Author Topic: SMACX AI Growth mod  (Read 7699 times)

0 Members and 1 Guest are viewing this topic.

Offline bvanevery

  • Emperor of the Tanks
  • Librarian
  • *
  • Posts: 2163
  • €1315
  • View Inventory
  • Send /Gift
  • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
  • Planning for the next 20 years of SMACX.
  • AC Text modder Author of at least one AAR
    • View Profile
    • Awards
Re: SMACX AI Growth mod
« Reply #270 on: December 05, 2018, 07:07:50 PM »
  • Publish
  •   ACK!!  My finished 1.25 merge is rubbish! 

    In fairness, I did release 1.25 almost a month ago, and I've been communicating many of my intents the whole time since then.  This is the reason I don't personally support Yitzi or other patches that require alteration to alphax.txt.  The right way to handle the problem, ongoing, is with an automated rules based merge script that simply replaces whatever Yitzi needs to be happy, every single time.  And to ship that script with some kind of updated Yitzi package, like bundle it with Python or Lua or whatever, so that it runs when the user installs.

    Pracx already has some scripted install infrastructure in it, BTW.  It's proof of concept, it alters some files, I forget whether just the Alpha Centauri.ini.  That might be worth looking at if you really want to solve "the Yitzi problem" once and for all.

    Another approach is to alter Yitzi to not need any change to alphax.txt in the 1st place, but that would require going up a learning curve about what he did.  Not worth it to me personally.  I don't need his mod features, and I'm not victimized enough by the bugs he fixed to care.

    By contrast, I checked on Scient 2.0 recently and it is compatible with my modding, as any changes it makes to alphax.txt are optional or unimportant.  Scient added the ability to specify the reactor size for predefined units, but you don't have to do that.  He changed some of the short facility and secret project descriptions to be more helpful, but those aren't necessary.

    Quote
    An integrated version of the best of Yitzi's and Inductio's stuff would definitely catch my attention.

    Don't plan on that.  I do skim Induktio's development posts occasionally.  Last I looked, he wasn't interested in that as a priority.  He has very good engineering reasons to hold that view, i.e. lotta work figuring out what Yitzi did, not a lot of obvious payoff for bothering.  If he did change his mind about that, I would still expect him to put it at much much lower priority than his own new changes.

    In the real world, that means it won't happen.  This is the sort open source problem that gets solved when someone who really wants something, steps up and does the work themself.  I fully expect Induktio mostly wants Induktio's work, not Yitzi's work.  I think if anyone with "ASM mucking chops" really wanted to follow in Yitzi's footsteps, they would have done so already.  It's a difficult arena to get into, much harder than writing new C/C++ code.

    Quote
      Of course, that would change the basic parameters and probably require a UI Growth Mod adjustment to conform.

    There's never going to be any "conformance" on my part for Induktio's work.  We do not have a good working relationship.  If he chooses to design a patch that will work with anybody's alphax.txt, no changes required, and he doesn't violate core assumptions about how alphax.txt works in the process, then we may finally have an entente.  Engineering-wise it's the Right Thing To Do when working in a community with other people, rather than assuming one's own work is the only thing that counts.

    I don't know what Induktio's views are nowadays, nor what he's implemented in that regard, nor what his implementation priorities are.  I suspect he has his hands full with basic AI design, testing, and stability issues, because he bit off a lot to chew in that regard.  I know how many bugs or misfeatures I typically come up with per X number of changes on my part, and I think he's working at a factor of at least 10x mucking things around.  So unless he's a godlike release engineer, which frankly I have no reason to believe based on past interaction with him, I expect him to be stabilizing and polishing for several months yet.

    I do think he may get there eventually, if he doesn't burn out, or something doesn't intervene in his life that prevents him from doing more work.  He's basically competent at shoveling out large quantities of work, i.e. he's made releases, he hasn't quit.  Engineers who keep that up, do eventually reach an endgame, a sort of "maintenance mode" rather than developing lots of new features.  And then they tend to exhale deeply and say, good grief, where's my time, where's my money....

