlds so that a multi-HW player doesn't have such an energy disadvantage compared to multiple single-HW players.

• Several minor factions and AI plots now have variable intensity that allows more game variety: want to play a game with TONS of human resistance fighters? Or where the Dyson Sphere has a dominant impact? Want just a few human marauders? Or hybrids on, but in much lower numbers? Or hybrids on, and massively amplified? It's all there.

• Totally revamped AI Carriers:
  • When there's too many AI ships on the planet for them to fully deploy (without unduly hurting game performance), they're no longer invincible but if you pop them they combine their contents into nasty mkV ships and mkV guardians.
  • Can no longer shoot through forcefields (but can still fly through them).
  • Now get more shots-per-salvo based on the ships being carried, so they're not so much less a threat when not-yet-deployed (previously high-carrier waves tended to be like several waves in series, which is much easier than intended).
  • There's now a toggle on the CTRLS window to tell your ships whether to auto-target carriers or not.

• Cross-Planet Attacks have been substantially reworked to be a much more serious threat.

• Each AI planet now picks 3 ship types to focus its reinforcements on, making each planet feel more distinct.

Big

• Re-implemented custom galaxy layouts for each player, so you can move planets around the map to get the strategic view looking the way you want.

• Substantially rebalanced waves, particularly on the higher difficulties.
  • Notably, 10/10 used to be totally lethal... unless you somehow pulled out some massive cheese and survived the first couple waves to get a foothold. Now the early 10/10 challenge is a much better "advertisement" for the kind of challenge that lies beyond the initial period. Accordingly, 10/10 has been massively rebalanced to be harder in those later parts of the game, as certain players have displayed the temerity to _win_ on that difficulty.
  • Also smoothed out several "difficulty cliffs" along the 7-to-10 range where one difficulty step was massively more significant than either the one just below or just above.

• The Botnet Golem has been split off into an entirely separate minor faction with its own costs/risks because it was so powerful. We could have just nerfed it down near the other golems, but this way is much more fun.

• The starfield backgrounds have been completely redone with something much more interesting.

• Rebalanced... honestly, it might be easier to list the units that were _not_ rebalanced during the 5.0-to-6.0 run than those that were.
  • Thanks to all the participants who contributed nominations, discussion, and votes to our "what's the worst unit?" (and other) community polls!

• Added a devious super-hybrid plot to Advanced Hybrids (on high enough intensity). That's a nice Dyson Sphere you've got there, it'd be a shame if anything happened to it.

• Finally found a way to detect loss-of-focus from within the Unity engine, allowing us to auto-pause the game when this happens (there's a toggle to suppress the auto-pause, if you like).
  • Note: this doesn't work on the Mac in windowed mode, but it does work on the Mac in full-screen mode and on Windows in either mode.

Important

• AI War engine upgraded to Unity 3.3 from Unity 3.1.

• The Siege Starship (formerly known as the Dreadnaught) was too effective as "park at extreme range and pick off the enemy" ship so it was nerfed into the Antimatter Starship. That couldn't really hit anything, so it was re-imagined again into the Plasma-Siege Starship, which is much shorter ranged than the old siege starship but fires special ammo that does aoe damage, engine damage, and causes extra splash damage to targets under any forcefield it hits.

• The MkIII Riot Control Starship was given some new exclusive module possibilities (grav tazer, etc) to better set them apart from the MkI and MkII.

• Added Hardened Forcefields: an alternate form of human-tech forcefield that relies on reducing incoming damage more than absorbing it.

• Added the ability to switch your UI to display what another human player sees, making team control significantly more convenient and paving the way for other features.

• Added the "Helper" player role which is much like the "observer" or "spectator" role from other RTS games but they can also control the units of any human player allowing team control.

• Added a new Ultra-Low unit-cap-scale with 1/8th the caps of High, for folks on low-end hardware or wanting to play something crazy like a 16-homeworld game.

• Added a scripting language for generating lobby game-setups within a set of desired parameters, and a number of default scripts that ship with the game (including a few "Beginner Game" scripts, a few scripts for randomized minor factions, and a few scripts for randomized _and undisclosed_ minor factions).

• Added short-lived "Drones" that the player's Neinzul Enclave Starships can build. These are based on unlocked turret technology and provide more of a "carrier with fighters" feel.

• Made defending multiple planets much more viable (relative to the viability of defending one planet) by changes to wave interval (which impacts wave size), re-imagining the Military Command Station's weaponry, and adding the new per-planet-cap'd Minifortress.

• AI Eyes now have 3 varieties: Sentry Eyes (the old behavior), Ion Eyes, and Parasite Eyes. Each is pretty lethal against no-thought-blob attacks, in a different way.

• The ramp-up from one AI tech level to the next is now much more gradual, with waves containing a percentage of the next-higher tech level that increases as you approach the threshold.

• Reclamation damage is now much more likely to have a consistent impact because any reclamator nanites unable to reclaim their host when it dies "hop" to another nearby eligible target.

• Exogalactic strikeforces can now contain Hunter/Killers. Be afraid. And throw the rotten fruit at chemical_art for suggesting it.

• Exogalactic strikeforces are now more wily about picking targets. Get used to defending your outlying assets if you play with any exo sources.

• Several golems have been substantially buffed, to provide an appropriate reward for the costs/risks of playing on Golems-Medium or Golems-Hard. Enjoy your new "has tractors that paralyze!" Black Widow Golem and "that thing can kill anything, once every 8 seconds!" Artillery Golem.

• The game's RAM usage has been substantially reduced in many cases, leading to far fewer out-of-memory errors in remotely-sane scenarios (and even some clearly-insane scenarios, we've noted).

• Massively reworked the way the AI determines how many ships it gets in a reinforcement, so that it properly pays attention to how much a ship should "cost" relative to others (based on relative ship cap, mainly). This makes it much less likely that the AI will pile up huge amounts of low-cap ships on a planet, and if it does create such a pile up at least the proper price (in terms of fewer other ships elsewhere) was paid.

• Fixed all remaining desync bugs; haven't had a reported desync in months.

• Now when a forcefield is damaged, every other overlapping allied forcefield also takes a minimal amount of damage, preventing the old far-too-effective-to-be-balanceable tactic of having engineers repair inner forcefields while the outer ones tanked the damage. It was fun, but properly micro'd it was effectively invincible before.

• AI Homeworld structure seeding now distinguishes between normal and "brutal" structures: the core raid engine, core CPA guard post, and AI Eye were far more significant than the other structures being considered, and the difference between an AI HW with none of them and an AI HW with 5 core raid engines could be the difference between "pushover" and "impossible".

• Made fortresses, forcefield generators, and command stations drop rebuildable remains when they die, greatly reducing the need for micro when rebuilding after a significant AI breakthrough.

• Added the first per-control-group control options so you have finer control over auto-Free-Roaming-Defender and auto-kiting behavior.

AI Updates

• Previously, allied minor faction and zombie ships would just patrol around player planets endlessly, which was helpful for defense, but problematic for two reasons: first, that it could cause massive amounts of slowdown after many hours of them stacking up, and second, that it could cause massive buildups of AI stalkers, who never dare to attack the planets with so many defenders, which leads to something akin to stalemate situations.
  • Thus the allied minor faction and zombie logic has been adjusted as follows:
    • Neutral planets are now considered the same as allied planets, and the ships will patrol to them like any others.
    • If there are more than about 100 of AI threat ships on an adjacent planet, the ships will also patrol to that planet to clean them off.
    • If the human-allied team has more than a token military force (more than about 10 ships) on a planet, then these ships will patrol to there, to help aid in that attack.
      • This allows for small raids to still happen with raid starships or whatever without assistance from these minor factions, but it enlists their help for larger-scale engagements.

• Attack Hybrids now use a variant of the behavior the standard AI free (threat) ships use to avoid charging a human planet with insufficient firepower. Combined with other hybrid changes in this release this may result in hybrids being a lot harder (and/or a lot more annoying) than they used to be.

• Reworked the CPA formula to be more aggressive on higher difficulties, since it's been sadly anemic lately.
  • Also added cpa-size calculations to the wave logs, when advanced logging is enabled.

• Removed a "go for the command station over everything else when AI" behavior from a number of units, including:
  • Leech Starships.
  • Flagships.
  • Zenith Starships.
  • Spire Starships.
  • Riot Starships.

• Barracks are now seeded on some AI planets at the start of the game (and retroactively into old saves):
  • The higher the difficulty, the more planets get them.
  • The higher the difficulty, the more ships in them.
  • None are placed closer than 3 hops from a human homeworld, to not be too cruel.
  • These serve two purposes:
    • Making it far less likely that the AI will simply not have enough ships for a CPA, while still letting players whittle away at "the reserve" if they want to.
    • Making the AI's defensive response to the loss of a planet (with a barracks on it) something more of an event.

• AI planets now pick a primary, secondary, and tertiary type of ship to focus their reinforcements on.
  • It's not purely random, sometimes it picks ships that fit certain roles.
  • When picking a new reinforcement ship for the central pulse or a non-special-forces guard-post-pulse, it has a 10% chance to pick any available ship like usual, and a 90% chance to pick one of its focus types. When picking a focus type, it has 3 times as much chance to pick the primary as the secondary, and 3 times as much chance to pick the secondary as the tertiary. If that doesn't make sense, don't worry about it.
  • Whenever the AIs unlock new ship types due to AI progress, all AI planets "reroll" their focus types.

• The AI now has a "strategic reserve" it can deploy to defend critical planets (the homeworld, the core worlds, and the superterminal world; though it still holds onto most of it when defending the latter two) and pulls back up (off the board) if the attack goes away.
  • Reserve ships are basically just normal mkV fleet ships, but they are not allowed to leave the planet so they can't be easily baited to their doom.
  • The reserve starts at zero and increases periodically, so it can be whittled away.
  • The max size of the reserve and its increase rate are proportional to the difficulty of the AI player in question (each AI has its own reserve) and AIP. On Diff 7 with non-insane AIP it's pretty tame. Higher up, it gets more intense.
  • If you're curious about the numbers, with advanced logging on they go into the special forces log (since the concept is somewhat similar, and didn't seem to need a separate log; might change that).
  • Thanks to Faulty Logic and others who voted on this poll for inspiring this change (more coming eventually in response to that, for that matter).

• AI Special Forces:
  • Now get periodic extra spawns (in addition to the usual trickle from reinforcements) if under a "special forces population cap" determined by difficulty, AIP, etc.
    • The spawns are spread out among all special forces posts (and special forces guardians) in the galaxy, except those in non-AI territory. Each post in non-AI territory instead increases the special-forces-population-cap by about 5%. So there's a reason to kill them, in tension with the +1 AIP from killing them.
    • If you're curious about the math, etc, turn on Advanced Logging and check SpecialForcesLogicLog.txt in your RuntimeData directory after the game has run for a bit. There's some minor spoilers in there about what ships each AI has, but nothing you probably won't find out looking at a few AI planets anyway.
  • Will now rally to defend an AI planet that is under significant attack and at least mildly important (AI Homeworld, core world, or a planet with a CSG, ARS, Advanced Factory, Super Terminal, or Broken Golem on it, giving priority to HWs/core-worlds).
  • Now keep the special-forces flag even when they have encountered the enemy, so that rallying to a planet does not just dump all the special forces there as normal threat.
  • When not rallying to defend a planet, will select a staging planet about 3 hops deep into AI territory and rally there until an eligible AI planet is attacked.
  • To prevent a player's first attack from running into several hundred AI ships out for blood, as entertaining as it was during testing, special forces ships are no longer seeded during the initial map seeding.

• Since "make the threatballs fish or cut bait" was #1 on the second round of the 6.0 poll:
  • Threat ships behave as they used to for about 30 minutes after being freed, but after that they are switched to an alternate "Threat Fleet" behavior that is somewhat similar to the new Special Forces mechanic.
    • Note: this only happens on Difficulty 6+, as it's not the AI getting anything extra, it's just behaving more intelligently (in theory) with what it has.
  • If an AI homeworld or core world is under attack, the threat fleet will rally to defend it.
  • Otherwise, if it sees an accessible non-AI planet with a significant human presence that it thinks it can take out, it goes to attack that (it will still pool up at the entry wormhole in some cases, until enough of them are there to pass the threshold, so they don't march in to the grinder one-by-one).
  • Otherwise, it picks a planet in AI territory to hang out at until either of the two above conditions are met.
  • If a carrier is spontaneously formed from ships that have a significant threat-fleet population, the carrier (and anything it spawns) is also considered threat-fleet.
  • When a special forces alarm post is triggered it converts all special forces in the galaxy directly into threat-fleet, and then blows up the alarm post (not causing AIP, unless it was triggered because it was shot to death).
  • For the curious who have advanced logging on, the target-planet selection process is logged in the special forces log file.

• Fixed a relatively longstanding bug where AIPerPlanetShipCap and AIPerGuardPostShipCap were generally not being enforced when loading a game (either just-starting or loading later).

• Fixed a bug where hybrids were... basically completely broken. Sigh. The "don't let hybrids rebuild modules if they've recently been damaged" code was inverted in such a way that they could never build modules. And because they could never get modules, they were never satisfied that they were ready to go attack or defend or whatever.

General Large Gameplay Additions

• Advanced Research Station Hacking:
  • Added the "Ship-Design Hacker" (for lack of a better name, let us know if you think of one): a 4th ship to the "science lab" line, very similar to the "Knowledge Hacker". It gathers no knowledge, but has another ability.
  • If you keep a Ship-Design Hacker on an AI-controlled planet with an AI-controlled ARS for 10 minutes (it must remain stationary during this time; it's one of those kinds of countdowns), it will "hack" the Advanced Research Station and reprogram it such that when you capture the planet:
    • Instead of being immediately granted a new bonus ship type, you will be able (from the tech menu of the ARS, or of a Ship-Design Hacker) to select one of three bonus ship types.
    • Note the hacker does not have to survive after the hack has been completed, and the ARS does not have to survive after the planet is captured, but you will need a science lab, ARS, or SDH (anything with a tech menu that normally lets you unlock II/III versions of fleet ships, basically).
  • However, the AI will react to Ship-Design Hackers the same way they react to Knowledge Hackers. Also, the number of previous ship-design-hacks you've pulled off will increase AI response to future hacking (of all 3 kinds). Each subsequent ship-design-hack makes things a lot worse than the previous one did, such that 1 isn't a big deal, 2 is a big deal, 3 is a huge deal, 4 is probably death if you hack seriously again, and 5 is really going to mess you up if you so much as look like you're hacking the rest of the game.

• Superterminal Hacking:
  • Now the superterminal never seeds on an AI homeworld or core world (can still seed next to a core world which is mean, but not nearly so mean as having to permanently alert a homeworld, or capture one, in order to use the superterminal).
  • Also, the superterminal's response is now based only on the reduction achieved through the superterminal, not total AIP-reduction, so it's no longer very important to do the ST before getting reduction from other sources.
  • Now the antagonism generated by knowledge raiding feeds into the strength of superterminal spawns, and the number of times a superterminal has "ticked" feeds into that same antagonism (thematically, they're both "hacking" and the AI dinnae like it).
    • Reducing AIP via superterminal hacking is significantly more "efficient" than knowledge raiding in terms of what you gain in return for making future hacking efforts more dangerous. The disadvantage is that this requires the superterminal, and doesn't stop if you lose control of the system (unless you destroy the superterminal).
  • Now also uses a "wild roll" system like what k-raiding uses, but it's somewhat more tame in terms of what kinds of crazy it can throw in there. Right now all it can do is surges (10% of the time, the same effect that was added in the previous version), and the "short range warp jump" (15% of the time, spawns the units away from the superterminal itself) wild-effect that the k-raid spawns can get.
  • If you have Advanced Logging enabled, each spawn will generate an entry in the "CounterSaboteurSpawns.txt" log file; it may not make sense to players but is very helpful to us if you submit it with your feedback.
  • The rationale for this is trying to make both k-raiding and superterminal-hacking more interesting and more of a strategic choice (since now using one makes the other more dangerous), but also to make it harder to use the superterminal for arbitrarily-high AIP reductions (we've seen cases where players could kill 400 AIP with it without even being in much danger).
    • As with k-raiding: In general, the balance of all this is tentative. It looks right in our testing, but it may be too easy or (perhaps more likely) too hard. Please give us feedback and we will continue to refine towards better balance and more fun :)

Lobby Setup Scripts

• Preface: over the years we've had requests accumulate on a number of different fronts that couldn't easily be handled the way things were, like:
  • Numerous new players asking "What's a good setup for my first game?"
  • Variable-strength minor factions and AI plots (which could be handled just by the above changes, but read on)
  • Randomized minor factions and AI plots, particularly combined with variable-strength.
  • Randomized and hidden minor factions and AI plots (so you don't know what's there until you run into it in the game, with potentially hilarious results).
  • Some kind of overall "difficulty ranking" for a particular game setup.

• So, from a small scripting language that was written for (and actually not ultimately used for) AVWW, a Lobby Setup Scripting system has been cobbled together. The actual scripting language is likely to completely change in the next version (it's really verbose, cumbersome, etc, but it does work), but the short story is it's now possible to define external scripts which populate the settings in the lobby.
  • Please note, those of you with scenario-building aspirations, that this really only sets stuff in the lobby. It cannot directly influence mapgen (placement of planets, units, etc) in any way.
  • To use a script, start a new game, and in the lobby on the Map tab pick an option from the Setup Script dropdown. It will then immediately execute the script and apply the changes to the lobby.
  • The following scripts are available in this version:
    • Beginner Game
      • A fairly simple, fairly easy scenario that may be a good next step for new players who have just completed the intermediate tutorial.
    • Beginner Game 2
      • A bit harder than the scenario generated by the Beginner Game script. May be a good next step for new players who have just completed the intermediate tutorial and are feeling confident.
    • Random Factions 1
      • This script sets up some randomized minor factions and AI plots from the base game (no expansions required).
        • Note (this goes for the rest of these scripts, too): it's not purely random, it starts with a certain number of "points" and randomly picks factions/plots to allocate them to. Friendly factions (in this case, human resistance fighters) actually increase the point total when chosen, rather than decreasing it. The point is that while there's simply no way to get any kind of overall "difficulty score" of a particular scenario, it is possible (with significant effort) to make a script which produces scenarios within a particular range of difficulty. Which, hopefully, scratches the real itch behind the "difficulty score" request.
    • Random Factions 2
      • This script sets up some randomized minor factions and AI plots from the base game and The Zenith Remnant expansion.
    • Random Factions 3
      • This script sets up some randomized minor factions and AI plots from the base game and the The Zenith Remnant and Children Of Neinzul expansions.
    • Random Factions 4
      • This script sets up some randomized minor factions and AI plots from the base game and the The Zenith Remnant, Children Of Neinzul, and Light Of The Spire expansions.
    • Unknown Factions 1
      • This script sets up some randomized minor factions and AI plots from the base game (no expansions required), and then hides them so that you cannot see what has been selected either in the lobby or in the game. If you then pick another script (or none) the hidden status will be reset.
    • Unknown Factions 2
      • This script sets up some randomized minor factions and AI plots from the base game and The Zenith Remnant expansion. It then hides them so that you cannot see what has been selected either in the lobby or in the game. If you then pick another script (or none) the hidden status will be reset.
    • Unknown Factions 3
      • This script sets up some randomized minor factions and AI plots from the base game and The Zenith Remnant and Children Of Neinzul expansions. It then hides them so that you cannot see what has been selected either in the lobby or in the game. If you then pick another script (or none) the hidden status will be reset.
    • Unknown Factions 4
      • This script sets up some randomized minor factions and AI plots from the base game and The Zenith Remnant, Children Of Neinzul, and Light Of The Spire expansions. It then hides them so that you cannot see what has been selected either in the lobby or in the game. If you then pick another script (or none) the hidden status will be reset.

Energy System Rework

• The motivation here:
  • The previous energy system was settled on after a series of different models were tried, and has been around for over 2 years, and has served pretty well.
  • But there were a number of problems pointed out by players over the years, the most bothersome being:
    • There's significant motive to maintain as close-to-zero positive net energy balance because energy costs metal+crystal per second and a positive net energy balance provides no real benefit.
      • A positive balance did protect against having your forcefields shut off in case of losing a reactor, but that could be easily protected against by a big pile of low-powered reactors that you can bring up as needed; or you can go low-power a hundred spider turrets, or whatever.
    • And maintaining that low-wasted-energy state involved _tons_ of micro. A couple of "automated energy hamster" hotkeys were added to greatly reduce that micro (by making the "low-power least efficient 'on' reactor" and "power-on most efficient 'off' reactor" operations a single keystroke instead of minutes of scrolling and clicking), but that was really just a bandaid.
  • Many people have requested automated energy management as a solution to all this, but whenever the game is simulating something that it then turns around and plays optimally without your involvement... that means something is wrong. In this genre, anyhow.
  • Thanks to Cyborg and other players for making this point, and being patient as we tried to find a better model.

• The overall result is that the energy/econ game should actually be easier, unless you were relying on low-efficiency reactors or low-powering tons of stuff to make ends meet. In those cases it shouldn't be much harder.
  • But please let us know if the new model really messes you up in some way, or if you have other feedback. It's not like this model can't be changed, it just seemed a definite step in the right direction from where we were.
  • If you're curious about the comparison math, check out this post: http://www.arcengames.com/forums/index.php/topic,11031.msg110117.html#msg110117
  • Anyway, on to the specific changes:

• Removed the Energy Reactors (all of them, MkI, MkII, MkIII).

• Added the "Energy Collector", which is like the energy reactors were except:
  • Produces 150,000 energy.
    • For reference, in the previous model an EnergyI+EnergyII+EnergyIII produced 125,000.
  • No ongoing m+c cost.
    • For reference, in the previous model an EnergyI+EnergyII+EnergyIII cost 57m+57c per second to run.
  • Can only build one per planet per player.
  • Output scales up with that player's number of homeworlds (to scale up for the extra stuff multi-HW players can build, similar to how reactors used to do it but without spamming lots of them).
  • Cannot be low-powered (there'd be no point).

• Added the "Matter Converter", which is like the energy reactors were except:
  • Produces 50,000 energy.
  • Costs 100m+100c per second to run.
  • No construction cap, and also no efficiency penalty for stacking.
    • Yes, this does mean that for now it's best to just pile these all up in a defensible place (much like manufactories); that will probably change in the future but probably not through an efficiency penalty.
  • Cannot be low-powered, since the main point of this rework is to avoid "player-time-tax" micro.
    • This can still result in mid-battle micro if you just lost a collector or whatever and need to build another converter, but that's real in-game costs at a critical moment rather than the only difference between optimal and sloppy play being how much wall-clock-time you personally want to spend "making change for a dollar" throughout the game.
  • In general, these are very inefficient compared to the old reactors, but more efficient than deeply-stacked low-efficiency reactors. The idea is that the collectors replace the output you used to get from normal-efficiency reactors, and the converters fill the role of the low-efficiency reactor piles.

• Putting units in low-power mode now no longer reduces their energy use.
  • Again: removing the motive to micro stuff for energy purposes is one of the driving motivations here.
  • Renamed "Low Power mode" to "Stand Down mode", since it's no longer for the purpose of reducing energy.
  • The mechanic for reducing energy use in stand-down mode is still there, just unused. It may be used for golems or stuff like that later, we'll see.

• Removed energy costs from Human Home Settlement and Cryogenic Pod structures, since you can't low-power the settlements anymore (and 8000e a pop isn't much fun in a tight early-game situation).

• Removed the auto-build-energy-reactors controls, and added toggles at the galaxy-level and planet-level for autobuilding collectors, and sliders at the galaxy-level and planet-level for autobuilding converters.

• Added new slider to galaxy-wide per-player-controls screen: Energy Buffer To Maintain
  • If this is set above zero, the game will try to maintain that amount of "buffer" energy above zero. A few examples:
    • Docks will not build ships that would take you below the buffer amount.
    • Ships will not be reclaimed if it would take you below the buffer amount.
    • Broken Golems will not be activated if it would take you below the buffer amount.
  • This can be restrictive, but it can also help if you're deliberately trying to keep all your defenses operational in the event of the AI destroying one or more of your energy producing units.
  • Note that this will not stop you from scrapping energy producing units (however low that may take your net energy), so be careful.
    • This is because going negative energy never stopped scrapping before, so that part wasn't changed.
  • Thanks to chemical_art for the suggestion.

