Prerelease 3.155 (Perf+, Cloaker Star+, Elec Bombers+, Bugfixes)

[x4000]

I'm getting what I think are CPWs even though I'm *fairly* certainly I had that disabled in my current game. I recently knocked out the two warp gates next to my home planet, and now I'm getting waves to "" which end up eventually jumping into my home planet after a few minutes. That sounds like a cross planet wave but I've never played with the feature turned on.

This is probably just this, which has always been a feature and was the inspiration for CPWs as a setting (people wanted for it to be an option to have it always do that): http://arcengames.com/mediawiki/index.php?title=AI_War_-_No_Warning_On_Waves%3F

Moral of the story is to keep at least one warp gate adjacent to your planets if you want to know where waves will be coming, or the AI will land somewhere else.

[Signata]

Ah, thanks, I just never noticed that because I rarely knock out all the gates. Fortunately in this particular case, I have two wormholes right next to each other, so the turrets are pulling double-duty and maintenance has been minimal.

[x4000]

Glad it works out! Good way to find out, versus getting stomped.

[Winter Born]

Ah, thanks, I just never noticed that because I rarely knock out all the gates. Fortunately in this particular case, I have two wormholes right next to each other, so the turrets are pulling double-duty and maintenance has been minimal.

When you take your next planet you can leave an adjacent warp gate intact for a known spawn point for incoming waves

[Buttons840]

So, what was the story behind the energy fluctuation bug?

[x4000]

So, what was the story behind the energy fluctuation bug?

I actually have no idea.  I never could find the problem, the only solution was to rewrite it from scratch, using a different premise, and that worked.

[Lancefighter]

how.. exactly did the shield boosters get nerfed?


(for the record, i just started a game with them.. only to find one of my AI's is a starship captain guy. Yay, shield boosters were slightly useless against shield-skipping enemies..)


as for energy use..
also, it was once at a time, that moving ships through wormholes sometimes incurred double energy use.. this was because of how the engine checked for energy use in systems (sometimes it would count them twice), so instead it was changed to use no energy when moving through wormholes? i think.. not entirely sure, it didn't flash my forcefields offline everytime my engineers jumped systems, so it was ok for me.

I'm kinda interested in how the new system differs from the old system, however...

[x4000]

how.. exactly did the shield boosters get nerfed?


(for the record, i just started a game with them.. only to find one of my AI's is a starship captain guy. Yay, shield boosters were slightly useless against shield-skipping enemies..)

I'm really not clear on what you're asking.

as for energy use..
also, it was once at a time, that moving ships through wormholes sometimes incurred double energy use.. this was because of how the engine checked for energy use in systems (sometimes it would count them twice), so instead it was changed to use no energy when moving through wormholes? i think.. not entirely sure, it didn't flash my forcefields offline everytime my engineers jumped systems, so it was ok for me.

I'm kinda interested in how the new system differs from the old system, however...

Yup, it's the same side of the other issue.  Nothing real special about the new system, the best part is how simple it is.  It just counts up all your ships, in transports or not, adding all their energies together about 4x per second.  That used to be something that was wrapped into the simulation loop, but that was the root of the other problems, hence the change to it being its own loop.  This also does let us skip it for non-human ships very easily, which might actually have a slight performance gain, anyway.  Certainly not a hit from testing, anyway.

[Lancefighter]

"In the last version, munitions boosters and shield boosters were a couple of orders of magnitude too strong. Fixed. "
what exactly changed?


and yeah, giving the energy stuff its own loop does seem to make slightly more sense then checking ships on a per-planet basis

[x4000]

"In the last version, munitions boosters and shield boosters were a couple of orders of magnitude too strong. Fixed. "
what exactly changed?

Well, in the immediate prior prerelease I had changed how their stuff was calculated -- how many ships they could boost at once, based on their square area in their new range.  But I forgot to account for the fact that that stuff was squared, and so the number of ships they were able to boost was often 10X or more too many.  This just fixed the bug to bring it back to what their stated values were more supposed to be in the first place.

[Buttons840]

how.. exactly did the shield boosters get nerfed?


(for the record, i just started a game with them.. only to find one of my AI's is a starship captain guy. Yay, shield boosters were slightly useless against shield-skipping enemies..)

I'm really not clear on what you're asking.

as for energy use..
also, it was once at a time, that moving ships through wormholes sometimes incurred double energy use.. this was because of how the engine checked for energy use in systems (sometimes it would count them twice), so instead it was changed to use no energy when moving through wormholes? i think.. not entirely sure, it didn't flash my forcefields offline everytime my engineers jumped systems, so it was ok for me.

I'm kinda interested in how the new system differs from the old system, however...

Yup, it's the same side of the other issue.  Nothing real special about the new system, the best part is how simple it is.  It just counts up all your ships, in transports or not, adding all their energies together about 4x per second.  That used to be something that was wrapped into the simulation loop, but that was the root of the other problems, hence the change to it being its own loop.  This also does let us skip it for non-human ships very easily, which might actually have a slight performance gain, anyway.  Certainly not a hit from testing, anyway.

Is this a case of "after writing it the first time, I feel like starting over because I've learned so much in retrospec."  I get that feeling a lot, and being obsessive, I usually end up fallowing it, meaning I start over a lot, and don't get much done.

[Lancefighter]

"In the last version, munitions boosters and shield boosters were a couple of orders of magnitude too strong. Fixed. "
what exactly changed?

Well, in the immediate prior prerelease I had changed how their stuff was calculated -- how many ships they could boost at once, based on their square area in their new range.  But I forgot to account for the fact that that stuff was squared, and so the number of ships they were able to boost was often 10X or more too many.  This just fixed the bug to bring it back to what their stated values were more supposed to be in the first place.

ah ok.

Oddly, i think my version was actually updated to 3.1.5.5 before i started, so no real change for me.

[Amonchakad]

Remains Rebuilders still cost 0 energy instead of 25 

[superking]

    * Ships in transports now still cost the normal energy they would cost outside the transport.
          o Previously, transports could be used to too great of effect in terms of minimizing energy needs simply by loading ships into them.

    * When the energy count is being shown for ships (either by pressing Alt+A, or via the net energy being low for you), transports now show their own energy plus that of all the ships inside themselves.

one less exploit 

[x4000]

Buttons -- yes, I guess you could say that. It was not anything real new that I had learned. But I gave up on one type of performance consideration and went with another instead. Just a general bit of refactoring, actually really common in a codebase so complex and large as ai war (which is actually tiny compared to some I have worked on).

Lancefighter -- exactly.

Amonchak -- thanks for the note, that is odd, I'll look at that again.