    I don't trust anyone to actually get to that point, until they've actually gotten there.  I've been through enough open source debacles to know what the lifecycles of people's energies are.  The main variables are 1) how much previous experience someone has doing "large scale" $0 work for.  Like if it's your 1st time doing something big, something is probably going to clobber you at some point.  2) how young the person is.  Financial realities become seriously pressing for most people as time goes on.  Where's my car, where's the roof over my head, where's my Significant Other....

    I'm an outlier and not immune to the financial pressure problem myself.  I definitely am feeling a pressure that I need to move on from this, to start a real commercial project.  1.26 could be my last release, I'd be perfectly happy if it was.  I'd feel I'd crossed some kind of finish line.  But people have this way of showing up and saying, "What about this?"

    Changing the Fundamentalist dialogue in xscript.txt so that it doesn't assume we're talking about God, is something I'd like to do.  However I already had enough stuff for 1.26, it was time to kick all that out the door.  The Usurpers talking like humans when berating people about Power, is another wart I'd like to fix.  But if my life moves on and these don't happen, then they don't happen.

    What I can guarantee, at this point, is I'm moving to a 1 month release cycle at the earliest.  This has to do with getting views and downloads.  /r/4Xgaming got me 2.5 times more traffic than previous forums I tried, and the downloads seemed to keep coming for basically a month.  I cleared 70 downloads, a new high for me.  I'm doubting there are more forums to be had, but I haven't managed to make a Steam Community post with a proper URL in it.  They were blocking those, I think because I didn't have a posting track record, or at least I hope so.

    The next frontier is YouTube.  In coming months, I will probably be focusing a lot less on developing the work, which I already did and it's good / fine.  I'll be focusing a lot more on getting exposure for the work.  I figure this is part of my learning curve as a game developer, something I can justify and has future value for me.

    Offline Vidsek

    Re: SMACX AI Growth mod
    « Reply #271 on: December 08, 2018, 08:58:37 PM »
  • Publish
  •    bvanevery,  commenting on redoing my merge was actually a very obscure complement on your continuing hard work.  I'm always delighted to redo or edit the merge if you've refined your mod :)  After all, I'm primarily doing it for myself as a quiet little pastime when I'm trying to avoid housework.  If and when I have a product that seems stable and faithful to your mod, I'll cheerfully offer it to everyone in our community who would like to use it, test it, improve it, or ignore it.

      The rest of my comment was actually never-gonna-happen-in-this-world wishful thinking.  I'm a semi-old guy with utterly trivial coding skills who's ok with being resigned to playing with whatever more competent people graciously offer :)

      I only referenced Inductio's work since he seems to have a bug-fix or two not in Yitzi's work that looked useful.

      Adjustments to preserve the intended balance and function of your AIGrowth mod I consider a proper and necessary part of any "merging", and to be done by the merger (eg. myself) as a part of that process.  Fortunately, with Yitzi's patch, it has so far been no more than setting some of his new parameters to match those of basic SMACX.

      I've found your insights, opinions, and advice to be informative and interesting, and to have given me a better understanding of the game mechanics.  For me, that makes actually *playing* the game more fun for various reasons.

      Much as I've enjoyed following you through your practice/learning experience here (with the bonus of a decent mod to play with), at this point I'm actually more interested in the game or games you'll produce yourself in the future.  Do keep us informed on progress, you know many of us would be delighted to discuss aspects of it with you.....obsessively.  Hmmmm, then again maybe you should keep us some of us out of the loop, just for sanity's sake ;lol

     
    All this talk of fungus and worms makes me hungry...

    Offline bvanevery

    • Emperor of the Tanks
    • Librarian
    • *
    • Posts: 2163
    • €1315
    • View Inventory
    • Send /Gift
    • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
    • Planning for the next 20 years of SMACX.
    • AC Text modder Author of at least one AAR
      • View Profile
      • Awards
    Re: SMACX AI Growth mod
    « Reply #272 on: December 09, 2018, 05:21:18 AM »
  • Publish
  • Heh, well I'm not always sure what people want, so thanks for letting me know.