• When upgrading saves from 5.039 and earlier:
  • For all EnergyI, EnergyII, and EnergyIII units:
    • If that player already has an Energy Collector on that planet, the reactor is replaced by a Matter Converter.
    • Otherwise, the reactor is replaced by an Energy Collector.
  • After that, while the player has > 60,000 spare energy and any remaining Matter Converters:
    • Scrap a Matter Converter on the planet with the most remaining Matter Converters controlled by that player.
  • The result should be that:
    • If you had a positive energy balance before this version, you will have a positive energy balance when you load the save in 5.040+.
    • It won't give you a lot of Matter Converters you don't need.
    • You'll still want to look at your energy situation and make adjustments for the new model. Particularly if this version-upgrade logic somehow messes up in your case ;)

• The automated energy hamsters have taken off without giving notice, but we suspect they're on vacation in Bermuda.

• In case you were wondering, all the tooltips and relevant tutorial components have been updated for the new model, but please let us know if we missed something.

Co-Op Improvements

• Fixed another rare desync that would occur when the "ship to kill" logic was loaded from savegames that were saved during a big battle. Not many people save DURING a big battle, so this one didn't come up much.

• Added in a new "aggregate hash" that sums up a few key integer stats from non-cold-storage ships (location xy, destination xy, is destination settled, "ship to kill" object number), to provide a sort of early warning system against desyncs.
  • This makes the game much likely to actually desync quicker when there is a desync going on under the hood, which will make them easier to find, and hopefully also easier to reproduce when a player does run into a desync.
  • This adds a bit to the CPU load, but since it's all addition, it's nothing particularly notable, which is one of the plusses of this.

• Fixed a bug where helper players were getting the overflow from other players resources income-past-cap.
  • The resources were basically being lost since helpers cannot gift resources, which is intentional as otherwise this overflow case would represent an in-game bonus from having slots devoted to "helper", which would invalidate the assumption that the game can be balanced as if they weren't there and thus the AI's strength doesn't have to go up due to the presence of helpers, etc.
  • Along with this: apparently in the past overflow was going to ALL human player slots, even those that had never had any players in them, causing the resources to be split more ways than necessary in MP games with fewer than 8 human players. Fixed now.

• Fixed a bug in multiplayer where hacking an ARS and having the first player pick an unlock on that planet would cause the other human players to be given a random unlock about a game-second later.

Performance Improvements

• HUGE Memory breakthrough.
  • Bad news first: the bad news is that this touched 63 different code files, and has resulted in approximately 4500 disparate lines of code being changed in a spiderweb pattern all throughout the application. So... expect some bugs. So far we haven't seen any, but with an organization change this drastic some are inevitable. Why, oh why, would we do this?
  • Well, that brings us to the good news: the baseline heap memory usage for the application has been literally HALVED, and in large savegames it's also... well, about halved in most cases. For those players that had been experiencing the dreaded "Too Many Heap Sections" error (thanks to Unity 3D's less than stellar garbage collector bounding capping out at about 900MB), this is tremendous news.
  • As a small ancillary bonus, this also makes both the program itself and individual savegames load a bit faster. The older your CPU, the more you will notice this effect (on older CPUs, it could shave off multiple seconds, actually).
  • As another not-so-small bonus, this restructuring cuts down on some confusing cases with the spirecraft and how some of the by-unit-scale stuff works. Long-term, this should mean a lot fewer bugs related to the unit cap scales, which is also welcome news. Some subtle bugs related to the armor rating have already been fixed as just part of making this transition, though we don't think anyone reported these specific ones.

• Put in a small code change that might lead to more efficiency during really large battles with a lot of things being drawn as well as a lot of sound effects being played. It might also make no difference, but it's a good precaution either way; our tests so far have not really been clear as to whether there is any benefit.

• Added new unit cap scale option: Ultra Low
  • Most large-cap ships have roughly a quarter of the standard cap (Fighter MkI = 24), and are roughly four times as expensive, four times as strong, etc.
  • This puts the lowest possible load on the your CPU when you have the biggest-possible battles. Also helpful if you're playing a game with 8 human homeworlds (in multiplayer or otherwise).

• Backported the draw order sorting from AVWW after all. This should help with the performance of lots of shots, lines, forcefields, or similar being on the screen at the same time. Or far zoom icons for ships, for that matter. It won't help for the ships themselves when you are very zoomed in, but that's comparably rarer.

• Removed the usage of Graphics.DrawMesh (which helped performance in AVWW, but seemed to harm performance in AI War for some people) in favor of Graphics.DrawMeshNow (which is what we've been using in AI War up until the last couple of releases). If you were seeing performance problems in the last few versions, please let us know if you continue to do so.

• Added a memory optimization to the aggregate-targeting system to only initialize the data structures as it needs them rather than all up front. This may cause some slowdown when fighting on a planet for the first time or with a bunch of new ship types at once, etc, but in general that shouldn't be too bad and will "work itself out" during the early phase of a battle.
  • Memory had been pretty ok, but every time we add new ship types that's more memory footprint all around so it was good to counterbalance that a bit.

• Removed the sprite pooling, including the sprite pooling option in the settings menu, and greatly improved both RAM use and CPU load with this and related underlying functions.

• In particular, improved the batching of things like destination lines, ship-to-ship beams, and so forth. These should now perform better than they did before we even started the backporting (which then caused a performance hit on these operations).

• Fixed a longstanding (since March 2011) bug where several per-unit computations were being done every frame instead of every 6 or so frames as designed.

Interface Improvements

• The "selection description" window in the bottom right of the hud that shows the count of each type in the selection now shows two percents: the first is the average percent-of-max-health of the living ships of that type (this used to be there before the Unity port, and it's back now!), the second is the lowest percent-of-max-health of the living ships of that type.

• The selection description window will now try to display up to 30 entries without resorting to a scrollbar, instead of only 5.

• Replaced the old "if CanUseNeinzulRegenerator then multiply by 2" wave-size rule with a more general UsefulnessInAIWaveMultiplier UnitData field.
  • Everything with CanUseNeinzulRegenerator (which is just the 5 youngling types), gets a 2 for this.
  • As an experiment, Bombers get 0.8 and Fighters and Missile Frigates get 1.2. Bomber waves will probably still be more dangerous (as is intended) but it might make the difference in "interestingness" less. Not planning to touch the bonus ship types because if the AI has a particularly bomber-like bonus ship type that should just be a characteristic of the kind of difficulty you'll be facing in that game, etc.

• On to #2 on the vote tallies list: The mouseover tooltip for the AI Progress display on the resource bar now displays each of the tech level thresholds for each AI Player.
  • Thanks to wyvern83 for the suggestion.

• Courtesy of being really high on the vote tallies: previously double-clicking a single unit would select all (your) units of that exact type on the screen. Now triple-clicking will do the same thing except it will include all units of the same general type (ShipType). Meaning that triple-clicking a fighter of any mark will select all your fighters of any mark on the screen.
  • Caveat: the general categorization being used is sometimes more general than mark; notably remains rebuilders have the same ShipType as engineers. Requests to make triple-clicking an engineer not also select remains rebuilders are likely to be ignored ;)
  • Also, we have to do this triple-click detection manually, so basically it's looking for a third click within 3 tenths of a second after the double-click selection. If this timing is problematic, let us know.

• In response to the number-one-on-the-vote-tallies mantis issue:
  • Added new PlanetView KeyBind: "Show Strong/Weak Info":
    • Defaults to Alt+W.
    • When this is active and the mouse cursor is over a ship, each planetary summary sidebar entry will display Win/Lose/Draw (and a % indicating intensity of win or loss, if it's not a draw) as a rough indicator of how effective a cap of the ship type under the cursor would be against a cap of the entry's ship type. Since planetary summary entries frequently include multiple distinct types of ships, this is perhaps most useful in conjunction with the "Guide" mode of the planetary summary that shows all ships of a given mark level (the default key for switching to guide mode is F1). However, even in normal mode this can be useful to at least get a rough feel for which of your ships on the planet are the best you have on hand against a specific target.
    • Caveat: this is not using a full simulation or anything like that, it uses the same simplified formula as the reference tab uses. So stuff with modules won't really get accurate results (because it's only counting the base hull), etc.
    • Suggestions on text color, default keys, etc, are quite welcome; we're not totally happy with the effect right now but don't have a lot of time to fiddle with alternatives.
  • Added new PlanetView KeyBind: "Strong/Weak => One Vs One":
    • Defaults to Alt+E (so the full combination would be to hold Alt+W+E).
    • When both this and the "Show Strong/Weak Info" keybind are active, the strong/weak info shown will use a one-ship vs. one-ship comparison instead of the normal cap vs. cap comparison.

• Courtesy of the new #2 item on the mantis vote-tallies: Added "Auto-Scout-Picket" command to the context menu (there's also an key bind, unbound by default):
  • Tell all cloaking scout units in the selection to try to station themselves on a planet where you do not currently have scout intel. Basically it:
    • 1) Makes a list of all scouts with cloaking in your selection.
    • 2) Makes a list of all planets you do not have current (less than 5 seconds ago) scout intel.
    • 3) Sorts that list of planets by distance (ascending) to the planet with your selection.
    • 4) Tells the first scout in the first list to go to the first planet in the second list, and so on. If it runs out of planets it loops back to the beginning of the planet list and thus it will double up, etc. When it runs out of scouts it stops.
  • Please remember that this order (and auto-explore) are only intended to automate simple scouting situations and thus save you time giving otherwise trivial orders. If your scouts all die before accomplishing anything that means that it's not a simple scouting situation and you'll need to take manual control to get good results.
  • Thanks to Malibu Stacey for the suggestion that inspired this command.

• Previously the controls window was not fitting vertically on 1024x600 (which isn't officially supported but we try to make it playable). Fixed to fit better.
  • Thanks to Talaraine for the report.

• Courtesy of the current #1 on the mantis vote tallies: Added "Copy To Disk" and "Copy From Disk" buttons to bottom of the controls screen.
  • Copy To File creates a copy of your global controls (excluding per-planet stuff, but including ship designs) as "copiedGlobalControlSet.dat" in your RuntimeData directory.
  • Copy From File prompts you to confirm first, and if you confirm it overwrites all your global control (excluding per-planet stuff, but including ship designs) from what's in "copiedGlobalControlSet.dat" in your RuntimeData directory. If you don't actually have such a file it doesn't do anything when you confirm.

• Courtesy of the new #2 item on the mantis vote tallies list: Added "Mobile Tractors With Full Load Rally" toggle to the global tab of the controls screen.
  • If this toggle is checked, when one of your mobile units with tractor beams (notably Etherjets, Spire Tractor Platforms, Spirecraft Martyrs, and Black Widow Golems) has a "full load" (all tractors active), it will look for a rally post to rally to. If that rally post happens to be a redirector, that will work as it normally does, etc.
  • When the rally trigger kicks in, all other orders (including FRD and attack-move, etc) will be canceled on the tractoring unit. But if you give the ship an order after the rally behavior is triggered, it won't retry the rally unless it it loses one of its tractor targets and then fills up again.
  • Note that this does not apply to Riot Control Starships with tractor modules, because the modules themselves are not mobile units.

• Fixed the close-zoom image for the AI Core Neinzul Spawner Guard Post having a height that wasn't a power of 2.
  • This would only impacted you if you were all the way zoomed in on one of these, and would have led to it looking stretched or (on older graphics card) looking gibberish-like.
  • But in the recent backporting we added some extra sanity checks that catch this when the player tries to look at it.

• An all-new, much higher-quality starfield/nebulae background effect is now drawn behind planets. These are vastly more varied and interesting, and require a similar amount of CPU/GPU processing to render as the old ones did.

• The game no longer scrolls the nebulae at the start of the game, instead giving you a really good static look at a really interesting nebulae background.

• A completely new method for rendering the planets has been put in place, and they are now more a part of the background starfield/nebula than anything else.
  • As part of this, 14 of the pre-existing planets have been kept although scaled down, and 40 new planet graphics by Eldon Harris have been added to the game. It makes quite a difference when combined with the new nebulae!

• The graphics of planets 3, 8, 11, and 16 have been substantially improved so that they have better borders and look more convincing as planets.

• Fixed a graphical bug with the border drawn around barracks icons in some contexts.

New Ships

• Added new unlockable support unit type: Hardened Forcefield Generator MkI-III. Basically the same as normal human generators except:
  • 1/4th the HP.
  • 1,000,000 Armor Rating instead of zero (this effectively reduces all incoming damage to 1/5th its normal power, unless the attacker has extremely high armor piercing or very high damage-per-shot).
  • They do not shrink as they take damage, making it easier to know how long a given target will be protected, etc. On the other hand this may make it harder to tell "at a glance" how long the generator has left to hold, so we might need to do something else here.
  • The first mark is not available by default, but costs 1000 knowledge (mkII and III are 2000 and 4000 knowledge each like the normal line).

• Science Lab III:
  • Has been renamed to "Knowledge Hacker"
  • Is now mobile, cloaked, and significantly more durable.
  • Now has a 15-second setup-after-moving time (similar to the MRS's 60-second setup time before it can repair) during which it cannot gather knowledge.
  • Since you can now build them off-planet where they're not triggering spawns, these changes significantly reduce the amount of player-hassle and wall-clock-time spent on a single knowledge raid; the raids themselves (beyond the first raid) are far more dangerous so it should balance out difficulty-wise, but be less annoying.

• Added new type of AI Eye: the Parasite Eye.
  • Has same trigger conditions.
  • Instead of spawning ships it powers on (it's in stand-down except when actively triggered) and starts firing. You really don't want that to happen. Multiple hyper-powerful railcannons that reclaim what they hit.
    • Standard "what can be reclaimed" rules apply.
    • Has the extra flag to prevent it from firing on non-reclaimable targets thanks to a test where multiple starships were dying every two seconds.
    • Note that this has a kill rate about 1/2 that of the Ion Eye, so the overall rate of "evening the field" should be similar.
    • In case you're curious, the results are not zombies and can be re-reclaimed if you're so inclined.

• Added new type of AI Eye: the Ion Eye.
  • Has same trigger conditions.
  • When seeding an eye, the game has an equal chance of seeding each type of eye, and only one eye is seeded per planet that's supposed to get one.
  • But when triggered, instead of spawning ships it powers on (it's in stand-down except when actively triggered) and starts firing. You don't want that to happen. It has the firepower of four MkV Ion Cannons.

• A new Ancient Shadows bonus ship class has been added: Neinzul Scapegoat (Mark I-V).
  • Mature Neinzul organism without the short lifespan of a Youngling. Capable combatant in its own right, but also sacrifices itself to regenerate military fleet ships on the same planet when they take critical damage.

• Another new Ancient Shadows bonus ship class has been added: Spire Railcluster (Mark I-V).
  • Large Spire fleet ship mounting multiple short-range railcannons. Highly effective against very small ships.

• New Ancient Shadows Bonus Ship Type: Zenith Reprocessor:
  • A larger, specialized version of the Acid Sprayer, this ship's shots actually "metabolize" the enemy target. When the target dies, a percentage of the target's metal+crystal construction cost will be granted to the player that did the most reprocessing damage to it. The more reprocessing damage the ship has taken, the larger the percentage of resources reprocessed.

• New Ancient Shadows Bonus Ship Type: Zenith Siege Engine:
  • Moderately large vessel built around an outsized plasma siege cannon. The ship does not have enough power to both move and keep the weapon online, so it must remain immobile for a full reload cycle before it can fire. When it fires, however, the enemy will feel it.

• Added new Ancient Shadows bonus ship: Spire Corvette.
  • A starship-sized vessel developed by retrofitting the smallest Imperial Spire combat design with human technology. Has a moderately powerful photon lance, but can also mount a small number of modules.
  • Built from the SCRV tab of the Starship Constructor.

• Added new Ancient Shadows bonus ship: Saboteur.
  • Fires hostile nanites at enemy ships. Unlike reclamation nanites, these focus on overloading the target's power systems with the goal of causing it to explode violently. This target does not go critical until its health is reduced to zero normally, but once that happens it will explode and damage up to 40 nearby friendly (to the exploding ship) units. The damage done to each victim of the explosion is based on how much damage the exploding ship took from overload-nanites, and generally cannot exceed one-tenth the maximum health of the exploding ship.

• Added new Ancient Shadows bonus ship type: Youngling Firefly.
  • Small, shortlived ship that builds up a charge as it successfully attacks other units. When it dies it explodes violently to damage up to 20 nearby enemy units. The explosion damge done to each victim starts at about the same strength as the Firefly's normal shot, but increases with the charge the ship has accumulated, up to a maximum of 20 times that.

• Added a new "Mini-Fortress" unit to the SUP tab of the command-station/mobile-builder buy menu (this is a base-game unit, not tied to an expansion).
  • These are basically a MkI fortress on a 1/10th scale (stat-wise), and you can only build 2 per planet (per player, also scales with multiple homeworlds), but there's no galaxy-wide cap so putting them on your satellite worlds does not take away any cap your main defenses could be using.
  • Unlocking these costs 1000 knowledge (for reference, MkI forts cost 3000 knowledge to unlock, and provide 25x as much firepower at cap for your main line but not the per-planet-cap flexibility of these).
  • These don't have the penalty vs polycrystal (or scout) hulls, making them actually able to deal with bombers.

Ship Logic Updates

• Changed the targeting logic for AOE units (that don't already use the alternate logic for engine damage, paralysis, reclamation, or insta-kill) to prefer non-AOE-immune targets. This is necessary because of the recent change to let flaks and some other aoe units be able to damage an aoe-immune ship they fire directly at, but since an aoe-immune ship is likely to be in a pile of other aoe-immune ships most of the dps is wasted when there are non-aoe-immune targets available.
  • Note that if the flaks are already locked on to something aoe-immune (like a missile frigate) and something non-aoe-immune (like a fighter) comes into range, the flaks will probably take a few seconds to retarget since the game only does the target logic every so often.

• Ships with engine damage now automatically repair their own engines, such that it takes a bit over four minutes to go from 0% engine-health to 100% engine-health.
  • Basically, until now it's been too easy to long-term-disable AI ships via engine damage. Not many players abused this thoroughly, but some definitely did, to hilarious (and disappointing, from a challenge perspective) results.
  • But tactical usage should still be very viable for kiting around a planet, spreading out an incoming attack, etc.

• The logic for when a carrier is popped early now tries a lot harder to produce something similar to what's actually in the carrier, and tries to avoid exo-like spawning logic.
  • Regardless of what it winds up spawning, it no longer scales up with difficulty level (since the ship numbers that led to the generation of the carrier were already a result of difficulty level).
  • If its contents fit under a certain threshold (under 150 if there are < 2000 AI ships on the planet, under 50 if there are < 3000 AI ships on the planet, etc), it just spawns whatever is in the carrier.
  • Else, it promotes all triangle/bonus fleetships in the carrier to max mark (mark 5 for most types, mark 4 for some) and scales down the quantity accordingly (mkI to mkV divides quantity by 5, etc).
  • If that's still not enough to get the quantity under the threshold, it divides the quantities in the carrier by some number (determined by how badly over-count it still is) and uses that extra strength to add mkV guardians to the carrier instead. Since a single mkV guardian is worth about the same as ~2.2 full caps of mkI fighters, that's pretty good compression.
  • If that's somehow still not enough, it falls back on the exo-like population it's used in the past, but without the extra multiplier from difficulty.
  • For the curious, having advanced logging on when a carrier does early-pop logic like this will generate a log entry in the corresponding log file in your RuntimeData directory, explaining the computation.

• Engineers have had a "what to assist" logic overhaul such that:
  • If is Prefer-Non-Military toggle is on, and one is military and the other is not, prefer the non-military one.
  • If one has more currently assisting engineers than the other, prefer the one with less.
  • If the repairer is not a teleporter and the difference in distance to each unit is larger than about 4000 range units, prefer the nearer.
    • Also, distance is now considered zero if the target is already within range, which helps a lot because previously it was masking other considerations.
  • If one object is a damaged irreplaceable unit and the other is not, prefer the damaged irreplaceable.
  • If one object is in stand-down and the other is not, prefer the one not in stand-down.
  • If one object's remaining-health-percent is less than half the other's, prefer the lower-health one.
    • This was being considered before (it always went for the lowest percent-health), but only by teleporting repairers.
  • If one object has taken more than 25% engine damage and the other has not, prefer the engine-damaged one.
  • If one object's metal+crystal cost is more than twice the other's, prefer the more expensive one.
  • Thanks to Wanderer and the other voters for inspiring these changes.

• Human-player teleporting units now use normal engines when group-moving.
  • Along with this, increased the normal-engine speed of all teleporting units to the speed of a fighter, for those lower than that

• Advanced Fabricators and Fabricators now use the same "don't put this near a wormhole" logic as human home command stations.

• Removed energy cost from Logistics and Military command stations, since the cost was pretty minor and the units certainly aren't in danger of overshadowing the econ station at this point, and the costs were interfering with the logic for rebuilding their remains in some cases.
  • Also removed it from warp-jammer stations, but since that wasn't so minor also increased that station's direct per-second m+c cost.

• Fixed a longstanding bug where attack-move units with no other orders would auto-target stuff that should not be auto-targeted (wormhole guard posts, AIP-on-death structures, etc) if the unit in question shot them first.

• Fixed a bug where Spire Blades, Spire Minirams, and Spire Rams would not actually automatically die (but just self-damage by some amount) when attacking.

Balance Updates

• Exogalactic strikeforces provoked by the Broken-Golems-Hard or Spirecraft-Hard minor factions (but not Fallen Spire ones) now have a chance (50% for Broken-Golems, 20% for Spirecraft) of using an alternate composition distribution that heavily favors a large lead ship and a single battlegroup.

• Broken-Golems-Hard and Spirecraft-Hard exogalactic strikeforce initial-strength, per-minute-increment-amount, and maximum-strength all increased roughly 10%.

• The Armored, Black-Widow, Artillery, and Regenerator Golems used in exogalactic strikeforces are now different than the "standard" AI golems (like those used by the Golemite AI Type). The main difference is that they lack the 1/10th-health-of-a-normal-golem rule (but the Armored variants still have a 1/2 health modifier, since they're so buff).
  • Thanks to Shrugging Khan for continuing to provoke the golems.

• Black Widow golem health and Attack Power boosted a bit (15%-ish).

• Regenerator Golem health doubled, in response to various concerns that they're less useful than other golems.

• The "shot travels at least 40 speed faster than its target's current speed" logic has been extended to include "shot travels at least 40 speed faster than its target's tractor-er's current speed", since in the "tow" case the target's current speed is almost always zero.

• Spire Blade Spawner:
  • Ship cap from 0.05 of normal to 0.03 of normal (in practice, from 9 to 5 for mkI). Note that this impacts AI usage as well as human.
  • BaseAIPerGuardPostShipCap from 2 => 1.

• Spire Gravity Drain:
  • Grav range from 3000*mk => 8000 (flat).

• Spire Gravity Ripper:
  • Attack power from 1000*mk => 2000*mk.
  • Seconds-per-salvo from 1 => 2.
  • So basically the gravity ripping effect is being nerfed to 50% of what it was, and the normal attack is better against armor. We'll see where to go from there.

• Spire Miniram:
  • Ship cap multiplier from 0.25 => 0.20.

• Spire Stealth Battleship:
  • Effective Attack range from 6500 + (500*mk) => 7000 flat (thus, they can no longer engage while still protected by their 8000 radar dampening range).

• Spire Tractor Platform:
  • Hull Type from Turret => Heavy.

• Previously, turtle AI types had over 13 times as many AI Eyes as "normal". Moderately defensive ones (Shield Ninny, Grav Driller, and Peacemaker) had only 2 times as many as normal. Turtles now have about 2.6 times as many as normal.

• Light Starship:
  • Base Health from 1.5m => 2m.
  • Armor Rating from 500 => 1000.
  • Given Radar Dampening of 8000 (to make it harder to deny a fleet its fleet starship munitions boost by ganking the fleet starships from long range). For reference, its munitions boost range is 3000.

• Flagship:
  • Base Health From 4,875k => 6m.
  • Given Radar Dampening of 8000. For reference, its munitions boost range is 6000.

• Neinzul Youngling Commando:
  • Base Health from 6600*mk => 8000*mk.
  • Base Attack Power from 1000*mk => 1200*mk.

• Impulse Reaction Emitter:
  • Now has a minimum multiplier of 5 (roughly what it would get vs a ship with 5,000 energy cost). This might be excessive, we'll see, but the ship was having significant "used as a paperweight" issues against anything at all small.

• Zenith Mirrors:
  • When a shot is reflected, the returned shot's power is now increased 4x. Thought to be a perpetual motion machine until the first power bill arrived. More seriously, this is an attempt to correct for the fact that things now tend to have a higher ratio of hp to damage (particularly against their own hull type) than when the mirrors were first added.

• When a ship is translocated away from its current location, all incoming projectiles now are also immediately translocated to its new destination, and thus will hit it instead of fizzling. This makes the military command stations a lot less unwieldy.

• Raid starship rebalance, thanks to many players for various suggestions:
  • Raid Starship health has been increased from 1.6 million health * mark level to 2 million health * mark level.

• Guard posts, fortresses, superfortresses, force fields, and exo-shields are all now immune to regeneration from regenerator golems/trains.

• AI Motherships are now considered large ships for purposes of bomber starships and siege starships being able to hit them.

• When an engineer is repairing a ship, and it was not set to repair that ship explicitly by the player, and it's not in free-roaming defender mode or attack-move mode, it will no longer try to chase the ship it is repairing. Instead it will just find a better target and keep its position. This allows for players to exercise more tight control over their engineers now that engineers have a huge repair range and no longer teleport.

• Astro Trains are now immune to translocation and gravity effects, since both of those cause more problems for players than they would ever solve

• The radar dampening range of marauders and resistance fighters has been changed to be about 1000 greater than their attack range in all cases, soo that they always have to come into range of their targets in order to fire on their targets. This makes it so that turrets and other fixed-position defenses are not completely powerless against them.

• Spire Civilian Leader Outposts are now included in the galaxy map filter for Detected AI Progress Reducers.

• Spire Civilian Leader Outposts previously did not have their minor faction stance properly set, so they could not be attacked. Now they are properly noted as AI Allies until captured by the human team. They are still direct-only to attack, though, so unless you accidentally mis-click on them your ships won't attack them.

• Armored Warhead health reduced from over 2 billion to 40 million. This is still a huge amount of health, but they are no longer infinitely survivable. This also fixes a bug where the interface was having an overflow exception and rendering their health as 1.

• When planet-wide EMP or tachyon detonations occur, as from a warhead or an EMP guardian, a chat message is now sent to the human players telling them who detonated what, and where. This message is also clickable. The goal here is preventing EMP guardians from punching QUITE so impenetrable a hole in player defenses without notice.

• Warhead Interceptors have been nonfunctional since slightly before 5.0 of the game (I disabled them, but never did get around to re-enabling them. They should work once again, but I haven't had a test case to work on it to be sure. If anyone has a save with them not working after this version, please let us know.

• The efficiency of gravitational slowing (gravity drains, gravity turrets) has been greatly improved, such that big battles that previously would grind to a halt because of them can now be played at 3x their prior framerate in one test case in particular.
  • One major improvement was to exclude checking of gravitational slowing for ships that can't move or aren't moving at the moment (since they don't need to be slowed!), and the other was to skip redundant gravitational slowing checks for gravitational sources that are the same and which are too close together. This last is very similar to how the speed improvements on the range data display (from way back pre-2.0 days) works.

• The devourer golem is now completely invincible—it was never meant to be killable in the first place, but recent changes to the game made that so that it was possible post-5.0. Like the dyson sphere, the devourer is more intersting when it is indestructible.

• Minor factions now shoot low-power ships without prejudice, which means that devourer golems actually will devourer more again.

• Zenith SpaceTime Manipulators, like forcefields, now have the ability to move very slowly but not to go through wormholes, to allow for better repositioning of them.

• Gifting a ship now copies across its current attack recharge, cloaking recharge, and similar, preventing a few exploits related to that and penetrators in particular.

• Dyson Gatlings are now immune to paralysis attacks.

• Slightly improved the targeting handling when a ship is targeting a teleporting ship that is out of range.

• Spirecraft Penetrator balance:
  • Mark I moved from pysite to xampite. Mark II moved to ebonite, and now produce two per asteroid. All the other marks not moved around.
  • Penetrator health reduced 8x.
  • Penetrator perma-cloaking replaced with normal cloaking

• AI ships that are spawned in response to saboteurs, such as the human mark III science lab, are now zombie bots

• Zenith SpaceTime Manipulators can no longer be swallowed or regenerated.

• Human-allied zombies and minor factions were not being properly limited in their pool of planets to fix, thanks to a typo in their logic. It should now be fixed.

• Zenith Electric Bombers:
  • No longer use scaled caps (due to the 0.1 ship cap multiplier, that can make the additional /2 or /4 from normal and low caps cause a more significant truncation than is desirable).
  • Event attack cost (used for exogalactic strikeforce computations) tiers increased from "Medium Corvette" through "Heavy Frigate" to "Light Frigate" through "Medium Destroyer", since the previous placement was putting mkIII zelecs at the same place as a mkV bomber. It's not a goal to balance the event attack costs terribly precisely, but that was a bit excessive.

• Sentinel Frigates:
  • Event attack cost (used for exogalactic strikeforce computations) tiers increased from "Heavy Corvette" through "Light Destroyer" to "Light Frigate" through "Medium Destroyer".

• Spire Blade Spawners:
  • Now have an attack again (not having an attack was causing all kinds of bugs and unintended behaviors), with a base power of 1000 per shot, 8 shots per salvo, 2 seconds per salvo, and effective range of 6000 (for reference, that's the same as the spire stealth battleship except it's 25% the base power and 1000 lower range).
    • Notably, this should make the BaseAIPerGuardPostShipCap flag work correctly again for them (it's set to 1 for them).
    • But since they were unable to guard stuff in the past due to zero attack, to correct old saves the AI ones just being completely stripped out when loading something from 5.009 or earlier. Human ones are not affected.
  • Have the new DoesNotMoveToPursueUnlessOrderGivenByPlayer flag, which prevents human ones from chasing an auto-targetted target (in FRD or not). Hopefully this will deal with the complaint that lead to the original removal of their guns.

• Made Cloaker Starships mark I knowledge-free now, instead of costing 1k knowledge. This helps to compensate players a bit for the generally-increased difficulty of the Zenith expansion.

• Same with the Neinzul Rnclave Starships mark I. They were 2k knowledge, but now they are knowledge-free.

• Mercenary Neinzul Enclave Starships, which would now be pretty well pointless, have also been revamped. Now they are the equivalent of the mark II Neinzul Enclave Starship, rather than the mark I starship. This is very powerful, but they are also now limited to only a single ship in their shipcap, rather than 6 ships.

• Starship costs tweaked around a bit:
  • Bomber starship metal cost reduced from 120k to 80k.
  • Fleet starship metal cost reduced from 60k to 40.
  • Siege starship crystal cost reduced from 120k to 80k.
  • Riot Control starships crystal cost reduced from 80k to 60k, and metal cost reduced from 30k to 20k.
  • Enclave starship costs reduced from 30k metal and crystal to 20k each instead.

• Zentih SpaceTime Manipulator energy costs reduced to 20k from 60k

• Due to widespread feedback that the martyr is very OP, we've drastically increased the asteroid "cost" of building them:
  • Previously Reptite could be used to build 2 MkI Martyrs. Now it cannot build martyrs at all.
  • Previously Pysite could be used to build 4 MkI Martyrs or 2 MkII Martyrs. Now it can only build 1 MkII (can't build MkIs).
  • Previously Xampite could be used to build 4 MkII Martyrs or 2 MkIII Martyrs. Now it can only build 1 MkIII (can't build MkIIs).
  • Previously Ebonite could be used to build 4 MkIII Martyrs or 2 MkIV Martyrs. Now it can only build 1 MkIV (can't build MkIIIs).
  • Previously Adamantite could be used to build 4 MkIV Martyrs or 2 MkV Martyrs. Now it can only build 1 MkV (can't build MkIVs).
  • Note that the units themselves are just as incredibly powerful as before (we really want them to be highly useful), but hopefully this change puts their real cost more in line with their utility and also puts a saner limit on the total number the players can have in a game. More changes may follow if necessary.
  • For reference, previously the total number of martyrs that could be produced from all asteroids in an average 80-planet galaxy was about 2200 on average (so someone holding 15% of the galaxy might reasonably field 220 over the course of the game, if they didn't use many other spirecraft). Now that total-possible number average about 480 (so that same martyr-happy player would probably only have about 48 to use, unless they take more planets).

• Zenith Viral Shredders:
  • Simplified the scaling of replication costs (how much damage a shredder has to do to replicate) and made it better at preventing game-breaking swarms of shredders. Now instead of it using a linear multiplier based on the owner's number of shredders on that specific planet, it quadruples the replication cost for every full cap of that mark of shredder owned by that player anywhere in the galaxy (but if they own less than or equal to a full cap, there is no extra cost).
    • So if the mkI cap is 98 (normal cap scale) and you have between 99 and 195 (inclusive) mkI shredders they will have to do 4x as much damage to replicate as normal. If you have between 196 and 293 (inclusive) mkI shredders they will have to do 16x as much damage to replicate as normal. And so on (maximum multiplier is 1024 to avoid arithmetic overflow; at that point a mkIV on low caps would need to do over 376 million damage to replicate).
  • Base Energy Use from 100 => 50 (note: all mkI types use half normal energy).
  • Replication-created shredders now copy the FRD, Attack Move, and any unit commands (movement waypoints and whatnot) of their "parent". Note that they already copy the control groups and "am I currently selected" status of their parent.
  • The end result of all this should be that if you play your shredders well you'll be able to maintain maybe 2 or 3 caps worth of them (a pretty significant advantage over other bonus ship types) but it's going to be difficult to get into the really absurd numbers. On the other hand, maintaining that swarm won't cost as much energy as it used to and it should require less micro due to the commmand copying.

• For a very long time it's been possible to ward off almost any amount of damage through sufficient micro of a pile of forcefield generators and engineers (to repair the ff's as they collapse within the protection of other generators, or go on low power mode, or are being newly built). While ff's are supposed to be extremely helpful in protecting stuff, it wasn't intended that they be able to do so indefinitely while under constant attack. So:
  • When a forcefield is damaged by enemy fire, all allied forcefields whose fields are in contact (or very close to it) with the damaged ff will take 1 point of damage from the energy conducted along the surface of the field. The damage itself is inconsequential since the ff's have millions of hp, but it will trigger the "cannot be repaired less than 6 seconds after being damaged" logic. This also applies to forcefields currently under construction or in low-power mode. So it won't be possible to repair some forcefields in a ff-pile while others in the same group are under fire.
  • Forcefield generators now cannot be assisted during construction if they have been damaged recently.
  • When a new forcefield is placed for construction, if it is in near any allied forcefields (low power or not) that is currently unrepairable due to recent damage, the newly placed forcefield inherits the longest assist-delay present among those neighbors. So it won't be possible to speed-build new generators under other fields that are currently under fire.
  • This also applies to all the "shield bearer" and module type shields.
  • If this leaves some situations unreasonably hard other changes can be made to compensate, we just wanted to close this "micro to dodge anything" loophole.

• Due to very widespread feedback that ion weaponry is very underpowered, all Ion Cannon and Spirecraft Ion Blasters now fire 4 shots per salvo instead of 1.

• For the ships listed below, armor piercing increased from "really high" (6 figures) to 999,999. They're really supposed to just flat out get through any armor, excepting maybe some of the ships so large we don't mention them in release notes because it would be a spoiler.
  • Zenith Polarizer
  • Spire Armor Rotter
  • Youngling Vulture
  • Youngling Nanoswarm
  • Raid Starship
  • Lightning/Armored/EMP/Tachyon-warheads
  • Botnet Golem
  • Devourer Golem
  • Sentinel Frigate
  • Spire Blade
  • Spire Gravity Drain
  • Spirecraft Siege Tower
  • Sniper
  • Electric Shuttle
  • Lightning Turret
  • Sniper Turret
  • Spider Turret
  • And a variety of lesser known but related types.

• AI Hunter/Killers are now eligible to be chosen for exogalactic battlegroups. The MkI is considered one notch above the Armored Golem cost-wise, and proceeding up from there. For anyone who remembers the slaughter those things could perform before, their number of shots-per-salvo has been halved but their health has been doubled.

• Added a new ability to the Spirecraft Attritioner in hopes of making it more generally useful. From the new tooltip:
  • One unique side effect of Spire attrition technology is that when an enemy ship is destroyed by any cause all its allies (that aren't immune to attrition) take a certain percent of that ship's maximum health as feedback damage. This feedback damage is cumulative with multiple Spirecraft Attritioners and higher mark attritioners cause a higher percentage, but this feedback damage is divided evenly across all affected targets and thus does not greatly damage any one target.
  • Basically, this functions as a small boost to overall fleet dps (averaged out over time), and thus makes the attritioner useful in short engagements where previously its contribution would have been minimal.
  • Current percentages are 0.5%/1.0%/1.5%/2.5%/5.0% for the 5 marks of attritioner. Also, this effect does trigger on enemies killed by attrition/feedback (otherwise it motivates micro to make sure a triggering-source kills big targets). In our tests this has not seemed overpowered though it is likely that further balancing will be needed. Bear in mind that mkV attritioners require the extremely rare Titanite, and mkIVs require the not-much-more-common Adamantite.
  • Thanks to Hearteater and many other players for weighing in with feedback that led to this change.

• Spirecraft Ion normal shot damage from 700 => 14000, and armor piercing from 0 => Max.
  • Previously their damage was just token, but this brings them up to roughly the DPS of the Spirecraft Siege Tower (which, for reference, also has max armor piercing), but can't target anything that's immune to insta-kill (which is basically anything bigger than a fleet ship). The result is that a cap (8) of MkI Spirecraft Ions is capable of doing well against a single cap of even MkV fleet ships unless those fleet ships have a bonus against them (bombers, chameleons, raiders, IRE's, etc), and so on. In general, a lot of these together can put some serious hurt on fleet ships (particularly low health ones) even when they can't insta-kill.
  • Related note: since their seconds-per-salvo is higher on the lower cap scales, and since their shot damage now actually matters, and since just about everything they can target scales with the caps, their damage will now be scale upward on the lower caps.
  • Updated the tooltip, which is reproduced here since it probably helps understand what we're doing here:
    • Like Zenith Ion weapons (which AI Ion Cannons use), Spire Ion weapons insta-kill most ships with a mark level equal to or lower than its mark value, and cannot fire at all on ships immune to Ion weapons (i.e. immune to insta-kill). Unlike Zenith Ion technology, these ships are mobile, much shorter-ranged, and actually do decent damage to higher mark ships.

• Fallen-Spire city-provoked exogalactic strikeforces "buildup" rate (this only impacts the time between attacks, not their size or composition) redistributed a bit:
  • Spire City Hubs now generate 1/4 the "aggro" they used to.
  • Spire Shard Reactors, Habitation Centers, and Shipyards now, instead of generating zero "aggro", generate 1/8th what the Spire City Hub used to generate.
  • So a city with all 6 building slots taken up causes exactly the same amount of buildup as before.
  • But the advantage is that if your cities get really beat up in an AI attack and you lose city structures you have more time to rebuild before the next exo. Not exactly smart of the AI, but the alternative is often a game that is fundamentally lost but just a matter of time.

• Youngling Nanoswarm:
  • Fixed bug in targeting logic that was making it focus on single targets when the intent was that they would aggressively spread themselves out across all available targets. The new behavior still doesn't reach quite the desired behavior (they still tend to only spread out over a subset of the available targets) but it's still a significant improvement. Hopefully more can be done later.
    • Note that this overrides the Non-Sniper-Focus-Fire control; there's very little reason one would ever want these the auto focus-fire since they do so very little damage. Of course if you really want a bunch of nanoswarms to all attack the same target you can give them a normal attack order.
  • Explosion range from 200/225/250/275/300 => flat 1000.
    • They still only hit a max of 3/5/7/9/11 targets when they explode, so this makes it more likely that they'll actually get multiple hits.

• AI Fortress health from 6x human fortress equivalents => 3x human fortress equivalents.
  • Previously it was easy to take out an AI Fortress with just bombers (and other polycrystal stuff, since fortresses can barely scratch the paint on polycrystal) but it could take a really long time. We want the fortresses to be challenging but not in the "how long do I have to wait before it dies" sense. Hopefully we can find some other more interesting (and not very time consuming) way of making them more challenging

• Ships that generate attack (munitions) boosts now show the magnitude of the boost in the tooltip, not just the range.
  • Thanks to SpaceBrotha for the question that reminded us that this info wasn't directly displayed anywhere.

• Ships that are not immune to attack boosting but have a attack-boost-ceiling below the default (which is 4, i.e. 5x or 500% of normal damage) now display their maximum receivable boost in their tooltip.

• Higher mark Hybrid Drone Spawners now spawn drones more quickly in addition to being capable of spawning higher mark drones (a mkIV drone spawner still only spawns mkI drones for a hybrid at the first maturity level, and so on).

• Hybrids: second-stage and third-stage maturity now takes less time (about half what it used to, still longer than first-stage), as previously most hybrids died either before or during the first stage.

• Neinzul Youngling Tigers:
  • Time-to-live from 2 minutes => 4 minutes, which is the same as other youngling types.
  • Metal and Crystal cost from 50/50 => 40/40.
  • Base Armor Rating from 300*mk => 600*mk.
  • Base Health from 11k*mk => 13k*mk.
  • Thanks to chemical_art for suggesting/inspiring these changes.

• Neinzul Youngling Vulture:
  • Metal and Crystal cost from 50/50 => 40/40.

• Spire Teleporting Leech:
  • Base Energy Cost from 400 => 200.
  • Base Metal Cost from 2200 => 1200.
  • Base Crystal Cost from 800 => 500.

• Science Lab MkII:
  • Base Health from 120k => 500k.
  • Base Armor Rating from 1500 => 5000.
  • Now has cloaking.
  • Long story short: they're still more expensive than MkIs, but now they actually are significantly easier to keep alive in dicey situations than MkIs.

• AI Superfortress:
  • Base health from 750 million => 450 million. This brings it from 5*human-version to 3*human-version, putting it in line with the other fortresses.
  • Health regen halved.

• AI Fortress I/II/III: health regen halved.

• Neinzul Preservation Warden spawn counts now scale up with AI Difficulty (the higher of the two is used, this is 1 at diff 7, 1.5 at 8, 2 at 9, 3 at 10) and AI Handicap (again, the higher of the two, negative handicaps don't apply to this), similar to other non-AIP-based threats. Previously it only scaled with the number of human players/homeworlds, and of course the number of human-controlled resource extractors.

• Sentinel Frigates and Zenith Electric Bombers are now considered large enough targets for ships like the Antimatter Starship.

• The earliest time that an AI player can even check for sending a wave has been changed to avoid both AIs sending their first waves on the exact same second on higher difficulties (which artificially increases the difficulty of the first wave, giving a false impression of how hard the second will be). Specifically, it's still (( 14 - Difficulty ) * 100) seconds for the first AI player, and 1.5 times that for the second AI player.

• Black Widow Golem:
  • Number of tractors from 100 => 200.
    • Tractors on non-scaling ships do need to be made to scale with unit caps, but we don't have time to implement that just yet. Bear in mind that when that scaling comes, the current number of tractors will be the _highest_ value: corresponding to "high" caps, because that's what the caps were before scaling was added (when the tractor numbers were established to begin with), so it won't make them any better. For now, enjoy :)
  • This golem's tractors now do 2 seconds of paralysis damage per second to each held ship. Ships immune to paralysis are not affected.
  • Base Health from 80M => 200M.
  • Basically just trying to make this golem a lot more useful and a lot less likely to die when not baby-sat.

• Some further adjustment to high-difficulty wave size calculation, this time not changing diff 10 much at all but making the 9.3-9.8 stretch build up more gradually rather than 10 being more than 2x as hard as 9.8.
  • Previously in step 8 any difficulty greater than 7 received a 2.5 multiplier (7 got 2.25, and lower the lower you went).
  • Now:
    • 7.3 through 9 still get 2.5.
    • 9.3 gets 2.75 (10% harder than before).
    • 9.6 gets 3.00 (20% harder than before).
    • 9.8 gets 3.80 (52% harder than before).
    • 10 gets 4.5 (would be 80% harder than before, but see next change).
  • Previously each wave on diff 10 would actually cause 2 separate waves. This was from an old rule where 8 through 10 got double waves, but that was changed to just 8 due to the challenge-cliff it caused between 7.6 and 8. Now this rule has been entirely removed, since 10's waves are now so much larger than before.
  • We're very much open to more feedback on this; basically diffs 9.3 through 10 aren't really where we expect people to play (indeed, we may need to make them harder if someone actually beats 10 in a fair fight again), but we want it to be fun for those who do.

• Previously Sniper and Spider turrets cost 8 to 10 times as much metal+crystal to build a cap of compared to all other turret types, so:
  • Base Metal Cost from 1500 => 300.
  • Base Crystal Cost from 7500 => 1500.
  • They're still the most expensive turrets, that's intentional, but they're no longer quite so punishing (before you could have built a fleet of starships for the cost of a cap of sniper turrets).

• Antimatter Starship:
  • Yes, this one is getting _another_ revamp. Some background:
    • First, they were called dreadnoughts. They did tons of engine damage. Too much. Nerfed.
    • Then, they were called useless. For a long time.
    • Then, we made them hit like a train from long range, but unable to fire upon anything remotely small. They were called Siege Starships (alias: "awesome").
    • But this broke a lot balance because you could just camp around your siege starships from across the planet and bomb everything to bits.
    • So they were given antimatter bombs that can't hit forcefields, fortresses, or... well, much of anything. In fairness they do great against starships and guardians, but that's proven too small a niche to really please their users. They were renamed "Antimatter Starship", which has since become a dire insult in 75% of the inhabited galaxy.
    • But the solution was unclear: we could shorten their range, but Bomber Starships (which were made normally available shortly after these were changed to Siege Starships) already fill the short-range big-stuff-killing niche.
    • And now, the next episode...
  • Base range from 16000 => 4000.
  • Shot type changed back from Antimatter Bomb => Energy Bomb (same basic type as bombers use, so no one is specifically immune to it like some are to antimatter bombs). Still cannot fire on small ships, like before.
  • If it hits a forcefield, it will also damage up to 200 units protected by that forcefield (or forcefields physically near that forcefield) with 1% (to each target) of the damage done to the forcefield.
    • Note: even stuff the starship itself cannot aim at (small units and cloaked units, etc) can be hit by the splash damage.
    • The 200-count doesn't change between low, normal, and high unit caps because that would change the dps of the unit pretty significantly. If it proves to be a problem we can try a few things.
  • Has been renamed to the Plasma Siege Starship.
  • The end result:
    • Still longer-ranged than the bomber starship, but much shorter than before, so not worrying that it will break balance in that way.
    • Still only about 1/3 of the bomber starship's single-target dps, but can up to triple the total damage output in a "siege" situation against a forcefielded group of targets. Still not the better choice in terms of raw damage, but the utility of being able to get at the protected targets should counterbalance that and create its own niche.

• Plasma Siege Cannon Modules (the ones for Fallen Spire modular stuff) now also have the forcefield-splash-damage attribute. Thematically, the Siege starship is (using) a plasma-siege-cannon now.

• Zenith Beam Frigate Base Range from 7500+100*mk => 6500+100*mk.
  • They're supposed to be long-ranged, but some marks were outranging their main triangle counter (missile frigate, which has 6500+500*mk base range), which isn't appropriate given the beam frigate's very high damage when it hits several targets per shot.

• Turret base energy costs:
  • Basic Turret from 150 => 100.
  • MLRS Turret from 150 => 100.
  • Flak Turret from 300 => 200.
  • Missile Turret from 150 => 100.
  • Counter-Dark-Matter Turret from 4000 => 1000.
  • Counter-Missile-Turret from 4000 => 1000.
  • Counter-Sniper-Turret from 4000 => 1000.
  • Sniper & Spider Turrets from 250 => 150.
  • Grav Turret from 500/8000/10000 => 400*mk.
  • Tractor Turret left alone (150/240/300).
  • Laser Turret (150) and Heavy Beam Cannon (750/1300/2000/20000) left alone. They're energy weapons.
  • Lightning Turret from 500/1000/2000 => 400 flat. Energy weapon, yes, but that was a bit high.
  • These really aren't that big a deal, but it does help rein in some numbers that were set long ago, and bring the energy-per-dps ratio of the main attack turrets under the stuff that shoots and has engines.

• Artillery Golem:
  • Base Armor Rating from 0 => 4000.
  • Base Recharge Time from 30 => 8.
  • Basically: when you see an Artillery Golem you can get on your side, we want you to be happy. When you see an Artillery Golem on the enemy side, we want you to be sad. Previously it wasn't doing a whole lot of either.
  • Thanks to Orelius and others for feedback on golems.

• Regenerator Golem:
  • Repair cost reduced by 90% to make regeneration not "cost" quite so much more than simply producing the ships from scratch, in metal+crystal terms.

• Cursed Golem:
  • Self-attrition reduced by 50% (now would go from 100% to 0% in 60 minutes instead of 40).

• Lightning Turrets
  • Attack changed to function like electric shuttles: Hits up to 200 targets, and if hits are left over can hit each target up to 5 times (meaning that groups as small as 40 take the full power of the attack).

• Lighting Turrets/Electric Shuttles: The per-planet-per-player-per-type-per-mark delay has been reduced to a very minimal duration, generally allowing a full cap of lightning turrets to unload its full power before the first one finishes reloading.
  • This still prevents a total alpha strike, but that also helps avoid a situation where the entire lightning turret cluster unloads on a couple of leading fighters before the rest of the wave shows up.
  • It also looked really cool in testing.

• Player-Ally Dyson Gatlings are no longer unable to shoot mkV units.
  • In previous iterations this rule was necessary because player-ally dysons were clearing off AI homeworlds and/or core worlds. Since player-ally dysons rarely wind up on AI planets anymore the rule no longer seems necessary, and "why are my dysons not shooting X?" questions have been happening on a regular basis for quite some time.

• Hybrids:
  • If either player has Advanced Hybrids enabled, hybrid maturity-xp contributions to the central dirty-tricks "bank" are now doubled compared to Normal Hybrids.
  • The first couple stages of the only such dirty-trick progression currently implemented now happen a bit faster, but the first and second nasty bits have been made less nasty

• Spirecraft Shield Bearers:
  • Ship cap doubled.
  • Amounts produced:
    • Reptite: from 0 => 1 mkI (so can now be built from Reptite)
    • Pysite: from 1 => 2 mkI
    • Xampite: from 1 => 2 mkII
    • Ebonite: from 1 => 2 mkIII
    • Adamantite: from 1 => 2 mkIV
    • Titanite: from 1 => 2 mkV
  • Hopefully this will bring their costs more into line with their actual utility (as non-repairable shields).

• Plasma Siege Starship
  • Shot now has a traditional (lightning-type) aoe that does 6.25% of normal damage to up to 14 extra nearby targets.
    • In order for this normal aoe to work, the restriction from firing upon fleet ships has been removed. Previously it was there primarily to keep the autotargeting from "wasting" shots on small stuff, but this has become less important. It still prefers targeting starships, etc.
  • Still has the "siege" effect when hitting a forcefield, but that's been changed from 1% of the damage done to up to 200 ships to 6.25% to up to 25. That's less total, adjusting somewhat for the addition of the traditional aoe and the greater ease of getting the whole bonus.
  • For reference, the Siege's dps (per unit, not per cap) against a single target is about 71% that of the Bomber Starship (previous version notes said 1/3rd, that was actually inaccurate). If the traditional aoe hits the full 14 targets, that brings it up to about 133% that of the Bomber Starship. If the shot also hit a forcefield that's protecting other units, that extra "siege" damage is just gravy on top (potentially up to a total of 245% or so).
    • If this proves to be too much, adjustments can be made, but typically we get more useful feedback for a new/changed unit that's stronger than it should be than if it were weaker.

• AI Carriers:
  • Are no longer invincible if there's >= 2000 AI ships already on the planet.
    • Previously this was necessary to avoid too many AI ships on a planet at once for decent cpu/ram performance.
  • Now, when a carrier deploys (it deploys when destroyed):
    • If there are fewer than 1400 AI ships already on the planet, and another carrier has not deployed in the last few seconds, it deploys normally as before.
    • Otherwise, it "combines" its contained units into a fairly small assortment of ships and deploys those. The population of this group uses similar logic as exo waves, but is not itself an exo wave (there's no leader, no beeline-human-homeworld logic, etc), it behaves as normal threat.
      • The more AI ships there are on the planet, the more "condensed" it tries to make this deployment, so in those cases you'll see more enemy starships, etc but far smaller quantity.
  • Also, if there are less than 500 AI ships already on the planet, and it is a non-AI planet and there's at least one enemy military (non-mine) unit, and another carrier has not deployed in the last few seconds, it automatically deploys.
  • Rationale: Previously it was possible to have carriers slip through otherwise impenetrable defenses because there were simply too many AI ships (which may have been somewhere else on the planet) to quickly kill. This was desirable from a design standpoint because we don't want a single layer of impenetrable defenses to be the end-all defensive strategy in the game. However, it has proved consistently frustrating (in a non-fun way) for some players so it seemed a good time to try something different. Now the presence of a ton of other AI ships doesn't make the carriers invincible, it just makes them very dangerous to destroy. Hopefully in an interesting way :)

• Spider turrets:
  • Base Metal Cost from 6000 => 1200.
  • Base Crystal Cost 15000 => 2400.
  • These weren't actually changed in last patch because they weren't inheriting the sniper turret costs. This is actually a good thing as spider turrets are substantially more useful than sniper turrets. They now cost 2x as much m+c as the sniper turret (3600 vs 1800).

• Riot Starships:
  • Ship Cap from 4/3/2 => 4 flat. Most of the special starship types have non-decreasing caps already.
  • Base Health from 900k*mk => 1.8M*mk. This brings their just-the-hull health up to just below the squishy end of the current target zone for starships (7.5M*mk cap-hp, though Raids and Leeches are currently way below that for other reasons).
  • Shield mkI/mkII Base Health from 500k*mk => 1.8M*mk. This brings their hull+shield health up to the middle of the target zone.
  • Base Speed from 20/17/14 => flat 20.
  • Base Energy Cost from 3000/4000/5000 => flat 3000.
  • Base mkI Metal+Crystal Cost from 20k+60k => 30k+50k (mkII is 2x, mkIII is 4x).
  • Knowledge Cost from free/3500/4500 => free/4000/5000. DPS starships run 0/5k/7k, more specialized ones like the raid run 0/4k/4k, etc.
  • Base Attack Power from 3600/5200/6800 => 4000*mk. This brings their non-bonus-dps up to the very low end of the current target zone for starship (20k*mk cap-dps). So it's really low (even with module damage), but it's a utility type.
  • Base Engine Damage from 40/55/70 => 40*mk.
  • Multiplier vs Medium from 2 => 4. Its other 4 bonuses were already 4.
  • All Modules:
    • No longer have an energy cost. 12k cap-e for the hulls is enough.
  • Machine Gun Modules:
    • Engine Damage Percent Floor from 50%/30% => 40%/20%.
  • Laser Modules:
    • Engine Damage from 17*mk => 30*mk. The machine guns are still 3x as efficient for total EDPS but have shorter range and cannot bring engines as low as these.
    • Multiplier vs CloseCombat from 5 => 4.
    • Multiplier vs Neutron from 3 => 4.
  • Shotgun Modules:
    • Base Attack Power from 40 => 400.
    • Engine Damage from 28 => 36. Makes these 2x as efficient as MGs for total EDPS.
    • Shots per salvo from 25/49 => 25*mk.
    • Multiplier vs Swarmer from 6 => 4.
    • Multiplier vs CloseCombat from 5 => 4.
    • Multiplier vs Refractive from 5 => 4.
  • Tazer Module:
    • Multiplier vs Medium from 3 => 4.
    • Multiplier vs Artillery from 3 => 4.
    • Now has a planet-wide stagger time of 2 seconds, to prevent permanently stunlocking 4000+ ships in a multi-hw setting with 4 Riot IIs. They can still be used to effectively halve the attack power (due to prolonging reload time) of non-paralysis immune enemies. Also tends to interrupt them while they try to get away.
    • Thanks to Wanderer for being the one to confirm the long-held suspicion that these were horribly exploitable.
  • Tractor Modules:
    • Tractor Range from 1700*mk => flat 4400. Same as mkIII tractor turrets.
    • Tractor Count from 10/40 => 72*mk. Same as mkII/mkIII tractor turrets (but mobile and not needing supply).
      • Note that this doesn't scale with unit cap scale and really should, but that's a bit out of scope for this time frame and these would be the numbers on high caps so you're not losing anything.
  • For reference, total cap-EDPS with all MG/SG builds firing within main-gun range is now 3080/4720/7080, compared to the spider turret's 7833 (from infinite range, insta-fire, and able to take engines down to 1%, but immobile and way more expensive).
  • This was another long-overdue rebalancing, hence the magnitude of changes.

• Orbital Mass Drivers:
  • Now when under a firepower-reducing forcefield they retain 75% of their attack power instead of the customary 25%.

• Leech Starship:
  • Base Energy Use from 10k/14k/18k => flat 5k. So cap-e is same now as Raids (the next highest for combat starships)
  • All vs-hull-type bonuses removed, to allow for a bit more of an overall damage buff and thus more general reclamation utility.
  • Base Health from 1.6M*mk => 3.2M*mk. Was well below the target range for starship cap-health, is now just on the low-end (bottom is 7.5M*mk).
  • Shots per salvo from 1 => 3. That brings it up to the target range for starship-with-no-bonuses cap-dps (about 66k*mk), and should help it reclaim more.
  • Hull type from UltraLight => Heavy because UltraLight tends to be a major pain to kill when in the hands of the AI (see: raid starships) and the health just doubled.

• Light Starship/Flagship/Zenith Starship/Spire Starship/Core Starship
  • Ship Cap from 5/3/2/2/2 => 4 flat.
  • Base Speed from 15/18/18/24/22 => 22 flat.
  • Base Armor Rating from 1000/2500/2500/2500/2500 => 1000*mk.
  • Base Energy Use from 2k/10k/20k/20k/20k => 4k flat.
  • Base Attack Range from 1500/4800/3000/10000/500 => 3000+(200*mk).
  • Base Health from 2M/6M/14,625k/12,675k/25,350k => 3.75M*mk. This brings them to 15M*mk cap-health, and the target range for starships is 7.5M-40M*mk (though there aren't any "Armor Starships" or whatever that would actually use the top of that range, etc).
  • Shots-per-salvo from 5/9/14/4/5 => 5 flat.
  • Seconds-per-salvo from 4/4/2/2/5 => 4 flat.
  • Multiplier vs CloseCombat from 12/12/1/1/1 => 4.
  • Multiplier vs Medium from 8/8/1/6/1 => 4.
  • Multiplier vs Light from 6/6/1/1/1 => 1/1/4/1/4.
  • Multiplier vs Polycrystal from 1/1/0.01/4/1 => 4/1/1/4/4.
  • Multiplier vs Artillery from 1 => 1/4/1/1/4.
  • Multiplier vs UltraLight from 1/1/1/4/1 => 1.
  • Multiplier vs Swarmer from 1/1/1/1/0.1 => 1.
    • Part of this is to make each of the mkI-IV varieties good against one triangle ship type, and the mkV good against them all (not that you can get the mkV normally, but the AI can!).
  • Base Attack Power from 3600/12k/30k/60k/60k => 8k*mk. This brings them to 40k*mk non-bonus and 160k*mk bonus cap-dps, which is about right for a 4x-bonus starship.
  • Munitions boost range from 3k/6k/5k/8k = 2k+(1k*mk).
  • Most of the mark-specific quirks (flagships do a little engine damage, zeniths and up have warp detection, spires are immune to gravity, different ammo types, etc) are still there.
  • There've been a lot of balance problems with this line for a while, and we've held off from comprehensive changes partly since they're really two separate lines (fleet starship and alien starship) and we'll probably split them up at some point. We'll probably still split them up later since that would be cool, but for now it's easier to balance them as a cohesive line according to the guidelines used for the other combat starships.

• Flak Turrets:
  • Effective range from 4000/5000/6000 => 4500/5500/6500 to make them a little easier to place around a wormhole.
  • Now when under a firepower-reducing forcefield they retain 75% of their attack power instead of the customary 25%.
  • Hopefully they'll now be closer to the general usefulness of other turrets in that they don't have to die immediately (if protected).

• Lightning Turrets:
  • Now when under a firepower-reducing forcefield they retain 75% of their attack power instead of the customary 25%.

• Heavy Beam Cannons (just the human standalone ones, not any of the spire variants, though the spire variants were impacted by the m+c and knowledge cost changes) :
  • All marks:
    • All vs-hull-type bonuses removed, these are just straight damage now, which is more thematically appropriate. Also, it's nice to have at least one turret type that doesn't have that kind of bonus.
    • For reference, these aren't really "aoe", they're "blast-through"; a beam loses strength as it does damage.
  • MkI
    • For reference, MkI unit cap is 12 for these, and each fires 3 beams every 7 seconds.
    • Base Health from 250k => 850k. This brings cap-health a bit higher than missile turrets, which are the squishy end of turret types.
    • Base Energy Use from 750 => 1650. This brings cap-energy to just slightly above MLRS and several other turret types.
    • Base Metal/Crystal Cost from 8000/14000 => 6500/11000. This brings cap-mc to slightly below most turret types.
    • Base Attack Power from 50000 => 140000. This brings cap-dps to be second highest turret non-bonus-dps (lightning being highest) but well below the any turret's bonus-dps.
    • Knowledge cost from 2000 => 500. The other mkI turrets are free.
  • MkII
    • For reference, MkII unit cap is 8 for these, and each fires 6 beams every 6 seconds.
    • Base Health from 1.25M => 2.55M. Basically: mkI health * (2 for being mkII) * (1.5 for smaller unit cap)
    • Base Energy Use from 1300 => 2500. (roughly mkI * 1.5 for cap)
    • Base Metal/Crystal Cost from 16000/28000 => 13000/22000. (mkI * 2, customary for mkII)
    • Base Attack Power from 50000 => 180000. Basically: mkI damage * (0.429 for rate-of-fire-difference) * (2 for mkII) * (1.5 for unit cap)
    • Knowledge cost from 3000 => 2500. Most other mkII turrets are 2000-3000.
  • MkIII
    • For reference, MkIII unit cap is 4 for these, and each fires 12 beams every 5 seconds.
    • Base Health from 2.25M => 7.65M. Basically: mkI health * (3 for mark) * (3 for unit cap)
    • Base Energy Use from 2000 => 5000. (roughly mkI * 3 for cap)
    • Base Metal/Crystal Cost from 32000/56000 => 26000/44000. (mkI * 4, customary for mkIII)
    • Base Attack Power from 50000 => 225000. Basically: mkI damage * (0.179 for rof diff) * (3 for mark) * (3 for cap)
    • Knowledge cost from 3500 => 3000. That's pretty normal for a mkIII turret.
  • MkIV
    • For reference, MkIV unit cap is _1_ for these, and it fires 40 beams every 4 seconds.
    • Base Health from 3.25M => 40M. Basically: mkI health * (4 for mark) * (12 for unit cap)
    • Base Energy Use from 2000 => 20000. (roughly mkI * 12 for cap)
    • Base Metal/Crystal Cost from 64000/112000 => 52000/88000. (mkI * 8, customary for mkIV)
    • Base Attack Power from 50000 => 290000. Basically: mkI damage * (0.043 for rof diff) * (4 for mark) * (12 for cap)
    • Knowledge cost staying at 3000. Thought about 3500, but it really is just one unit.
  • These haven't seen a real rebalance in... a long time, and had fallen behind pretty drastically through the various changes since the 3.x days. Hopefully this will make them statistically competitive with other turret types (that have much larger caps). Feedback on how it all actually works out in practice is very welcome.

• All modules (except those that generate forcefields) are now invincible and thus untargetable, etc.

• When computing the number of reactors of a specific mark controlled by a specific player on a specific planet for the purposes of reactor-inefficiency, that number is now divided by that player's number of homeworlds (if that player has more than 1). This allows them to get basically the same energy output per m+c cost as an equivalent number of separate players with the same total number of homeworlds.
  • On the same note, the "auto build energy reactor" toggles on the CTRLS window (both galaxy-wide and per-planet) have been converted to sliders allowing you to specifying that more than one should be auto-built. If both the galaxy-wide one and a per-planet one are greater than zero, that planet will auto-build the higher of the two numbers (but not the sum, unlike with engineers).
  • The energy difference between multi-HW and multi-players has been known for a long time and not considered a priority. However, recent scientific analysis has concluded that multi-HW AARs have significant potential entertainment value, hence the change.

• Everything that fires Siege Plasma (Plasma Siege Starship, Plasma Siege Cannon modules), Flak grenades, grenades, or self-explodes (autobomb, nanoswarm) can now fire directly upon aoe-immune targets, but the resulting aoe will still not hit any aoe-immune targets (just the directly hit target).
  • Notably, this restores the Plasma Siege Starship's ability to target Raid Starships, which was unintentionally lost in the previous version.

• Youngling Nanoswarm:
  • Paralysis time from 1*mk seconds to 3*mk seconds. Partly because really short paralysis times work a bit oddly, partly because these needed to be more useful.
  • Engine damage from 2*mk => 10*mk.

• The "The Tank" AI-type wave multiplier from 0.5 => 1, as the intent behind it is that it be actually stronger on offense than usual, rather than weaker. The average should be fine, but possibly another increase can come later.

• Vorticular Cutlass:
  • Base Health from 7000*mk => 10500*mk. This puts them at the top of the target fleet ship range with 30M*mk cap-health (Armor Ships and Shield Bearers were the only other ones up there before this).
  • Base Speed from 16 => 54. So from slower than a fighter to slightly faster than an etherjet.
  • Base Metal/Crystal cost from 200/50 => 30/30. Now the same price as infiltrators, a bit lower than tele-raiders, etc.
  • Base Energy use from 100 => 20.

• Acid Sprayers:
  • Base Attack from 450 => 600. This puts the non-bonus cap-dps pretty exactly at the fighter's.
  • Base Movement Speed from 40 => 50. Now same as the ether-jet; this is a very short range ship and needs to be able to close the distance.
  • Vs Hull Type Multipliers from 6 => 8. The fighter's vs-polycrystal bonus is 5, for reference.
  • Base Metal/Crystal cost from 0/300 => 40/180. So 10% more expensive to build to cap than a fighter (though it's still twice as much energy).

• Riot Tazer:
  • Paralysis Time from 1 => 3 (internally 2 => 4, but the first second generally gets eaten due to processing order).
  • Reload Time from 2 => 8.
  • Per-planet stagger from 2 seconds => 6 seconds.
  • This helps prevent some transient (and difficult to reproduce) stunlocks due to processing order, while maintaining the ability to keep things locked down 50% of the time.

• Anti-Starship Arachnid V:
  • Has become Spider Bot V, which it has been internally since the beginning but had a bunch of special changes to make it an anti-starship unit.
    • Long ago this was a really important unit for the AI because it would spawn them in response to human usage of starships, but that hasn't been the case for rather a while.
    • In the meantime, they generally only see service if a human player captures a fabricator that produces them, and while the AI does use some starships generally these units are very, very low on usefulness to a human player because they literally can't hit anything that isn't a starship. So now they're just Spider V.

• Spider Bots:
  • Engine Damage per shot from 15 => 60.
    • This brings mark I cap-engine-damage-per-second to about 30% of a cap of Spider Turrets; a cap of mkII (which has fewer ships) to 50-55%, and so on.
  • Base Health from 3600*mk => 7200*mk. They were right at the flimsiest end of fleet ship cap-health (5M), now just in the lower 3rd (10M).

• Beam Starship:
  • Rebalancing as a MkV starship, since the fleet ship fabricators produce MkV stuff.
  • Base Move Speed from 14 => 22.
  • Base Health from 4.5M => 22.5M. Now on par with the Core Starship for cap-health.
  • Max targets hit per shot from infinite => 9. Same as Zenith Beam Frigate.
  • Seconds-per-salvo from 5 => 2. Not the same as Zenith Beam Frigate.
  • Base Attack Power from 28k => 44k. If it hits 9 targets every 2 seconds, that puts cap-dps at about 3/4 that of the Core Starship's bonus cap-dps.
  • All vs-hull-type multipliers removed; the "bonus" for ships like this is typically obtained by hitting more than one ship; max bonus is against large clumps where you can always hit the max.
  • Energy and m+c costs looked ok; energy a bit high, m+c rather low for mkV but since you have to get a fabricator to build these we can let it slide unless it becomes a problem.

• Warbird Starship:
  • Rebalancing as a MkV starship, since the fleet ship fabricators produce MkV stuff.
  • Made eligible for spawning in exos again; were originally removed from that because if they were lead ships they made the whole group go terrifast. Well, Raid Starships are even faster and are still eligible, so these are going back in. Generally these won't be picked to lead all but relatively early exos with lighter overall firepower.
  • Base Health from 3M => 17M. Now about 70% of Core Starship for cap-health (or: a bit over 10M*mk, where 7.5M*mk is about as low as any combat starship should go in the current approach).
  • From no vs-hull-type multipliers => 4 vs polycrystal, heavy, ultra-heavy, and structural.
  • Seconds-per-salvo from 5 => 2.
  • Shots-per-salvo from 40 => 16.
  • Base Attack Power from 2k => 8k. This brings non-bonus/bonus cap-dps to a bit under 200k/800k, very similar to the core starship in that regard.
  • Energy and m+c costs looked ok; energy a bit low, m+c rather low for mkV but since you have to get a fabricator to build these we can let it slide unless it becomes a problem.

• Paralyzers (specifically, zenith paralyzers, but also nanoswarms to some extent) are now much more aggressive about spreading out their fire across available targets to make better use of the paralysis effect.
  • It still will focus more than is optimal in cases where a group of more than 100 paralyzers (of the same mark, controlled by the same player, relatively near one another) has more than 100 non-paralyzed targets in range (they'll generally all focus on 100 of the targets in the first volley, and spread out afterward) due to our desire to not have target lists longer than 100 for RAM-consumption reasons. But it's a lot better than it used to be.

• AI Super-Terminals:
  • Previously these became active after you finished building a command station on the planet, and went dormant again as soon as your command station was destroyed. Now they never go dormant after becoming active: now, the only way to shut down an active superterminal is to destroy it.
    • Thanks to Atomjack for pointing out that it was too easy to keep turning the terminal on and off previously.
  • Now each super-terminal "pulse" has a 10% chance to be a "surge":
    • It puts a message in the chat log letting you know there was a surge.
    • It intensifies the pulse by (highest AI difficulty / 2) (so diff 7 means that pulse will be 3.5 as strong in terms of enemy ships generated).
    • It sets the time until the next pulse to 1 second instead of the usual 15. "Chain surges" are therefore possible, albeit fairly unlikely.
  • Previously when the super terminal reached a certain size of mkV pulse it simply didn't get any more dangerous; now it will start shortening the time between pulses as the strength goes past that point.
  • In all, super terminals were being a little too tame. It's great that proper playing of an ST can make the difference between winning and losing, but they seemed to need more danger and unpredictability to balance out their fairly powerful advantages. It's more like "riding the bull" now, and if it gets out of control it can kill you.

• Mapgen will no longer put the dyson sphere adjacent to an AI core world, because having a core world on alert the whole game was nearly an automatic loss on some more intense scenarios. It's better if you die because of something you actually had influence over.

• Science Labs I/II/III, the Survey Ship, and the new Ship Design Hacker are now all immune to the maw's swallow effect.

• Impulse Reaction Emitter, in recognition of winning the 3rd "worst ship" award:
  • Effective Range from 4200 => 6000.
  • Base Move Speed from 20 => 30. This makes it slightly faster than a fighter.
  • Base Metal+Crystal cost from 300+300 => 250+250. That's mkI; other marks use normal progression.
  • Minimum multiplier from 5 => 8.
    • The base-cap-dps was already high at 73k and is now in the top 10 among fleet ships (including self-destructing ships and ships with no bonuses) at 117k, but this helps counterbalance the fact that it doesn't get more than the minimum multiplier against most fleet-ship targets.
  • Multiplier from (target energy use)/1024 => (target energy use)/256. So it starts getting a higher-than-minimum bonus against ships using 2049+ energy, up to the max multiplier of 30 against targets using 7680+ energy (note that the IRE is not any better damage-wise against max-multiplier targets, since it was already very powerful there).

• Space Plane, in recognition of doing really well in the 3rd "worst ship" poll:
  • Vs-Hull multipliers from 2.4 => 3.2.
    • For reference, that's against Light (Fighter), Polycrystal (Bomber), Heavy (... a lot of things), and UltraHeavy.
    • Eyebots have 3.2, and a somewhat lower base-dps, but can shoot through forcefields.
  • Seconds-per-salvo from 3 => 9.
  • Base Attack Power from 600*mk => 1800*mk.
    • So that's the same dps as before (aside from the multiplier increase), but higher alpha capacity for hit-and-run using their radar dampening, etc.
  • Base Energy Use from 100 => 50, because with the 1.75x ship cap the default energy use (100) was kind of excessive.
  • Base Metal+Crystal cost from 90+90 => 75+75 (that's mkI). More in keeping with the dies-due-to-light-breeze-when-it-can-catch-them nature of the ship.
  • And the AI is really happy about this part: the Space Plane is now immune to AoE damage. That's bargained down from immune-to-gravity, so be happy.

• Vampire Claw, also in recognition of runner-up-for-worst-ship:
  • Base Health from 5000*mk => 10000*mk.
    • Previously they only had 4M cap-health, which was below even stuff like eyebots. Kind of rough for a melee ship even with self-regen. Now at 8M they're still really low compared to most fleet ships.
  • Now have cloaking. Be afraid. Bring tachyons. And garlic.
    • Note that this means that games with cloaking ships disabled won't have claws. But any game already started that has them will continue having them, regardless of that setting.
  • Base Energy Use from 200 => 160.
  • Base Attack Power from 650 => 850.
  • Added Artillery to the hull types this has a bonus against (om nom missile frigate tasty).
  • Now has a 0.3x multiplier against the command-grade hull type.

• Youngling Tiger:
  • Base Metal+Crystal from 40+40 => 25+25, same as commandos and weasels.

• Youngling Vulture:
  • Base Metal+Crystal from 40+40 => 25+25, same as commandos and weasels (and tigers, now).

• Spire Gravity Drain:
  • Base Metal+Crystal from 0+3000 => 500+1000. These were tied for most expensive fleet ship, and 2x (or more) as expensive as all but 6 other fleet ship types. Wasn't really anything to justify that.

• Spire Gravity Ripper:
  • Base Metal+Crystal from 1800+1200 => 1000+500. Was similar to the grav drain's situation.

• Spire Armor Rotter:
  • Base Metal+Crystal from 800+1900 => 450+900. Was 3rd most expensive.

• Space Tank:
  • Base Metal+Crystal from 1100+0 => 600+100. Was 5th most expensive, and more expensive than bombers, now slightly less expensive than bombers.
  • Base Attack Power from 2600*mk => 2300*mk.
  • Vs-Hull-Type multipliers from 3.2 => 6.
    • Bombers have 6. Space Tanks now have a roughly 30% higher base-dps, since the general idea is for bonus ship types to be 30-50% more useful than triangle types.
  • Effective range from 4800+200*mk => 6300+200*mk. Since they're so slow.

• Grenade Launcher:
  • Base Energy Use from 400 => 150. It had twice the cap-energy-use of the next highest fleet ship type.

• Sentinel Frigate:
  • Base Attack Power from 14000*mk => 31000*mk. It had a cap-dps (base=max for this one) of 44k, which was really laughable. 100k is bit on the low end for a type with no bonus of any kind, but it is an infinite-range ship.

• Made end-of-Fallen-Spire spawns yet more enthusiastic, due to the opposition they sometimes face on 10/10. That AI sure is stubborn...

• Dyson Antagonizer (hybrid plot) :
  • Now tries to be built somewhat closer to the AI's border.
  • Health from 80M => 40M.
  • The more dangerous values may be restored for Advanced Hybrids or some other option in the future, but the previous string of buffs to this plot was proving too much for normal.

• Knowledge Raiding:
  • Now scales more granularly with difficulty (for example,previously 9.3, 9.6, and 9.8 all had the same intensity of response as 9; no longer). This somewhat increases the difficulty of k-raids on non-integer difficulties.
  • Now considers the ship cap of ships being spawned when deciding how many to spawn, so a laser gatling doesn't "cost" as much as a fighter, but a Maw "costs" quite a lot more. This makes the difficulty of k-raids not nearly as variable based on what bonus ship types the AI has, or the unit-cap-scale being used.
  • Now only the AI controlling the planet spawns response ships, but spawns twice as many, rather than both AIs spawning ships separately. This makes the composition different if the AIs have different bonus types.
  • Now the spawns happen every 10 seconds instead of (11 - Difficulty) seconds, and the strength is multiplied by 2.5 to roughly normalize it to what diff 7 was doing. The base strength of the spawn is still linearly proportional to difficulty, but this second difficulty-based factor which made it heavily higher-than-linear is no longer there. That really steep ramp-up at the end was good before, but would be excessive with the other changes below.
  • Now, the more knowledge you raid for, the more intense the response. Up to 1 planet's worth of knowledge there is no change, after a full 2 planet's worth of knowledge the response is 1.5x as intense (having increased linearly during the raids after the first planet's worth), at 3 planet's two times as intense, at 4 2.5x, and so on.
    • In multiplayer games this ramping-up is adjusted, e.g. if 3 players have raided 9000 total knowledge (3000 each), the ramp-up is the same as in a single-player game where that player has raided 3000 knowledge. Multi-homeworld players only count as 1 for this purpose.
    • Now each spawn makes a number of "wild rolls" which apply various effects to what happens.
      • These range from "nothing" to "make it spawn around the planet edge" to "multiply it by 1.5" to "make it all raid starships" to "make it spawn on a neighboring planet", etc.
      • At first there is only 1 roll per spawn, but this goes up by 1 for every 1500 knowledge raided.
  • If you have Advanced Logging enabled, each spawn will generate an entry in the "CounterSaboteurSpawns.txt" log file; it may not make sense to players but is very helpful to us if you submit it with your feedback.
  • For the detailed discussion see http://www.arcengames.com/forums/index.php/topic,9906.0.html , but in brief the rationale is:
    • K-raiding's main purpose is to let players dig themselves out of a hole by gaining power without increasing AIP.
    • K-raiding also needs to not be a way of circumventing normal AIP-gain on a wide scale. To some extent it's good; avoiding 200 AIP via it is a bit much.
      • If a player gets in a hole multiple times in a row, it's ok if the k-raiding gets harder and eventually either they die to it (sparing them a longer game) or they have to bite more AIP (making progress, probably towards sparing them a longer game).
    • If possible, K-raiding also needs to not be boring. Partly that means making it so that "playing optimally" doesn't involve 10+ k-raids in a game, and partly that means making the actual AI response to the raids more interesting.
    • In general, the balance of all this is tentative. It looks right in our testing, but it may be too easy or (perhaps more likely) too hard. Please give us feedback and we will continue to refine towards better balance and more fun :)

• Scout Starship:
  • Base Health from 120k/480k/960k/1.2M => 300k*mk.
  • Base Armor Rating from 1500 flat => 1500*mk.
  • Tachyon Range from 750/950/1150/1550 => 2000*mk.
  • Ship Cap from 5/4/3/2 => 5/5/5/2.
  • Energy Use from 2000/4000/6000/6000 => 2000/2000/2000/5000.
  • Base Move Speed from 224/464/704/704 => 240/480/720/720.
  • Now can cloak-boost up to 20*mk allied cloaked ships (range: 4k/5k/6k), except the MkIV which isn't being given cloak-boosting due to potential issues with a tachyon-immune ship having a lot of cloak-boosting.
  • The MkIV Scout Starship:
    • Now provides counter-missile coverage.
    • Will now not use the evade-after-exiting-wormhole logic at all, since it is permacloaked (the non-starship MkIV Scout has the same attribute already).

• Spirecraft-Medium and Broken-Golems-Medium no longer have a score penalty.

• Fortresses (mkI, II, and III; human and AI) now have a radar dampening range of 30000. Since most ships have ranges lower than that it doesn't affect them, but this prevents sniping them (notably with zenith-bombards and sentinel frigates) from way out of range.

• Now when a unit dies while under the influence of a parasite infestation ("reclamation damage"), those parasites will behave far more efficiently:
  • If there are enough of them on the ship to take it over (that is, reclamation damage is >= half the ship's max health) then enough will remain to perform the takeover.
  • Any remaining parasites (which means all of them, if there aren't enough to take the dying ship over) will spread to nearby (within 2000 range units) valid targets that have not been fully infested.
  • The end result is that while the exact same amount of reclamation damage is being done, the efficiency of it actually doing something in condensed fighting is way more satisfactory, particularly towards the end as all the partially-reclaimed ships die off and the parasites all pile up on the survivors.
    • In fact, it's possible that reclamators are now back to being OP, but not nearly as much as in the days when a single leech-starship salvo could "tag" 40 ships for guarunteed reclamation. Your feedback on the new balance will be appreciated.

• Youngling Nanoswarm:
  • Base attack power from 400*mk => 800*mk. Still not much at all, but it allows:
  • Inherent "damage does 2x as much reclamation damage" rule removed. So it does the same amount of RD as before, just without the fiddly rule (which is now completely gone, the nanoswarm was the only thing using it).

• Mines:
  • Base Health is now 16 times the damage done, to allow 16 detonations before "wearing out" instead of the 2/3 that were probably common before. There are 16 mines in the graphic: the logic is unassailable, no?
  • Standard Mines:
    • Base Metal+Crystal cost from 360+1200 => 20+40, which puts them at about 2x the cap m+c cost of zenith autobombs (which are in many ways simply mobile mines).
      • Note that the build time for all 3 mine types is still 90 seconds.
  • Area Mines:
    • Base Metal+Crystal cost from 1080+3600 => 40+80.
    • Now hit a maximum of 10 targets with their aoe.
    • Base Damage per hit from 7500 => 52500 (half as much damage as a normal single-target mine).
  • EMP Mines:
    • Base Metal+Crystal cost from 720+2400 => 60+120.
    • Fixed an error in the tooltip: it said the paralysis was 60 seconds long, but it's really only 10 seconds.

• Harvester Exo-Shields, in honor of winning the 4th "Worst Unit" poll (and the first one for which it was eligible) :
  • Now has cloaking.
  • Now cloaks the protected harvester.
  • No longer halves production of protected harvester.
  • Energy Use from 1000 => 5000. At maximum non-ZPG energy efficiency (running everything off mkIIs) that's a bit under 4 (m+c)/s to run, or 1/5th of a mkI harvester.
  • The result is that:
    • They have a much lower operational cost which no longer scales up with higher harvester marks or with your energy efficiency.
    • More importantly, they allow a new tactical choice: do you want attacking enemies to potentially be distracted some/all of the harvesters in the system, or do you want to make sure they stay focused on the command station or some other objective? Some defensive setups get much higher efficiency when the enemy stays focused in a particular radial or linear area, and this can help with that. On the other hand, sometimes the AI's tendency to split up after harvesters is a big help.

• Metal/Crystal Harvester II/III, in honor of getting 2nd place (by one vote) in the first "Worst Unit" poll in which they were eligible:
  • A bit of background:
    • Unlocking Econ Station II costs 4000 knowledge and gives a total of 576 m+c above the "standard" command station (econ I) if you build all 6, and that bonus is still relevant if you unlock Econ III. That's about 0.144 (m+c)/s per knowledge point.
    • Unlocking Econ Station III costs 5000 knowledge and gives a total of 1536 m+c above the "standard" command station. Benefit/K Ratio is about 0.3072.
    • Previously, unlocking mark II of either metal or crystal harvester cost 3250 knowledge and, assuming a mid-game situation of 13 total planets (i.e. all 6 of econ II and econ III could be placed) with an average of 2 spots of that resource per planet (ymmv, but it's probably not far off) gave a total of 208 resources above mkI harvesters. Benefit/K Ratio was about 0.064.
    • Previously, unlocking mark III of either harvester cost 4000 knowledge (but really 7250 since you don't get the benefits of mkII and mkIII at the same time like you do with econ stations) and, across 26 ressource spots, gave a total of 416 resources above mkI harvesters. Benefit/K Ratio (assuming 7250 "real" K cost) was about 0.0574.
  • So, to do something about this pretty vast disparity:
    • Harvester II:
      • Knowledge cost from 3250 => 2000.
      • Production from 28 => 31 (so +8 => +11).
    • Harvester III:
      • Knowledge cost from 4000 => 2500 (so 4500 total).
      • Production from 36 => 73 (so +16 => +53).
        • Yes, that's a lot, but that's what parity with econ III looks like. And nerfing econ III would have been an outrage considering how much waiting-for-resources happens on challenging games _with_ econ IIIs.
  • In short, harvester upgrades should now be competitive with econ station upgrades. Getting mkIII in both harvesters also now costs the same amount of knowledge as getting mkIII econ stations, but has a somewhat lower "barrier to entry" in that the individual knowledge costs are smaller.
    • Further balance feedback is, of course, welcome. This change will have a fairly dramatic impact in favor of harvesters on multi-HW games, for one, but it's primarily important that choices be interesting in single-HW games and adjustments can be made for the less common scenarios (as long as multi-HW doesn't become excruciating, of course). Of more concern is the impact on multiplayer games if all players gift all their harvesters to one player who upgrades to mkIII harvesters (and gives resources back) and other such tomfoolery, but that can also be adjusted for as the need arises.

• Warp Jammer Command Stations, in honor of tying for third in the first "Worst Unit" poll it was eligible for:
  • Now prevent their planets from putting adjacent AI planets on alert.
    • Those adjacent planets can still be put on alert any other way (for example, but another bordering human world without a jammer or by a significant human presence on the planet or a different adjacent planet).
    • But this still allows a new strategic choice in shaping the AI's response to your presence.
    • Notably, this could be used on a planet adjacent to a particularly hard-to-crack AI planet to allow you to get supply on the target planet and thus deploy turrets and such without putting the target planet on permanent alert (which is normally a distinct no-no for core worlds and homeworlds). But beware, if the jammer is destroyed, its effects go away. Also, jammers are far from free.
  • Also fixed a longstanding bug where these counted (for display purposes only, apparently) as an AI reinforcement warp gate.

• Missile shots (from the missile frigate, MLRS fleet ship, MLRS turret, Missile Turret, Anti-Armor Ship, etc) now can acquire a new target from their firing ship's target list if their current target is destroyed or otherwise no longer available (note that all shots insta-hit a ship trying to leave the planet, but missiles now won't "overkill" them in that case).
  • Bear in mind that the parent ship's target list does not necessarily include all viable targets, and is often limited to 50 nearby targets, but this still significantly decreases DPS wastage for missiles (which generally move fairly slowly)

• Mobile Repair Stations, in belated honor of doing well in the last "worst unit" poll
  • Maximum simultaneous repairs from 100 => 20.
  • Repair rate reduced.
  • Time it must remain stationary before starting repairs from 60 seconds => 5.
  • Now has cloaking.
    • This doesn't break when repairing, but beware AI tachyons.

• Metal/Crystal Harvester Upgrades:
  • MkII production from 31 => 30.
  • MkIII production from 73 => 55.
  • Results/Rationale:
    • Assuming 12 resource spots on the homeworld and 4 resource spots per captured planet (the latter is a well-attested average, to the author's surprise).
    • MkII upgrades are better than EconII if you have 1, 2, 3, or greater than 12 planets, but are worse from 4 to 11.
    • MkIII upgrades are better than EconIII (plus EconII since it's unlocked on the way) if you have 1-4 or greater than 13 planets, but are worse from 5 to 12.
    • Harvesters still have the advantage of not occupying the command station "slot" but also still have the disadvantage of being at the mercy of mapgen: sometimes you don't have the luxury of picking a planet with 4+ resource spots instead of a planet with 1-2.
    • All in all, the question of "do I upgrade harvesters or econ stations?" should now be more situational and map-based.

• Fixed a null exception that happened when gifting a spirecraft mining enclosure to another player.

• Neinzul Enclave Starships, in honor of winning the fifth "Worst Unit" poll.
  • Base Armor Rating from 1000 flat => 1000*mk.
  • Base Health from 500K*mk => 3M*mk.
    • This gives an individual one 80% the health of an individual corresponding fleet-boosting starship; not using cap comparisons here since these are not military ships (i.e. ships with an attack), but for reference those fleet-boosting starships have 2x as much cap.
  • Now have radar dampening of 8000/7500/7000/6500.
    • Between this, the armor, and the health it should be a lot more reasonable to keep these alive unless you get overwhelmed, in which case these should die (cloaking was experimented with, but was way too exploitable; that much cheese requires cloaker starships).
  • Knowledge Costs from 0/3000/5000/10000 => 0/2000/2000/14000.
    • This is because the first two upgrades are purely for in-the-field construction (which is worth something) and the last is a game-changing ability to have mkIV production capacity without an advanced factory.
  • The rate at which they construct ships used to be the same as a space dock, and is now 4*mk as fast as a space dock (so 16x for a mkIV).
    • This is because previously using these efficiently basically required engineer support (specifically, engineer mkIII support). Engies still help, but it's not as critical.

• Armored Warheads, in honor of placing second in the fifth "Worst Unit" poll:
  • Base Attack Power from 300k/1.5M/2.4M => 1M*mk.
  • Explosion Size from 1500/1250/750 => 1500*mk.
  • Now do 10k*mk armor damage, enough to strip all the armor off most units that aren't some kind of boss fight.
  • In general you still aren't going to want to use these a lot, it's still a "rot your teeth" sort of unit where heavy use is rarely a good sign for how well your game is going. But these should be more useful than they were, for sure.

• Lightning Warheads:
  • Base Attack Power from 300k/1.5M/2.4M => 800k*mk.
    • Just keeping up with the armored warhead changes; only a significant change to the mkI, really. So the armored ones pack about 25% more punch, for whatever that's worth.
  • Explosion size changed to match the new armored warheads: 1500/1250/750 => 1500*mk.

• Metal/Crystal manufactory input and output scale multiplied by 10 (so consumes 200 of a resource to produce 140, instead of 20 for 14), in honor of placing 3rd in the fifth "Worst Unit" poll.

• There's now a galaxy-wide population cap on each of the three types (enemy-to-all, player-ally, and ai-ally) of Dyson Gatling, equal to the current effective AIP.
  • So if AIP is 80, there could be 80 enemy-to-all gatlings, 80 AI-ally gatlings, and 79 player-ally gatlings and if the Dyson was currently player-ally it would still spawn another player-ally gatling before ceasing to spawn until either the player-ally gatling count went down or the AIP went up

• Amended the "if base max health >= 1,000,000, then immune to fusion cutters" rule to exclude fleet ship types whose mkI versions don't pass that >= 1,000,000 threshold (to prevent situations where fusion cutters were getting way less useful against higher-mark AI planets, etc). The fleet ship types that still have the immunity due to health:
  • Spire Stealth Battleship
  • Spire Blade Spawner
  • Spire Maw
  • Spire Tractor Platform

• Ships killed by a zombie-reclamator no longer spread any excess reclamation damage they have to other ships. The spreading mechanic was intended to make non-zombie-reclamators more reliable; the zombie reclamators were already plenty powerful enough :

• Teleport Battlestations, in honor of doing pretty well in the last worst-unit poll:
  • Base Max Health from 14k*mk => 28k*mk (bringing it from 4.8M cap-hp, which is basically the minimum for a fleet ship, to 9.6M cap-hp where about 18 other types will be lower than it).
  • Base Attack Power from 400 => 700 (bringing it from 39.2k cap-base-dps and 94k cap-max-dps, where some ships have a base dps higher than that max-dps, to the midst of the respectable-attack types)

• Marauders and Resistance Fighters can now be targeted by ships that cannot target small units (Bomber Starship, etc).

• Since the Riot Starship MkIII did pretty well in the last worst-ship poll:
  • Gave them tachyon coverage matching the MkIII scout starship.
  • Added a couple additional options for its highest-level hardpoint:
    • Grav Tazer
      • Similar to the Riot IIs tazer but also applies the "halt" effect (same as the grav ripper) to everything hit, making it take longer for even paralysis-immune ships to get away.
      • Note that these share the same planet-wide stagger category as Riot II tazers, to prevent these making permanent tazer lock possible again.
    • Grav Generator
      • Basically a ship-mounted grav turret, requires research of Grav Turret Is.

• CPAs now don't get queued up if the result based on difficulty, AIP, etc would be way too small (under 200 ships, for high caps, adjusted appropriately) (once a CPA is announced, however, it will happen).

Rebalancing High (and-not-so-high) Difficulties

• Since the many changes to high-difficulty wave calculations a few months ago the 9+ difficulty range has gotten a fair bit of play and generally looks easier than it has been. And just recently the game was beaten fair-and-square on 10/10. In response to that bug:
  • A Diff 9+ player will seed 1 less data center than normal (but no fewer than 1 if it would have seeded any).
  • On Diff 8+ the first step of the wave size formula has been changed from:
    • ( AIProgress * AIDifficulty ) / ( 13 - AIDifficulty )
    • to ( ( ( AIProgress * 0.8 ) ^ 1.1 ) * AIDifficulty ) / ( 13 - AIDifficulty )
    • This produces very similar sizes in the early game but adds a gradual exponential increase as AIP gets higher, having both the effect of making later planet captures not feel as much less significant (in terms of wave size) than early planet captures and increasing the overall "throw weight" of the AI's attacks on these higher difficulties.
  • On Diff 7+ the randomization of wave intervals (which feeds directly into step 5 of the wave size formula) has been changed to consider the number of planets that the AI can send waves against (and thus the probable defensive power of those worlds, i.e. chokepoints, and the viability or lack thereof of smaller waves):
    • If there are 6+ wave-targetable planets it uses the normal (pre-5.036) formula of adding a random number from (AIDifficulty*-60) to (AIDifficulty*120).
    • If there are 5 wave-targetable planets the random is from (AIDifficulty*-30) to (AIDifficulty*120).
    • If there are 4 the random is from 0 to (AIDifficulty*120).
    • If 3, from (AIDifficulty*30) to (AIDifficulty*120).
    • If 2, from (AIDifficulty*60) to (AIDifficulty*120).
    • If 1, from (AIDifficulty*90) to (AIDifficulty*150) (making it actually able to hit single-chokepoints harder than it used to).
    • The upshot is that it makes the AI a little "smarter" about figuring out that it may as well not bother sending small waves against concentrated defenses and should just save up for stronger attack forces. Accordingly, since the AI on <7 is intentionally a bit dumb it doesn't get this new rule at all.

• Previously, an AI of difficulty 9.3 or higher would automatically be one tech level higher than normal. This was generally fine but did cause a pretty massive difficulty cliff between 9.0 and 9.3, and we prefer to keep each step a relatively smooth progression from one step to the next. Now:
  • That automatic tech level increase has been removed.
  • AI's with difficulty > 9 get an additional reduction to their tech thresholds (on top of the normal difficulty*10 one) of ((difficulty-9)*200). So the threshold for reaching tech 2 is:
    • 9 = 210 (same as before).
    • 9.3 = 148.
    • 9.6 = 85.
    • 9.8 = 43.
    • 10 = 0 (so diff 10 starts at tech 2).
    • Of course, that makes the thresholds for tech 3 and tech 4 much higher since it's using the normal base of 800 and 1200 for those instead of 300 and 800 (which they used to use due to automatically being one tech level higher), thus actually making things a lot easier for those higher AIPs. See next change.
  • To add back some of the difficulty but in a more granular fashion than before, on diff 9+ (not just 9.3+) waves will have a portion of their fleet ships "promoted" to the next tech level according to how close AIP is to the next tech level.
    • So if you're playing on 9.6 and have 50 AIP, approximately 50/85= 58% of the mkIs fleet ships in a wave will be promoted to mkII.
    • Of course, they'll also run into the tech level multiplier which will reduce their actual numbers significantly, but roughly 58% of the wave's "strength" of that ship type was spent on the higher tech level.

• Previously carriers were way less dangerous than the ships they contained, tending to make a wave + a stack of carriers feel a lot more like that many back-to-back waves. This is better than the pre-carrier condition of waves capping out at 2000 ships, but still tends to cause the actual difficulty of waves to increase much much slower with the size of the wave than is desirable. So:
  • Now gets +1 shots-per-salvo per 16 ships inside the carrier (that's on high caps; per 8 ships on normal, 4 on low, 2 on ultra low). So a carrier containing 1000 ships on a high caps game has 63 shots per salvo.
  • Base Attack Power from 1400*mk => 19200*mk (the base strength of 16 fighters on high caps, but without the fighter's bonuses).
  • Seconds-Per-Salvo from 2 => 4 (same as fighter, cuts down a bit on shot-spam).
  • Can no longer fire as though ignoring through forcefields, but it can still move through them.
  • Shot type from dark matter => energy bomb as the idea is not for these to be easily counterable via counter-dark-matter turrets.
  • Hull-bonus against Turret from 0.1 => 1.
  • Hull-bonus against Structural from 0.1 => 1.
  • Hull-bonuses against Scout and Command-Grade staying at 0.1.
  • In summary, a full carrier is now pretty dangerous, but the firepower is still significantly lower (by about 3x, probably) than its total contents. That's fine as they'll probably still feel more powerful than you want them to be.

• Previously it was possible for AI Homeworlds to get really brutal combinations like an AI Eye + 3 (or more) Core Raid Engine guard posts (which is pretty much the RNG saying "good game" on higher difficulties but you don't know until you scout it), or nothing really all that threatening at all. Randomization is core to AI War, but this particular bit of it could literally be the difference between a fairly easy endgame and a mathematically-impossible-to-win situation.
  • Now, for AI Homeworlds, it randomizes the three most brutal structures (Core Raid Engine, Core CPA, and AI Eye) separately from the rest of the structures:
    • On Diff 9+ it always gets two picks from this new set. So you could get an Eye + a Core Raid, or two CPAs, or two Eyes (which is actually kind of nice to you since they don't really stack), or whatever. But not 3 Core Raids. And not none of the above.
    • On Diff 8+ it randomly rolls (per homeworld) between one and two picks.
    • On Diff 7+ it always gets one pick.
    • Below Diff 7 it randomly rolls between zero and one picks.
  • As the set of core guard posts expands in the future (expansions, mainly), this "brutal" set can also be added to.

New Drones For The Neinzul Enclave Starship

• Didn't want to let the AI have _all_ the fun, so: Added 4 new drone ship types that only the Neinzul Enclave Starship can build:
  • Neinzul Laser Drone I-IV
    • MkI and II require Laser Turret II technology. MkIII and IV require Laser Turret III technology.
    • Has 5x bonuses against the hull types that Laser Turrets have a bonus against.
  • Neinzul Needler Drone I-IV
    • MkI and II require Basic Turret II technology. MkIII and IV require Basic Turret III technology.
    • Has 5x bonuses against the hull types that Basic Turrets have a bonus against.
  • Neinzul MLRS Drone I-IV
    • MkI and II require MLRS Turret II technology. MkIII and IV require MLRS Turret III technology.
    • Has 5x bonuses against the hull types that MLRS Turrets have a bonus against.
  • Neinzul Missile Drone I-IV
    • MkI and II require Missile Turret II technology. MkIII and IV require Missile Turret III technology.
    • Has 5x bonuses against the hull types that Missile Turrets have a bonus against.
  • A Neinzul Enclave Starship can build any of these up to its own mark level once you have the corresponding turret technology, just select an enclave and click the "DRONE" tab on the buy menu (similar to how the Starship Constructor has a "RIOT" tab for the Riot Control Starships).
  • Stat wise:
    • Each drone only lives 30 seconds; no repair, no attrition-stops-on-low-power, no regeneration chambers, nada. It's got 30 seconds to hurt something.
    • Drones do not have the lightning speed of the normal Youngling types (not wanting these to horn in on what makes the younglings special). One consequence is that the "mothership" has to be within a certain range of the stuff you want hurt.
    • Drones are pretty flimsy, but dying from enemy fire isn't much of an event for these.
    • Ship cap is about 25% of normal because these aren't supposed to be like a whole new ship unlock, they're just a bonus on top of what you already get for turret (and enclave starship) research.
    • Drones cannot traverse wormholes.
    • Drones are super, super cheap in both metal/crystal and energy.
    • Attack-power-wise on an individual basis they're about as strong as a normal fleet ship with 5x bonuses, making them fairly effective against the stuff with those hull types.
  • The motivation behind these:
    • High-difficulty play can make it feel very painful to spend knowledge unlocking something that doesn't give an offensive boost. That's fine, but turrets in particular could get very little love due to their... very limited offensive use, shall we say. So having turret research provide a modest amount of offensive boost (if you have CoN and thus the Neinzul Enclave Starship, anyhow) seemed like a fun way to do things. And we've already done the cross-tech thing with having turret research unlock spire capital ship modules in LotS, so doing a bit more of it seemed fine.
    • Despite the fairly massive buffs Neinzul Enclave Starships received not long ago there's still been a lot of feedback that they're underwhelming and that they don't have a lot of point. Well, now they can build lots of little sharp points that will really annoy the AI for you, and they're the only way to build them.
    • But this is also not so powerful that you have to do turret research or even use the enclaves at all, it's just making it nicer if you do.

Better Scaling For Multi-Homeworld Games

• Previously CPA size was multiplied by the number of (non-Helper-role) human players, fixed to be multiplied by the (starting) number of human homeworlds.

• Previously if there were multiple human homeworlds (and thus human ship cap was much higher) each AI player would send that many extra waves (and various threats like normal CPAs and exo-attacks would be multiplied in size accordingly, though not always linearly).
  • This was generally fine but:
    • Particularly on high homeworld counts that could lead to a ton of waves all smaller than the carrier threshold that would nonetheless add up to larger than the max number of AI ships that the game generally allows on a planet (5000), causing those extra AI ships to just disappear.
    • Also, now that carriers will be more of a threat, it would be nice for them to see use in those circumstances.
    • Some important varieties of AI strength were not being increased at all, like: Reinforcements, Counterattack Guard Post spawns, Raid Engine spawns, Core Raid Engine spawns, Core CPA Guard Post spawns,
  • So now:
    • If there are one or two human homeworlds, the number of waves per AI player is calculated the same way as before: one and two, respectively. But if there's more than that it still doesn't go above two and instead increases the size of the individual waves proportionately.
      • So 3 HWs means twice as many waves as single-player (instead of three times as many) but they're 1.5x as large as they would have been. 4 HWs means twice as many waves that are 2x as large, and so on.
    • Reinforcements, Counterattack Guard Post spawns, Raid Engine spawns, Core Raid Engine spawns, and Core CPA Guard Post spawns now use the same scaling-with-HW-count as is used for normal CPAs: multiply by 1 + ((HW_count-1)*0.33).
      • This isn't 1:1 because the power increase from multiple homeworlds isn't (closer in MP, but the problems of coordination and efficiency are in play there).

• When Spirecraft are enabled and there's more than one human homeworld, the number of asteroids seeded will go up.
  • The main reason is that on Hard you're facing much harder exos on multi-HW, but weren't getting any additional benefit to match. Medium was not as much of a concern, but it could use it too.
  • The increase is 50% for each additional human homeworld, so 2 HHWs => 150% asteroids, 3 HHWs => 200%, 4 => 250%, etc. A straight-up 1:1 was also possible but seemed excessive. In general multi-HW isn't really a 1:1 increase, but let us know if this scaling feels like too much or too little.

• Spirecraft Shieldbearers:
  • Can now be put in low power mode (shutting off their projected shield).
  • Now their projected shield no longer shrinks as the shieldbearer loses health.
    • Something more like shrinking-to-90% had been considered, but that would still have resulted in stacks of them getting whittled down all at once (while still taking up ship cap) rather than dying one at a time while leaving the remaining ones at high health (depending on how they were stacked).
  • Now have AI-only variants that can be chosen for exos.
    • To make them actually work with exos, this variant actually has normal guns. They're only moderately powerful (about 1/4 as much DPS as a same-mark siege tower), but they avoid the logical problems the gunless ones had in exos.
    • These still have the shrink-when-damaged behavior, to make them a bit easier to handle.
    • Adding shieldbearers may make exos significantly harder, but that may not be a bad thing. Let us know if you find it particularly fun or unfun; adjustments can be made.
  • Thanks to TechSY730 and others for inspiring these changes.

• Spire Maw vacuum-recharge-seconds from 2 => 5/4/3/2/1.

• Knowledge Hacker and Ship Design Hacker collision priority set to 890 (below only stationary objects, standard/hardened forcefield generators, command stations, command station cores, and rally posts).

• Core CPA Guard Posts now start a timer-for-CPA instead of instantly triggering a CPA.

• Core CPA Guard Posts now trigger in response to >= 1 enemy human military ship being present, rather than > 1.
  • So that Artillery Golem won't get by on the ol' "But I'm just *one* ship!" excuse anymore.

• • EMP mines:
    • Health halved, which halves the number of times a single one will go off before needing to be rebuilt.
    • Knowledge cost from 1500 => 1000.
  • Area mines:
    • Knowledge cost from 4000 => 2000.

• Vampire claws are now immune to being camouflaged by the Camouflager AI Type because the combination of melee and cloaking didn't work very well there (could wind up with non-decloakable enemy vampire claws just sitting on your planet for a long time).

• Made Maws try to vacuum their shoot-at target if for some reason vacuuming their "assist" target failed.

• Carrier hull type from Scout => Ultra-Heavy.

• Now when a unit is actively making progress on a build queue it loses the ability to be cloaked until 2 seconds after the last build queue activity.
  • Notably, this prevents neinzul enclave starships from spamming a planet with drones (and whatever else) with impunity. Can still FRD built non-drones through a wormhole to a target planet, though, and that's fine.
  • Also this will break the cloak on modular ships that are building new modules, but since modules mostly can't die in combat anymore (except shield modules) and their internal build queue is disabled while they're in low-power mode, and if they're not in low-power mode and they're on an enemy planet they've probably lost cloak due to shooting at something... all told, shouldn't be a big change for them

• Human Forcefield generators (including hardened ones) now drop remains and can be rebuilt.
  • Note that the normal under-construction-forcefield-generator restrictions still apply; namely that it cannot receive construction assistance if it is close to another forcefield that has recently taken damage (this was added a few months ago to prevent forcefield+engie stacks that were effectively immortal).
  • Still, the fact that the rebuild starts at 50% instead of 0% (and is automatic if there are remains rebuilders nearby, as opposed to requiring player intervention) is a buff.
  • Also note that this does not apply to the "player home forcefield generator" that starts next to each human home command station: those are supposed to be irreplaceable.

• Human command stations (other than the home command station) now drop remains and can be rebuilt.
  • Also removed the rule where colony ships and mobile builders on the planet (or in a transport on the planet) when a human command station dies also die.
  • That rule (and the lack of rebuildable command stations) was there to prevent exploits based around very rapidly rebuilding command stations. It could often tie up AI attack forces much longer than would normally be possible, etc.
  • However, one of the most unnecessarily-micro-intensive operations in the game was rebuilding lost planets because you have to build a colony ship, select it, give it a move order to the lost planet, switch to that planet, select the colony ship, select the command station type you want, and place it. Way more player interaction than should be necessary for "rebuild what was there". But we couldn't make it a lot easier without making those exploits possible again.
  • So, to replace the previous exploit-prevention, now when a non-home human command station dies then it "stalls" the construction of any human command station on that planet for 4 minutes (1 minute if the station that died was a military station); during that time a new station can be placed or the previous one's remains rez'd by a remains rebuilder, but the resulting under-construction station will not actually make any progress on construction until the stalling timer has expired.
    • Note that all ways of removing a human station will trigger this timer: being destroyed by enemy fire, being scrapped, and being replaced by another human command station. The latter is something of an inconvenience but the thought is that it is far less than the convenience gained by being able to automatically rebuild command stations.

• Remains Rebuilders no longer require supply, to make rebuilding a "satellite" planet at least possible.
  • If this makes some wild exploits possible please let us know, but since most of the stuff that can be rebuilt requires supply anyway..

• Grenade Launcher grenade explosion radius from 200/300/400/500/600 => flat 500.

• Flak grenade (from guardians and turrets) radius from 300/350/400/450/500 => flat 500

• AI Guardians energy cost from 500*mk => 1250*mk; this is basically irrelevant (AI players don't have an energy model, really), but does help Impulse Reaction Emitters hurt them, giving the IRE another situation in which they are useful that isn't dependent on superweapons or something like that being enabled.

• Shield Bearers (the normal ones; this was already true of the spirecraft ones) are now immune to insta-kill.

• All non-shield modules are now immune to armor boosting (they're invincible, it'd be a waste).

• AI Carriers:
  • Effective range from 6000 => 3000.
  • Base Health from 3M*mk => 2M*mk.
  • Thanks to Wanderer for inspiring these changes.

• MkII+ Basic, Laser, MLRS, and Heavy Beam Cannon turrets now have radar dampening equal to their range minus 1000 to prevent them being inevitably alpha-striked by really substantial waves.

• The Zenith Trader has finally yielded to the threat of sanctions and now only sells no-AIP-on-death versions of structures to the AI. The toys will still annoy you when the AI gets them (indeed, a new toy has been added to the AI-only list to maintain annoyance equilibrium) but the Trader will no longer be an indirect source of AIP.

• In honor of it winning the latest worst-unit poll, Zenith Reserves:
  • AIP cost on death from 5*mk => 1*mk.
  • Amount of ships granted increased roughly 33%.

• In honor of coming in second in the most recent worst-unit poll, the Decloaker:
  • In general, moving this from a mobile tachyon unit with outdated stats to a mobile tachyon unit with roughly mkI-starship stats.
    • Note the tachyon range of 20k was already huge, no buff seemed necessary there.
    • Also note that it isn't really a starship from the perspective of starship disassemblers, etc, it's just beefed up because it has a ship cap of 4.
  • No longer has the IsBlind flag.
  • Now has cloaking.
  • Base Damage-per-shot from 36k => 50k.
  • Seconds-per-salvo from 10 => 2.
  • Effective Attack Range from 18k => 8k (the only reason they had such a long range was that we hadn't revisited this one since the last major range changes).
  • Base Health from 40k => 2M (again, in the general range of mkI starships).
  • Base Energy Use from 100 => 1000.
  • Base Metal Cost from 5000 => 10000.
  • Base Crystal Cost from 7500 => 20000.
  • Now has most common-to-starship immunities.

• Captive Human Settlement, in honor of coming in third in the "worst unit" poll: AIP-on-death from 100 => 15.
  • As was understood in the discussion in the nomination/poll threads, these are supposed to be a "penalty" unit, but players had a very good point that 100 AIP was just gameshattering for something like that (rebel colonies have the same AIP-on-death, for instance).

• Spire Armor Rotter, in honor of coming in fourth in the latest "worst unit" poll:
  • The general idea is to make these pretty good combat ships in their own right, at least until armor matters more in the general case and armor rot thus means more.
  • Base Health from 14,500*mk => 23k*mk (this gives them a similar cap-health as fighters).
  • Multiplier vs UltraHeavy from 2 => 3.
  • Multiplier vs Heavy from 2 => 3.
  • Multiplier vs Polycrystal from 1 => 3.

• Microparasite, because it (like other experimentals we need to get back to) had really outdated stats:
  • In general, rebalancing these to be something like the MkV Parasite while maintaining their distinguishing characteristics (like having twice the cap and firing 4 times as fast).
  • Metal Cost from 1800 => 1400.
  • Crystal Cost from 240 => 3400.
  • Base Health from 7200 => 36k.
  • Base Armor Rating from 500 => 1500.
  • Base Damage-per-shot from 8000 => 2000.
    • Previously they had 6.4x the dps of a mkV parasite, even though they were considered mkIII! This is still 1.6x the dps of a mkV parasite.
  • Base Armor Piercing from 0 => 3000.
  • Effective Attack Range from 4600 => 6000.
  • Multiplier vs Polycrystal from 20 => 4.
  • Multiplier vs Artillery from 1 => 4.

• Special Forces Guard Posts no longer count as reinforcement-warp-gates and are no longer autotargeted (by player or player-ally-minor-faction ships).
  • They still get reinforcements when their planet is reinforced, and those reinforcements use the special forces behavior, but they can no longer make a planet eligible for reinforcement all by themselves.
  • This removes the "cheese" that was possible by keeping them alive to trick the AI into reinforcing a planet with nothing but a special forces guard post on it (generally a waste).
  • With the cheese gone, there's no longer a need to make it hard to keep these alive, so the autotargeting could go away.
  • Which in turn removes the annoyance of getting +1 AIP from actions over which you did not have direct control (particularly with minor faction allies).
  • We may do something more clever with the special forces ships themselves in the future, we'll see. In the meantime these two changes seem to be a net improvement.

Core Guard Post Rebalance

• All Core Guard Posts:
  • Base Energy Cost from 500 => 10000 (doesn't matter except for IRE fire).

• Core Heavy Beam Guard Post:
  • Total base attack power (this is divided amongst all its beams) from 8M => 4M.
  • Multiplier vs Swarmer from 4 => 3.
  • Multiplier vs Refractive from 3.5 => 3.
  • Multiplier vs UltraLight from 3.25 => 3.
  • Multiplier vs Medium from 2.5 => 3.
  • Multiplier vs Heavy from 1.75 => 3.

• Core Electric Guard Post:
  • Basically balancing this against a cap of theoretical-mark-V electric shuttles, but still only about half that power with less than a quarter of the health.
  • Can now hit a maximum of 200 units per shot
    • This is the same mechanic as the electric shuttle and it can multi-hit targets up to 5 times each if hits are left over from the first pass. So it gets more-or-less optimal DPS as long as 40 ships are in range.
  • Base Health from 6M => 12M.
  • Multiplier vs Artillery from 4 => 1.
  • Multiplier vs Neutron from 2 => 1.
  • Seconds Per Salvo from 13 => 20.
  • Damage Per Shot from 3k => 50k.
  • Base Attack Range from 13k => 8k.

• Core Missile Guard Post:
  • Basically balancing this against a cap of MkV missile frigates, but with way less durability.
  • Base Health from 6M => 18M.
  • Multiplier vs UltraLight from 8 => 4.
  • Multiplier vs Medium from 6 => 1.
  • Multiplier vs Light from 1 => 4.
  • Multiplier vs Swarmer from 1 => 4.
  • Multiplier vs Neutron from 1 => 4.
  • Multiplier vs Composite from 1 => 4.
  • Multiplier vs Refractive from 1 => 4.
  • Effective Attack Range from 36000 => 12000.
  • Seconds Per Salvo from 5 => 1.
  • Damage Per Shot from 80k => 200k.

• Core Sentinel Guard Post:
  • Basically balancing this against a cap of MkV Zenith Beam Frigates, but with maybe half the power and way less durability.
  • Can now hit a maximum of 20 targets per shot.
    • Using the same mechanic as the Zenith Beam Frigate: the line attack does full damage to each target hit, but gets no compensation for not hitting the full number of targets.
  • Base Health from 3M => 18M.
  • Damage Per Shot from 4k => 100k.
  • Effective Attack Range from 9000 => 8000.
  • Multiplier vs Light from 16 => 1.
  • Multiplier vs Swarmer from 8 => 1.
  • Multiplier vs Neutron from 4 => 1.

• Core Leech Guard Post:
  • Basically balancing this against a cap of MkV Parasites, but with way less durability.
    • Interestingly, the attack power was basically already spot-on.
  • Base Health from 6M => 12M.
  • Multiplier vs Medium from 3 => 4.
  • Multiplier vs Neutron from 1 => 4.
  • Effective Attack Range from 5000 => 6000 (slightly further out than its radar dampening range).

• Core Zenith Bombard Guard Post:
  • Basically balancing this against a cap of MkV Zenith Bombards, but with way less durability and about half the attack power.
  • Base Health from 6M => 12M.
  • Damage Per Shot from 60k => 900k.
  • Effective Attack Range from 53k => 34k.
  • Multiplier vs Heavy from 1.75 => 2.
  • Multiplier vs UltraHeavy from 1 => 2.
  • Multiplier vs Structural from 1 => 2.
  • Multiplier vs Swarmer from 4 => 1.
  • Multiplier vs Refractive from 3.5 => 1.
  • Multiplier vs UltraLight from 3.25 => 1.
  • Multiplier vs Medium from 2.5 => 1.

• Core Zenith Fortress Guard Post:
  • Was going to balance this against a cap of MkV Fighters (but without the hull-type-bonuses), but with way less durability and maybe 1/2 to 2/3rds the attack power... but it's actually already there.
  • Shot type from (default? it doesn't appear to be set?) => Flamewave.

• Core Spire Shield Guard Post:
  • Base Health from 392M => 200M.

• Core Neinzul Melee Guard Post:
  • Basically balancing this against a theoretical cap of MkV cutlasses, but with way less durability.
  • Base Health from 6M => 36M (it's a self-damaging melee unit, it needs it).
  • Base Move Speed from 30 => 44.
  • Damage Per Hit from 106k => 300k.
  • Multiplier vs Neutron from 3 => 4.
  • Multiplier vs Light from 3 => 4.
  • Multiplier vs Scout from 2 => 1.
  • Multiplier vs Composite from 1 => 4.

• Core Booster Guard Post:
  • Base Health from 6M => 18M.
  • Munitions Boost and Armor Boost range from 8k => 16k.
  • Intentionally leaving the attack alone here: that's not really the point of the unit.

• Buffed reinforcements a bit in general. Judging by CPA size and other indicators since the major rework of reinforcements this has been lower than it was by a significant margin.

• When an AI gets 2 "brutal" picks for its homeworld (only happens on high difficulty), it now cannot pick 2 AI Eyes, since that's too nice.

• Previously it was very easy to keep an AI from ever reinforcing its homeworld until the very end during the assault upon it. This is generally desirable as part of the game is not stirring up that hornets nest until you've got the extermination squad in position. But the complete lack of reinforcements there may be letting that end-challenge get a little too tame against careful players. So now there's a 0.25%/0.50%/1% chance on Diffs 7+/8+/9+ (respectively) of a homeworld receiving high reinforcement priority per reinforcement pulse.

• AIP floor minimum is now 10 instead of 1.

• Data Center AIP-reduction-on-death is now reduced by 25% on Diff 8+, and by 50% on Diff 9+ (so 15 and 10, respectively). If the two AIs are different difficulties, the higher is used in this case (tried using the difficulty of the AI player that owns the data center but that leads to some display inaccuracy problems as the game is pretty used to units not having stats that vary between players).
  • Along with this, Took out the change that reduced the number of Data Centers seeded on Diff 9+ (as pointed out, this did reduce total AIP reduction available but also reduced the challenge of getting it).

• Net AIP reduction for killing all the coprocessors is now reduced by 25% on Diff 8+ and by 50% on Diff 9+ (so 45 and 30, respectively; that's 105 and 90 total reduction but killing all the coprocessors costs +60 AIP). If the two AIs are different difficulties, the higher is used in this case.

• AIP floor no longer has a cap of 300 (not that this should impact games where you were caring about AIP anyway, that requires really high total AIP to reach).

• AIP floor was previously total-AIP-gained / 5. Under Diff 8 that's still true. Now on Diff 8+ it is TotalAIP / 4, and on Diff 9+ it is TotalAIP / 3. If the two AIs are different difficulties, the higher is used in this case.
  • This effectively reduces the return on superterminal hacking since it does +1 and -2 AIP per trigger.

• Zenith Electric Bomber is now immune to insta kill, like other low-cap fleet ships.

• Spire Archives, in honor of winning our seventh "Worst Unit" poll:
  • Now try to seed on a non-home, non-core planet adjacent to an AI core planet. So taking one doesn't mean taking or alerting a homeworld, and doesn't mean taking a core world, but does mean alerting a core world (unless you claim the archive planet with a warp jammer, at least).
  • Knowledge gain rate from 1/sec => 5/sec.
    • So to get the full 9000 knowledge takes 1800 seconds = 30 minutes; can cut that down a bit by having science ships help on the first 3000 (theoretical minimum total time of 20 minutes because the science ships can't help you on the last 6000 knowledge).
  • When their planet is exhausted of all 9000 knowledge, they automatically die without causing AIP.
  • In short: instead of being literally "during the homeworld assualts" plays for knowlegde these are a mid/late-game high-risk-for-knowledge plays, but if you pull them off (holding the planet for 20–30 minutes to get all the knowledge) you get 3 planets worth of knowledge for only one planet's worth of AIP and without a permanent possible +80 AIP hanging over your head if something slips through. If you were going to take that planet anyway, all the better.

• Warheads killed by warhead interceptors no longer cause AIP-on-death.

• Spire Shield Guard Posts (the normal ones, not the Core ones), in honor of tying for second in our first "Aim The Nerfbat (AI-Side)" poll, have had their health changed from 56M*mk => 28M*mk.
  • Galactic Sandpaper Incorporated has already filed for bankruptcy.

• Spire Stealth Battleship, in honor of tying for second in our first "Aim The Nerfbat (AI-Side)" poll:
  • Radar Dampening from 8000 => 10000.
  • Base Move Speed from 18 => 16. (for reference, a fighter's is 28, this is before a variety of other calculation steps)

• The Player Home Forcefield Generator:
  • Now projects a grav well of 6600 radius (3 times its forcefield radius, 1600 more than the MkI Grav turret's radius) that allows a max speed of 8 (same as the MkI Grav turret).
  • Now projects tachyon coverage of 6600 radius (3 times its forcefield radius, 600 more than a MkIII Scout Starship).

• Since the Botnet Golem won (by a landslide) our first "aim the nerf bat" poll, a plan was set in motion. The Botnet, however, saw it coming. Pulling a koolaid-man, it busted out of its former niche into a whole new minor faction, to wit: Botnet Golem.
  • This is basically like the Broken Golems faction except:
    • It only seeds the Broken Botnet, and it only seeds one of them (Broken Golems no longer seeds any botnets, needless to say).
      • It tries to seed on a planet 4 or 5 hops from the human homeworld(s); if it can't it tries 3, 2, and then 1. This is a far more controlled distance than normal golems (which is just 3+ from human homeworlds) because it's best that the challenge of having this faction on not vary so widely as a simple 3+ distance seeding would.
    • The botnet's energy cost on Moderate went from 400k => 800k.
    • The broken-botnet's AIP-on-metamorphsis on Moderate went from 40 => 100.
    • The exo response on Hard is the same as Broken Golems, but it's separate from that so you're really getting all that pain just for the one unit.
  • Also increased the Botnet's health from 36M to 100M (still 1/5th of an Armored Golem) to make it less likely you'll lose the sole advantage granted by this faction to a silly mistake.
    • Note: the golemite AI's version of the Botnet's health is still what it was.
  • We're still very open to further balance changes; the main thing this move enabled was balancing the botnet separately from the other golems: we could have brought it down to the level of the other ones, but where's the fun in that?
    • Note: none of this affects the old saves except the change in base energy from 40k to 80k (and potentially the Moderate numbers for the Botnet will automatically go up, not sure).

• The exos from BrokenGolemsHard and SpirecraftHard are now more aggressive:
  • The threshold for the first exo of each has been doubled, which means (all else being equal) it would take twice as long to happen and be twice as powerful.
  • The accumulation (which is AIP-based, unlike Fallen Spire exos) now acts as if effective AIP is always at least 50 + (game_second/360). In other words, the minimum starts at 50 and goes up by 10 each hour (1 every six minutes).
    • This minimum stops going up at 250 (20 hours). (thanks to TechSY730 for suggesting that)
  • Thanks to Faulty Logic, Kahuna, rabican, and others for playtesting examples where even on high difficulties these exos were fairly underwhelming, at least early on.

• Golem construction costs adjusted due to the recent fix in repair cost computation (bear in mind that golems are never built, only repaired, so these numbers only impact the cost of repairing one, which jumped about 6 fold due to that fix) :
  • Armored golem from 30M+30M m+c => 10M+10M.
    • Combined with the halving of base repair costs in this release, this means that the armored golem now costs the same to repair as it used to take with engie 1s since those used to have a 6x repair multiplier. The difference is now it doesn't matter if you use MRS's or EngieIIIs except in terms of how fast it happens.
  • Artilery golem from 10M+10M => 8M+8M (lots of feedback that these are OP lately).
  • Black Widow golem from 25M+25M => 10M+10M (ditto).
  • Regenerator golem from 60M+60M => 20M+20M (note that the regenerator's post-reactivation repair cost is 10% of what it would normally be, due to an earlier change).
  • Cursed golem from 10M+10M => 3M+3M (a slight buff, but these aren't all that popular in general).
  • Hive golem from 20M+20M => 7M+7M (a slight nerf, but these are quite strong).
  • Botnet golem from 40M+40M => 20M+20M (in keeping with last version's changes, trying to make the cost of these more closely equal the benefit; let us know if this is still off)

• MkI Engineer repair multiplier from 6 => 2; this isn't because it needs a nerf but because with the repair fix people need time to adjust to how much stuff costs to repair when the math isn't bugged and it's easier to do that when the repair happens over more than a couple seconds. If this is a problem an alternate approach can be found, but our understanding is that repair _time_ (as opposed to repair cost) hasn't really been a problem for folks, but vanishing economies have been.
  • MkII Engineer still has a multiplier of 9 and MkIII still has a multiplier of 12, the idea being that if you're unlocking those you specifically want them to do things quickly

• Repair cost in general is now based off half the construction cost instead of the full construction cost.

• Repairing engines now does not cost m+c (if health is also being repaired at the same time that costs m+c).

• Spire Refugee Ship (the starship recovered in the second stage of Fallen Spire) can no longer be put in a transport, since the chase logic doesn't work right with stuff in transports and jumpships make it pretty trivial.

• The game now tries to seed human home command stations further away from wormholes and tries harder to find such a spot. This should help prevent "AI says: I have a plasma siege starship, gg" situations.

• Heroic AI type balance changes:
  • Is now considered a technologist type for the purposes of lobby logic and display (specifically, being displayed in RED), though it does not get higher tech levels than usual.
  • Its non-wave champion spawns now start at Frigate-level at tech-level 1, not Destroyer-level.

• Neinzul champion hulls are now immune to black hole machines.

• Plasma Siege Starship ship cap from 5 => 4. Most of the others here were at 4 already, and it's been pretty widely considered the best of the combat starships lately.

• Leech Starship:
  • Ship cap from 3 => 4.
  • Seconds Per Salvo from 4 => 2.
  • Shots Per Salvo from 3= > 6.
  • Base Attack Power from 30k => 2k (24k base-cap-dps, parasite has 12.25k)
    • Previously Leech Starships had nearly _five times_ the base-cap-dps of the equivalent fleet ship type (the Parasite). That parasite has 4x multipliers where the leech has none, so they don't need to be all that close, but the leech having a higher base dps than the parasite's bonus dps... no.
      • This change was done rather than buffing the parasites because the parasites seem to be reclaiming at a good rate already; further feedback on this is quite welcome.
  • Base Armor Piercing from 0 => 750*mk (same as parasite).

• Knowledge costs:
  • Flagship was 0/2000/5000.
  • Zenith Starship was 1000/2500/6000.
  • Starship Starship was 1000/2500/6000.
  • Bomber Starship was 0/5000/7000.
  • Plasma Siege Starship was 0/5000/7000.
  • Leech Starship was 0/5000/6000.
  • MkI costs (or lack thereof) unchanged.
  • MkII costs are now all 2500.
  • MkIII costs are now all 4000.
  • The idea being that you're paying for 1x/2x/3x the raw stats, with a 0/500/1000 "surcharge" for the fact that 2x health and 2x attack power is more than 2x as useful, etc.

• Health:
  • "Normal" cap-hp for a fleet ship type is about 15M (ranging from 10M to 30M), Normal for a starship type is about 22.5M (but normal cap-dps is lower).
  • Flagship from 3.75M*mk => 5M*mk (20M cap-hp).
  • Zenith from 4.5M*mk => 6M*mk (24M cap-hp).
  • Spire from 3M*mk => 4M*mk (16M cap-hp).
  • Bomber from 4.7M*mk => 7M*mk (28M cap-hp)
  • Plasma Siege unchanged at 5M*mk.
  • Leech from 3.2M*mk => 4M*mk (16M cap-hp).

• Metal+Crystal/Energy costs:
  • Cap-m+c for fleet ship types varies widely, from about 40k for fighters to about 160k for bombers even just within the triangle. Starships are supposed to be substantially more expensive than fleet ships in exchange for being easier to keep alive. Also, the econ part of the game has been a lot easier since changes to the energy system and harvesters. So targeting about 400k m+c for a mkI cap here.
  • Cap-e is more consistent for fleet ships, typically about 20k (halved for mkI types). So targeting about 40k for a mkI cap here.
  • The m+c costs listed are for mkI, mkII are 2x, mkIII are 4x. E costs do not change with mark.
  • Flagship from 40k+24k / 4k e => 60k+40k / 10k.
  • Zenith from 24k+40k / 4k e => 40k+60k / 10k.
  • Spire from 30k+34k / 4k e => 45k+55k / 10k.
  • Bomber from 80k+8k / 2k e => 85k+15k / 10k.
  • Plasma Siege from 8k+80k / 2k e => 15k+85k / 10k.
  • Leech from 60k+40k / 5k e => 60k+40k / 10k.

• Fighter (including tachyon-micro and bulletproof variants):
  • Base Move Speed from 28 => 32.
  • Multiplier vs Polycrystal from 5 => 6.
  • Multiplier vs CloseCombat from 2.4 => 6.
  • Multiplier vs Medium from 2.4 => 6.
  • Base Attack Power from 1200 => 1000 (similar for the variants).

• Bomber Base Move Speed from 28 => 26.

• The following guardians now have the medium hull type:
  • EMP (was Neutron)
  • Heavy Beam (was Heavy)
  • Laser (was Swarmer)
  • Lightning (was Composite)
  • Special Forces Rally (was Heavy)
  • Spire Implosion (was Artillery)
  • Starship Disassembler (was UltraHeavy)
  • Tachyon (was Refractive)
  • Tractor (was Neutron)
  • Vampire (was Artillery)

• Spire Starship:
  • Hull type from Neutron => Medium.
  • Base Attack Power from 120k*mk => 600k*mk.
    • The previous number was the result of multiple math errors during the balancing process. Specifically: its dps was computed using a seconds-per-salvo of 8 when it's actually 10 due to the photon lance's firing time, and its base dps was balanced as if it had 4x bonuses despite not having any bonuses.

• All fleet ship caps have been adjusted to be multiples of 8 (generally by rounding down, unless something was 1 away from the next higher multiple, in some cases), except for those fleet ships which don't change caps at different unit cap scales.
  • This involved about 5 metric tons of relatively minor changes to health, attack power, metal cost, crystal cost and energy cost. Mostly cap-health and cap-dps didn't change by more than 1% in either direction; cap-m,c,e costs vary a bit more but nothing earthshaking. The numbers are internally arranged in a much better fashion now which should help with other changes going forward.
  • The main immediate benefit of this is that the ultra-low-caps setting no longer impacts the balance of fleet ship types whose caps are not multiples of 8, because, well, they are now.

• Previously on Diff 9+ if you were part of the way from one AI-tech-level to the next, each wave would have a proportionate percentage of its ships "promoted" to the next tech level to make the transition in difficulty feel smoother. Now this happens on difficulty 5+
  • Mathematically speaking this makes the game somewhat more difficult and that's why it was restricted to 9+. In practice it seems to make the game easier because you're testing your defenses against waves more like the ones you're going to get at the next tech level, rather than getting the jump all at once.
  • The main motivation for changing this now is that since the double-application of the mark-level downwards-multiplier on wave sizes has been changed to only applying it once (as mentioned above), the higher-mark waves will be larger and the feeling of a "challenge cliff" between AI tech levels greater. This really helps avoid those cliffs.

• Zenith Siege Engine:
  • MkI-cap-health from 16.8M => 10M.
  • Corrected an... interesting balance-math error where it had 720k mkI-cap-dps AND 10x multipliers (against structural, heavy, ultraheavy, and turret):
    • Multipliers from 10x => 5x.
    • MkI-cap-dps from 720k => 360k (which is still really, really high for a base-dps even with no multipliers, but the must-reload-after-move mechanic is still really new, etc).
  • Reload time from 15 => 10.
    • Damage adjusted to maintain dps.
  • Now pay no attention to auto-FRD or auto-kite.
  • Thanks to Faulty Logic for inspiring these changes, as little consolation as that may be to the many ships (including spire capital ships)
• All modular forts now pay no attention to auto-kite (previously some did, and some didn't).

• Zenith Reprocessor:
  • MkI-cap-health from 8.4M => 15M (now in the neighborhood of fighters).
  • MkI-cap-dps from 17.2k => 30k (has 8x multipliers).

• The Shards in the Fallen Spire campaign now move 4x as fast (still immune to all forms of speed-boosting, combat style, etc) and the response "chase spawns" now happen about 4x as often.

• AI Type reinforcement/wave multipliers, since some of these are really old and the reinforcement ones in particular can have a disproportionately large impact on the feel of the game (to the extent that picking a particular AI type can make reinforcements look "broken", as in barely happening) :
  • Mad Bomber reinforcement multiplier from "you get one ship" on non-core-worlds (and 0.25 on core worlds) => 0.7, on all worlds.
  • Neinzul Youngster reinforcement multiplier from 0.25 => 0.8.
  • Vicious Raider, Extreme Raider, and Technologist Raider multiplier from 0.3 => 0.8.
  • Attritioner, Zenith Descendant, Spireling, Starfleet Commander, Experimentalist, and Grav Driller reinforcement multiplier from 0.4 => 0.9.
  • The Tank reinforcement multiplier from 0.3 => 0.9, and wave multiplier from 1.0 => 1.25 (same as the attritioner, zd, etc group above).
  • Feeding Parasite, Thief, Technologist Parasite, Bully, Assassin, and Speed Racer reinforcement multiplier from 1.0 => 0.9 (since they already had a 1.25 wave multiplier).
  • The Core reinforcement multiplier from 0.3 => 0.5 (same as its wave multiplier).

• Corrected a balance-math problem where the game was counting the 14 "secondary hits" of siege plasma without considering the fact that they only hit for 1/16th of normal strength. This means that the zenith siege engine's mkI-cap-base-dps is 45k instead of 360k.
  • The actual attack power of the unit is unchanged, this is just fixing a bug in the internal balancing model.
  • But since it's now clear that the numbers are more reasonable than previously thought, the siege engine's multipliers have been increased from 5 back up to 10.

• Human Cryo Pods and Home Settlements (the buildings that start next to a human home command station) are now immune to aoe (but not to beam weapons, i.e. not immune to linear-aoe), to avoid excessive-frustration issues with units like the firefly.

• Spire Corvette shield modules now cost 2000*mk each of metal, instead of 1m+1c to avoid the ship being effectively invincible due to immediately rebuilding its own shield. As it is a mkI shield can be "brought back online" in less than 20 seconds, but as normal engineers cannot assist if the ship has taken damage in the past few seconds.

• Zenith Reprocessor:
  • MkI-base-cap-dps from 30k => 45k (for reference, Acid Sprayers have 58.8k, and the same bonuses).
  • Now immune to tractors.
  • Now have cloaking (and thus will not be included in future games with cloaking turned off, but will still be present in games already created with cloaking turned off).

• Military Command Stations:
  • Base Health from 500k/1.5M/3M => 1M/4M/9M. (puts it on the low end of a fleet ship or starship with a cap of 10, but with the mkII/mkIII versions much stronger to help justify knowledge cost)
  • Metal,Crystal production from 16,16/32,32/64,64 => 24,24/48,48/96,96 (the same as logistics stations).
  • Now do not suffer damage reduction when firing from under a normal human-tech forcefield.
    • Yes, it's now just an obvious move to put a shield on these; presumably that was already true if you cared about keeping the planet. And it still costs shield cap, so not 100% obvious.
  • Base Damage per shot from 3200*mk => 5k*mk. (puts it on par with a fleet ship or starship with a cap of 10, but since shots-per-salvo is also mark-based this makes mkII/mkIII versions much stronger to help justify knowledge cost)
  • Is now immune to radar dampening.
  • Shot type changed from translocating lightning to a new "knockback railgun" mechanic. So the shots are now insta-hit, and actually still do the translocation code path but instead of sending the target to a random angle and random distance from the planet center they send the target directly away from the military station out to a distance about 2000 units short of the station's maximum firing range (assuming it wasn't further out than that already).
  • MkII and MkIII versions now do 200 engine damage per shot.
  • MkIII versions now apply 5 seconds of paralysis per shot.

• In honor of basically winning the "what most needs attention for 6.0?" poll, the remains rebuilders have received a logic overhaul and now:
  • Try not to target the same thing as another remains rebuilder. If there are fewer rebuild targets than rebuilders on the planet they'll generally spread out pretty evenly across the targets.
  • If in FRD, they now are much less likely to "yoyo" back to the FRD-point between rebuilds until all the targets are rebuilt.
  • Recharge time from 10 seconds => 1 second. The recharge time wasn't particularly easy to notice before because rebuilders would all pile up on the same target and while the ones not actually doing the rebuild would not incur the 10 second recharge time they would have to wait about 1 second to retarget. With a pile of 20 rebuilders it was hard to notice the one that did the initial rebuild took 10 seconds before it could actually rebuild something.
    • This is actually a fairly significant buff to the unit, but it's probably not a big balance impact to start the rebuilding faster.

• Spire Tractor Platform:
  • MkI-Cap-Health from 10.5M => 8M.
  • Base move speed from 18 => 15.

• The AI Core Grav Reactor Post now no longer receives protection from the Core Shield Generator network, to avoid generating maps (mainly snake ones) that were literally impossible due the homeworld having an invincible black-hole-generator-effect unit.

• To provide a counterbalance against the player's ability to build up a powerful champion (and nebula-ally fleet) without increasing AIP much (if at all, depending on map layout and tolerance for deepstrikes), each AI homeworld now gets "nemesis" champions in proportion to the number of human champions, the highest champion hull size unlocked for human players, and that AI's difficulty.
  • Critically, the "population cap" of these nemesis spawns actually goes _down_ as AIP increases (bottoming out at 150 AIP) to simulate the AI having less to invest into its nemesis spawns as AIP increases and waves, special forces, strategic reserve, and other responses are being given more and more resources. More importantly, this allows the nemesis spawns to counter an "early AIP" champion rush without making a more "normal AIP" approach massively harder to the point of stalemate.
  • Nemesis spawns are not immediately replaced up to the population cap (takes a little over 10 minutes to go from zero to the cap), and spawns remain even if the cap drops below the current population (but once you kill them off the extras don't come back).
  • Nemesis spawns never go through wormholes, so they won't ever leave the homeworld they're defending. This is good and bad for the player, but good all around for their intended purpose.

• The tech level of the AI is now only inflated by 1 artificially when playing on greater than difficulty 9, rather than greater than difficulty 8.

• Handicap now has the following effects, all of which is new unless otherwise noted:
  • AI
    • Previously it affected the speed at which the AI reinforced. It no longer does.
    • Increases the number of ships in each wave and reinforcement, and decreases when negative.
    • Roaming Enclaves and Preservation Wardens are spawned sooner or later into the game.
    • Scrap waves are scaled accordingly.
    • Specifically "forced waves" are scaled accordingly.
    • Cross planet atacks are scaled accordingly.
    • Saboteur and deep strike reactions are scaled accordingly.
    • For now, event attacks are NOT affected by this.
    • Initial seeding of AI planets are intentionally not affected by this—same as human players get no starting benefit from handicaps.
  • Player
    • Increases the amount of resources gathered from producers (using a new formula), and decreases when negative (also using the new formula).

• Some triangle rebalancing:
  • The rationale here is that bombers have been having their way with forcefields a bit too much, and having fighters be so much more "general-dps" than the other two has made them much less a natural predator of the Bomber. Also, the Missile Frigate is still being reported as the least desirable by a significant margin.
  • Fighters (including the tachyon and bulletproof variants) :
    • Bonus vs Polycrystal from 2.4 => 5.
  • Bombers:
    • Bonus vs UltraHeavy from 10 => 6.
    • Bonus vs Structural from 10 => 6.
    • Bonus vs Heavy from 10 => 6.
    • Bonus vs Artillery from 10 => 6.
    • Base Attack Power from 1900*mk => 2400*mk.
  • Missile Frigates:
    • Bonus vs Light from 10 => 6.
    • Bonus vs UltraLight from 10 => 6.
    • Bonus vs Swarmer from 10 => 6.
    • Bonus vs Neutron from 10 => 6.
    • Bonus vs Composite from 10 => 6.
    • Bonus vs Refractive from 10 => 6.
    • Base Attack Power from 1600*mk => 2400*mk.
    • Base Crystal Cost from 700 => 500.

• Tractor Beam Turrets:
  • Base Health from 210k/840k/1680k => 560k*mk.
  • Base Armor Rating from 1200 (flat) => 450*mk.

• Beam Guardians:
  • Bonus Vs Turret from 8 => 1.
  • Bonus Vs UltraLight from 8 => 1.
  • Bonus Vs Artillery from 2 => 1.
  • Base Attack Power from 4000*mk => 4500*mk.
  • Base Armor Rating from 600*mk => 300*mk.

• Laser Guardians:
  • Base Armor Rating from 1000*mk => 300*mk.
  • Base Health from 1.4*(standard guardian health) to 1.1*(standard).
  • Shots Per Salvo from 3 => 15.
  • Seconds Per Salvo from 4 => 2.
  • Base Attack Power from 17k*mk => 1700*mk.
  • Base Armor Piercing from 500*mk => 300*mk.
  • Base Attack Range from 7000 => 5000.

• Raider Guardians:
  • Now Have Radar Dampening Range of 8000.

• Lightning Guardians:
  • Base Armor Rating from 800*mk => 600*mk.
  • Now use the same "can hit up to 200 targets max, but can do max damage to as few as 40 by hiting each target up to 5 times" logic as Electric Shuttles.
  • Base Attack Power from 1600*mk => 2400*mk (for reference, Electric shuttles on low caps have 1600*mk, albeit somewhat slower reloading).

• Tractor Guardians:
  • Base Armor Rating from 1200 (flat) => 450*mk.

• Spider Guardians:
  • Base Armor Rating from 1000 (flat) => 150*mk.

• Sniper Guardians:
  • Base Armor Rating from 1000 (flat) => 150*mk.

• Tachyon Guardians:
  • Base Armor Rating from 400 (flat) => 150*mk.

• Gravity Guardians:
  • Max Target Base Speed from 10/8/6/4/2 => 11/10/9/8/7.
  • Grav Beam Base Range from 8k/10k/12k/14k/16k => 7k/8k/9k/10k/11k.
  • Base Armor Rating from 100*mk => 300*mk.

• Zombie Guardians:
  • Shots Per Salvo from 2/3/4/5/6 => flat 2.
  • Base Armor Rating from 5000*mk => 300*mk.
  • Base Attack Range from 9000 => 7000.

• EMP Guardians:
  • Base Armor Rating from 1200 (flat) => 300*mk.

• Carrier Guardians:
  • Base Armor Rating from 1200 (flat) => 300*mk.

• Fixed a bug with the electric shuttle "chain lightning" mechanic where it could hit a single forcefield way more times than intended. Used to be as many as 200 per shot, but now down to 5 (anything can be hit 5 times per shot).

• Fixed bug where electric shuttles were able to chain-hit the same target 10 times in one blast (thus getting maximum efficiency against groups 20 or larger) instead of 5 (thus needing to hit at least 40 to get full damage).

• Reclamators (excluding zombie-reclamators) :
  • Replaced the "cannot do reclamation damage to ships more than one mk level higher" rule: the reclamation effect of the actual damage done is multiplied as follows:
    • If the reclamator is 4 mks higher than the target (mkV shooting mkI), multiply by 64.
    • If the reclamator is 3 mks higher than the target, 48.
    • If the reclamator is 2 mks higher than the target, 32.
    • If the reclamator is 1 mk higher than the target, 16.
    • If the reclamator is the same mk as the target, 8.
    • If the reclamator is 1 mk lower than the target, 4.
    • If the reclamator is 2 mks lower than the target, 2.
    • If the reclamator is 3 mks lower than the target, 1.
    • There is no 4-mk-lower case because mkV are not reclaimable.
  • Leech Starship base attack power from 120k*mk => 30k*mk.
  • Parasite base attack power from 4000*mk => 1000*mk.
  • Nanoswarms inherent 16x-reclamation property down to 2x, but no reduction in actual damage.
  • Spire Teleporting Leech base attack power also not reduced, because the general feeling is that these are pretty underpowered already. Might nerf these later if this proves to be too much.

• Parasites:
  • Effective range from 3700 => 6000.
  • Armor piercing from 0 => 750*mk.
  • Base Health from 7200*mk => 14400*mk.

• Changed Engineer II and Engineer III caps from 0.75 and 0.5 of the Engineer I cap to 1.5 and 2.0 of the Engineer I cap, respectively.
  • Gives a bit more motivation to unlock the higher engineer marks, if you're having trouble getting enough assistance/repair out.

Misc Changes

• AI War engine upgraded to Unity 3.3, from previously being Unity 3.1.

• Spire Civilian Leaders are now no longer giftable, to avoid a bug whereby gifting them back and forth would repeat the AIP decrease.

• The alternate-victory condition on the Fallen Spire condition has been made a bit more vigorous about winning, since the recently added AI Hunter Killers were refusing to go quietly.

• Mostly as a visual improvement (via more variety of shot speed) and somewhat to make some weapons a little more effective:
  • Laser shot speed tripled (wait, why does a laser shot have a speed?)
  • Minor electric shot speed doubled.
  • Shell shot speed increased to 1.5x what it was.
  • Ion shot speed doubled.
  • Combat style now applies to all shot speeds, not only artificially-increased ones.
  • Note: shots will always move at least 40 faster than their target.

• Made the "signal markers" in Fallen Spire invincible in the same sense as the Dyson Sphere (well, without a certain caveat which applies to the Dyson)

• Just under "Enable Advanced Logging" on the Advanced tab of the Settings window, added a "Enable Reinforcement Logging" toggle.
  • Similar to Enable Advanced Logging, this enables logging of AI reinforcements to the ReinforcementsLog_MainThread.txt and ReinforcementsLog_AIThread.txt files in your RuntimeData directory.
  • This is off by default because:
    • This is really only for when you want to see if there is some problem with the reinforcement logic (short of logging, it is very hard for anyone to tell) and want to show evidence to the developers
    • Reading one of these logs in midgame will give "spoilers" about what the AI has, etc. It's not considered a cheat, however, because all the actually-important info it gives you would be available if you had turned off fog-of-war in the lobby before the game.
    • Lots of disk I/O during the game causes instability on some systems.

New Variety for Existing Minor Factions

• Human Resistance Fighters:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" increment frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Marauders:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" increment frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Human Colony Rebellions
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Dyson Sphere:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" spawn-check frequency (one check per 20 seconds), 1 is 1/4th of 4 (one check per 80 seconds), 8 is 2x what 4 is (one check per 10 seconds), and 10 is insane (one check per 6 seconds).
  • The galaxy-wide population caps on each of the three gatling types are modified by this 0-10 value. 4 is cap = AIP; 1 is cap = AIP / 4, 8 is cap = AIP * 2, etc (it's linear).
  • So if you want a game-dominating dyson sphere, you can still get it :)

• Zenith Miners
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Broken Golems (Hard)
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" strength, 1 is way less than that, and 10 is way more than that (it's not linear).
  • Note: the only thing impacted by the intensity value for Broken Golems (Hard) is the strength of the exogalactic strikeforces sent by the AI as a response to this faction being on. The benefit you derive from it (number or strength of golems, etc) is not affected.
  • Don't use a high value unless you want to see VERY strong exogalactic strikeforces (their frequency does not change).

• Neinzul Preservation Wardens:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" increment frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Neinzul Roaming Enclaves:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" increment frequency, 1 is way less frequent than that, and 10 is way more frequent than that (it's not linear).
  • Don't use a high value unless you want to see these happening a LOT.

• Neinzul Rocketry Corps:
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" number seeded at the beginning of the game, 1 is way fewer than that, and 10 is way more than that (it's not linear).
  • Don't use a high value unless you want to see a LOT of these (Would you like to play a game of... ?).

• Fallen Spire
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" strength, 1 is way less than that, and 10 is way more than that (it's not linear).
  • Note: the only thing impacted by the intensity value for Fallen Spire is the strength of the exogalactic strikeforces sent by the AI as a response to this faction being on. The benefit you derive from it is not affected.
  • Don't use a high value unless you want to see VERY strong exogalactic strikeforces (their frequency does not change).

• Spirecraft (Hard)
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" strength, 1 is way less than that, and 10 is way more than that (it's not linear).
  • Note: the only thing impacted by the intensity value for Spirecraft (Hard) is the strength of the exogalactic strikeforces sent by the AI as a response to this faction being on. The benefit you derive from it (number of asteroids or strength of spirecraft, etc) is not affected.
  • Don't use a high value unless you want to see VERY strong exogalactic strikeforces (their frequency does not change).

• Note that the Trader, Devourer, Easy and Moderate Golems (which don't fit into this idea of "faction intensity", they're really just different things than Hard Golems), Spire Civilian Leaders, and Easy and Moderate Spirecraft are all still just on/off. This isn't because we don't have any ideas on how they might be made so, just none that were very simple to implement at this stage.

New Variety For Existing AI Plots

• Hybrid Hives
  • Now instead of being an on/off toggle in the lobby, it can be set to a value from 0 to 10. 0 is disabled, 4 is the previous "normal" number of hive spawners seeded at the beginning of the game, 1 is way fewer than that, and 10 is way more than that (it's not linear).
  • Don't use a high value unless you want to see a LOT of these.

• Amended the reinforcement logging to better reflect the impact of alert on the order that planets are checked for reinforcement.

• Renamed the AI Eye to the Sentry Eye.

• Bomber Starship renamed to Heavy Bomber Starship.

• Added 25 champion-related achievements.

• The shard seeding in Fallen Spire now only cares about how far from the human homeworlds a planet is, rather than the current border of AI territory. This can make it easier depending on how much territory you've taken (particularly on something like a snake map, but that's something of an edge case), but in light of various recent changes this may not be a bad thing, and the seeding being relative to what you'd conquered was pretty annoying and encouraged some strange playstyles.
  • The last shard is an exception: it still doesn't pay attention to what you've conquered, but it always tries to seed on an AI core planet (so bordering a homeworld, but never on a homeworld).

Bugfixes

• Fixed a moderately longstanding bug where having the first AI player at tech level 1 and the second AI player at tech level 2 would actually draw as the text "II/II" instead of "I/II" on the resource bar.

• Fixed bug on the stats window resource-flows tab where most construction/repair/etc outflows were reported as positive instead of negative.

• Fixed bug that was greatly delaying the onset of gravity slowdown on units entering grav range in some cases.

• FINALLY fixed a long-standing and excruciatingly rare desync that nevertheless was reliably hit by two of our players, Fleet and Tssbackus. Huge thanks to them for their persistence in helping us figure this out. For all the details, here's the mantis issue: http://www.arcengames.com/mantisbt/view.php?id=3077

• The Zenith Trader previously had incorrect line breaks. Fixed.

• Previously, when command stations were replaced, colony ships and mobile builders were still dying even though they shouldn't have been. Fixed.

• Improved the clarity of the colony ship description.

• Previously, when a player was controlling an ally's ships and tried to put that ship into a control group, it would add it to the ally's control group instead of the local player's control group. Having cross-player control groups is not possible without really expanding the control group data structure, which doesn't seem like a good idea to do at the present time, so for the time being allied ships simply won't go into control groups if a player tries to put them in there, at least cutting out the worst of the confusion.

• Fixed a bug with zombie or minor faction electric shuttles, where the shuttles would just sit there. This also should improve the behavior for any ships that explode (including warheads, etc) when controlled by a minor faction, when zombified, when in FRD or attack-move in general, and possibly when in the hands of the AI in some circumstances.

• Previously, zombiefied ships were being bound by player ship caps, but they should not have been. Fixed.

• Fixed a bug where Defender games were always being recorded in the high-score list as lost.

• Fixed a bug where lightning and armored warheads were following the unit-cap-scale.
  • One of the side effects of this is that some Neinzul Rocketry Corps silos (only the lower-mark ones, and only on low or normal caps) were "scaling down" to a maximum internal capacity of zero, effectively making them Neinzul Rocketry Bricks. No longer.

• Fixed a bug where a self-destructing aoe weapon with a limited number of secondary targets (autobomb, nanoswarm, etc) would fail to do any damage to anything in some cases where only the main target was hit.

• Fixed bug where copying in your global controls from disk could lead to building engineer IIs and IIIs without the required tech.

• Fixed a bug that has been causing astro-trains to not be able to attack for the last six months or so.

• Fixed a bug where the "I Have A Bad Feeling About This" achievement was not being awarded.

• Fixed bug where the command-station foldouts each player got on each ally's planet were not executing galaxy-wide or per-planet controls (like Auto-FRD, auto-build engineers, auto-build energy reactors).

• Fixed bug where Core Starships were not immune to translocation.

• Fixed a bug that would cause ships to sometimes ignore their wormhole orders to instead stay and fight when they have a target. This was most noticeable with astro trains since they are never supposed to stop to fight, but it actually was affecting all AI ships since sometime prior to 5.0 most likely. Possibly as far back as 4.0. This had serious implications to the AI's ability to effectively retreat; the AI should be a lot more swift-footed now, while still getting in pot-shots when it decides to leave.
  • Additionally, a potential desync was discovered where players that were using the F3 debug menu and hovering over ships might cause a desync in multiplayer. Fixed that.
    • Fixing this also reduced the per-ship memory footprint by about 8 bytes, which we're always pleased to do.

• Fixed a null exception that could happen rarely when a unit was looking for the next wormhole to move to while moving from planet to planet.

• Fixed an (apparently) rare array-index-out-of-bounds error in the buy-menu "tooltip" window.

• Fixed a longstanding bug where modules would not always decloak their parent ship when firing.

• Fixed a longstanding bug (ever since starships were allowed to be loaded into a transport) where putting a ship with modules into a transport, saving the game, and loading the game would cause the modules to be removed from the game because the game thought they were no longer accessible (the "has a parent ship" link wasn't being checked).

• Fixed some longstanding (since introduction of group move trying to handle immune-to-speed-boosting units) bugs with group move:
  • Previously, for example, a speed-boostable 140-speed unit group-moving with a non-speed-boostable 152-speed unit would result in the boostable ship going 152, and the non-boostable ship going 140 (_below_ its normal speed!). This was because the non-boostable one was taking the minimum speed of it's boostable friends, and then setting its speed limit as if those friends were not actually boostable.
    • This bug has been captured, stuffed, and placed in the museum. It couldn't get away fast enough.

• Fixed a few longstanding bugs where ships were able to fire on targets significantly out of their range because in a few places the range checking was early-out'ing with max(dx,dy). This would have been fine if the target's current-forcefield-radius (and possibly other factors, none confirmed) were applied before the range check instead of after, but that wasn't the case and so the early-out threshold was wrong. Well, it's right now.

• Fixed a problem with too-low throttles on minor faction / zombie ships preventing them from even attacking at all in some cases.

• Fixed a bug where a superterminal could be "paused" by scrapping your command station, saving, and reloading.

• Fixed the bug with two buttons for mark V fighters showing up in the mark V fighter fabricator.

• Fixed a bug where if the AI had a pile of over 2000 zombie threat ships (and over 4000 total threat) it was possible for it to try to redeploy those zombies to carriers, and the carrier would be created with the appropriate contents, but the zombies would not actually be scrapped. Causing the AI to redeploy them again, creating still more carriers, and failing to scrap the zombies again. It was just carriers all the way down, leading rapidly to tens and hundreds of thousands of threat. Fixed so that the scrapping works, and thus the duplicate redeployments no longer happen.

• Fixed a regrettably longstanding bug where it was possible for modules (notably hybrid modules) to become "stranded" without a parent ship and to basically act as somewhat buggy (and, recently, the non-forcefield ones are now invinicible) turrets.

• Fixed a longstanding bug where the number of spire shipyards and spire hab centers supported by a spire shard reactor was being multiplied by the number of human home command stations. The number of spire capital ships of each type supported by a spire shipyard (or refugee outpost, for the frigates) is still multiplied by that number, as is intentional.

• Fixed a bug where a ship with an inherent kiting range (raptor, zenith bombard, etc) would not correct for the radar dampening range of its target.

• Fixed a bug in the auto-build-energy-reactors controls where they were trying to check for remains of other energy reactors but were not checking the right list for those.

• Fixed bug where autobombs/nanoswarms/etc could now target and fly into an aoe-immune ship but couldn't actually do anything to it. They'll now do damage and their other effects as normal, though if it's just a big pile of aoe-immune targets you'll probably want to hold them off for efficiency reasons: they'll only hit the direct target, unless there's something aoe-able in range.

• Fixed a bug where if the AI had a pile of over 2000 zombie threat ships (and over 4000 total threat) it was possible for it to try to redeploy those zombies to carriers, and the carrier would be created with the appropriate contents, but the zombies would not actually be scrapped. Causing the AI to redeploy them again, creating still more carriers, and failing to scrap the zombies again. It was just carriers all the way down, leading rapidly to tens and hundreds of thousands of threat. Fixed so that the scrapping works, and thus the duplicate redeployments no longer happen.

• Fixed a really, really longstanding bug where the AI got 1 bonus reinforcement at diff 8, 1 bonus reinforcement at diff 9, and 2 bonus reinforcements at diff 10, but no bonus reinforcements at any of the non-integer difficulties between those. Now it's 2 for 10 and 1 for >= 8

• Fixed a longstanding bug (since ship cap scales were added, roughly) where the ship-cap-scale multiplier was being applied twice to reinforcement calculations (so high got the normal amount, normal got half as much as it should have, low got 1/4 as much as it should have, etc).
  • The effect isn't as severe as it probably sounds, because most of the calculations ended in "if less than 1, set to 1" and other bugs (see below) were letting the AI get a lot more mileage than was mathematically sanitary out of that 1.

• Corrected some longstanding issues where reinforcements were frequently getting stuff like Spire Blade Spawners in roughly the same proportion as Fighters; the ship cap multipliers were being used, but if a particular individual group of spawns in a reinforcement (there could be over 10 in a single reinforcement-of-planet event, and multiple such events can happen on a single planet per overall reinforcement) only had 1 ship to pick it could freely pick either a fighter or a spire blade spawner (or whatever it had unlocked) and the fact that it had just gotten something like 30x as strong as what "1" normally means was simply ignored. If it had 20 ships to pick and picked a blade spawner that would be the end of the pick, though, so it wasn't totally ignorant of that dynamic.
  • Now it does a bit of arithmetic "carrying" so that if one individual spawn really picks something low-cap like that (and it still can do that, otherwise stuff like blade spawners just won't happen in most reinforcements) then it remembers that for further spawns in that reinforcement event, which can make certain reinforcements a little lopsided but in general is at least a balanceable situation now,

• Fixed a bug where the Dyson Sphere itself was alerting its planet and neighboring planets on the AI thread, but not on the main thread. It now does not alert on either (but the actual enemy-to-all gatlings and player-ally gatlings count for alert purposes, of course).

• Fixed a bug in the handicap modifier computations that caused negative handicaps to be far harsher than they were supposed to be in some circumstances (like making CPA size zero when it was supposed to be, say, 24, because the multiplier was about 51x smaller than it was suppposed to be).

• Fixed a longstanding bug where AOE "canister" shots (flak, grenade, plasma-siege) were not properly hitting forcefields when targeting a ship protected by them, unless the targeted ship was also in range of the explosion from the point where the shot hit the forcefield.

• Fixed a longstanding bug where an Ion Cannon killing a warhead would cause it to not properly detonate as if it had been killed by a Warhead Interceptor.

• Fixed a bug where the AI would not attack armored warheads (due to a "very low priority target" flag on those unit definitions).

• Fixed a bug where when the AI picked engineers or remains-rebuilders for reinforcements, it was "spending" too much on them.

• Fixed a fairly longstanding bug where player ships would not autotarget an AI ship that was under a (non-module) forcefield and guarding something because that ship had not set its "angry" flag (to avoid leaving the protection of the forcefield). Now it considers the ship eligible for shooting if either it or its guard post has been angered recently.

• Fixed a bug in the last version where the laser turret upgrades could show the stats on laser drones instead of... laser turrets. Similar for other turret techs that also unlock drones.

• Fixed a longstanding bug where a unit could both be reclaimed and regenerated, which would sometimes lead to an AI unit being reclaimed, zombified, and warped off somewhere deep in AI territory. Not great for threat management.
  • Fixed a related bug where a unit that had swallowed other units or was tractoring other units would take them with them when regenerated. Now those units simply cannot be regenerated unless they have no such "load".

• Fixed a bug where neinzul drones (which cannot go through wormholes) could inherit a go-through-wormhole order from the enclave starship building them, and would actually go through the wormhole.
  • Now they'll generally _move_ to a point near the target wormhole, but they won't try to traverse the wormhole itself.

• Fixed the colony ship tooltip which was still saying it had to be near the target construction point, and also amended its hint about replacing existing command station to reflect recent changes.

• Fixed a bug where beam and warbird starships were still in the "Starship mk III" buy-group for the AI; they're now in the mkV group where they're actually balanced to be now.

• Fixed a longstanding bug where the "prevent construction of new command station by a different player than the current controller of the planet" rule was preventing gifting of command stations.

• Fixed a longstanding omission preventing low-power ships under a forcefield protecting a passive guard post (and perhaps other guard posts) from coming out of low-power when the forcefield/post/guards were attacked.

• Fixed a longstanding bug where Decloakers in waves (only possible by having 1 wave-sending AI and 1 support-corps AI) were not having their ship cap multiplier applied and thus were appearing in vastly larger numbers than they should have been. Wasn't a big deal before, but would have been devestating with their new stats.

• Fixed some bugs with the Spire Archive gain-knowledge logic producing erratic display results (and not displaying how much it had gathered out of the total 9000, etc).

• Fixed a longstanding bug where repairing units with a repair boost (which is most of the ones that can repair) were repairing units faster but still "spending" metal and crystal at the rate they would have if they had no repair boost. The cost difference between repairing with an MRS and with an EngieIII was pretty big. No longer.

• Fixed a bug where multiple human homeworlds (in SP or MP) were not increasing the guard post part of reinforcements.

• Fixed a longstanding (back to the unity port) bug where the buy/tech menus would also draw the buttons from the menus above them, commonly resulting in "ghosts" of build queue buttons, etc.

• Fixed a longstanding but very intermittent null-exception in tech-menu-button rendering.

• Fixed bug where core/starship/experimental fabricators were not being considered by the D keybind.

• Fixed a bug where remains currently in the process of rebuilding could get "stuck" at 1 hp despite being under very heavy fire until its actual stored resources were exhausted.

• Fixed a longstanding bug where buy menu buttons were showing up red outside supply even when the ship doing the building can build fine outside supply.

• Fixed a longstanding bug where the AI would only produce carriers if the target planet already had 2 barracks, instead of 1.

• Fixed a bug where Fallen Spire events would spawn the recoverable objects (and thus anything they go on to build) as belonging to the first player even when the first player was champion-only. Now it gives them to the first normal or normal+champion player.

• Fixed omission in the new CPA-size logic that was making it not consider one of the later parts of the wave-size calculations (making it a bit too high on lower difficulties and not as high as intended on higher difficulties).

• Fixed a longstanding bug in FRD target sorting that was sometimes being overly sensitive to differences in how far away two targets were.

• Fixed a bug where it was possible to start placement of an item (like an engineer or turret) and switch to another planet via control-group-selection and still be in the middle of placing that item (and if you had supply, actually be able to place it).

• Fixed a bug where a bunch of science ships on auto-knowledge-gather but with no eligible targets could "clog" the auto-explore throttle such that auto-explore for scouts would stop working. Now auto-knowledge-gather has its own separate throttle.

• Made "what can I autotarget?" logic more consistent so that player-ally minor factions don't pop AI carriers.

• Fixed a bug where viewing a Core Spire Corvette Fabricator could cause unhandled exceptions (and the queue buttons to not show).

• Fixed a really longstanding bug where giving a ship an attack order against something already in range would cause it to move towards the target a little bit before stopping again. This was mildly annoying with most units but very troublesome with units like fortresses and zenith siege engines that have to recharge (either their ability or their weapon) after moving at all.

• Fixed a bug where minor faction and zombie ships could be affected by the player auto-kite controls.

• Fixed a bug where a cpa deploying could cause unhandled errors.

• Military Command Stations:
  • Fixed a bug where the mkII and mkIII versions no longer had a unit cap (leaving it that way was entertained, but ultimately would require more sweeping changes to command stations than is a good idea right now, if it would be a good idea at all).
  • Fixed a bug where the mkII and mkIII were still showing the mkI descriptive text, despite that now not being the same.
  • The knockback logic now launches stuff 2000 range units outside the station's effective range rather than 2000 range units inside, to help the autotargeting not get stuck on a few targets when its main use is best served by spreading shots out.

• Fixed a bug where the fallen spire campaign could spawn two of the same shard-signal.

• Fixed a longstanding bug where the galaxy-nav code of AI-ally Dyson Gatlings was getting distracted from its main purpose.

• Fixed a fairly longstanding bug where disabling an expansion from within the lobby would not properly remove several expansion-specific options (map types, setup scripts, minor factions, etc), which could lead to very odd behavior.

• Fixed a couple longstanding bugs that would cause very old saves to not load properly.

• Fixed a bug where hybrids, exo ships, and the like could switch to threat-fleet behavior.

• Fixed some bugs where special forces and strategic reserve spawns could result in an AI player controlling a type of ship only available to the other AI.
  • And a related bug where some of the special forces spawn logic could only access the special ship types of the first AI player.

• Overhauled how line-place determines space between points to be closer to how the game will actually accept or reject a given placement. Hopefully this will avoid "skipping every other one" bugs without adding excessive padding between points on a packed-line pattern.

• Fixed rare null exception that could happen when changing control groups.

• Fixed a bug where changing per-planet settings on one planet, then opening the CTRLS window from another planet would do the confirm-discard-changes check, which led to various other tomfoolery.

• Fixed a bug where MkV Spider fabricators were still considered experimental fabricators instead of core fabricators, which incidentally handles an issue where there was a very high chance of there being 2 of these per map.

• Now if you use the mid-game Manage Players interface to switch a player from the Normal role to Helper and then save that, it properly tells you that you can't do that rather than trying (and generally causing a desync).

• Fixed a moderately longstanding bug where if while starting a new game you select a planet as a homeworld, then deselect it (in favor of some other homeworld(s) ) before actually starting the game, and that planet happens to have the superterminal on it, the superterminal would start the game in "active mode". Needless to say, that would be a short game.

• Fixed a really longstanding bug where shots on a planet were sometimes visible even through the fog of war.