    I think you're going to find 1.26 plays rather differently than some previous releases.  Moving all the social engineering choices (except Democracy) to Tier 3 has a pretty substantial effect on gameplay.  I find I have to spend more time coming up with solutions for basic early game stuff, without a pile of tools and gewgaws to do it.  I'm ok with that, I think it's a good thing.  If ever it was a goal to get rid of the boring "here's another weapon, here's another armor" dynamic, well I've certainly done that.  I did relent from making it into a total rockfight though, there's Synthmetal Armor at Tier 1 and strength 2 Lasers at Tier 2.  Particle Impactors really do feel like the supreme weapon though, because they're strength 4 and are typically shooting at strength 3 armor.

    I'm going to start my own game effort soon.  I'm just putting out real life fires right now, with my car having gone to Transmission Heaven unfortunately.  Almost got a really sweet replacement, a 2003 Toyota Avalon in immaculate condition, but for sentimental reasons the owner decided they're not ready to part with it.  So it's back to the drawing board.

    Offline Vidsek

    Re: SMACX AI Growth mod
    « Reply #273 on: December 09, 2018, 05:23:12 PM »
  • Publish
  •   Curious.  I had always assumed transmission issues were somehow associated with Hell...

      Condolences on the Avalon.  I'm fresh out of anti-sentiment drugs or I'd send you some to slip to that owner.
    All this talk of fungus and worms makes me hungry...

    Offline bvanevery

    • Emperor of the Tanks
    • Librarian
    • *
    • Posts: 2163
    • €1315
    • View Inventory
    • Send /Gift
    • Allows access to AC2's quiz & chess sections for 144 hours from time of use.  You can't do without Leadship  Must. have. caffeine. -Ahhhhh; good.  
    • Planning for the next 20 years of SMACX.
    • AC Text modder Author of at least one AAR
      • View Profile
      • Awards
    Re: SMACX AI Growth mod
    « Reply #274 on: December 09, 2018, 08:37:45 PM »
  • Publish
  • Sadly, the objecting party may go his own way with dementia soon enough.  But in the present, his wife died of cancer, his daughter has his wife's car (the Avalon), and the high school age granddaughter does not in fact want it.  So he's very sad right now, things precious to him slipping away.  Too bad a high school girl can't see that car with the cold pragmatism I can!  I mean, what's wrong with driving around with a sun roof and 4 of you in the super nice leather seats?  I wonder what she thinks is a good car instead.  It's not my place to get into it though.  They're neighbors across the street, but I don't know them well.

    In modding news, I hereby claim that earlier hovertanks is not in fact overpowered.  They don't fight any better than other stuff, particularly when attacking cities.  I'm really kind of bored with them.  I consider them grossly inferior to building rails right up to a city's doorstep.

    Barring that, mindworms and a high PLANET rating are mostly more effective at attacking cities, unless the defender starts building Trance and Trance 3-Res units.  The AI doesn't do so much of that, probably because I made the predefined Trance 3-Res Sentinel unit a Tier 4 tech.

    I think it's kind of dumb that I can't really do strategic bombing.  The distances I have to cover on a Huge map are too large for Penetrators to be viable.  Well I suppose I could build a bunch of air bases but it all feels so tedious.

    Offline vonbach

    Re: SMACX AI Growth mod
    « Reply #275 on: Yesterday at 01:17:15 AM »
  • Publish
  • Sorry to hear about your car and his troubles. My condolences.
    « Last Edit: Yesterday at 10:57:00 AM by vonbach »

     

    * User

    Welcome, Guest. Please login or register.
    Did you miss your activation email?


    Login with username, password and session length

    Select language:

    * Community poll

    SMAC v.4 SMAX v.2 (or previous versions)
    -=-
    14 (6%)
    XP Compatibility patch
    -=-
    8 (3%)
    Gog version for Windows
    -=-
    62 (27%)
    Scient (unofficial) patch
    -=-
    22 (9%)
    Kyrub's latest patch
    -=-
    14 (6%)
    Yitzi's latest patch
    -=-
    83 (36%)
    AC for Mac
    -=-
    2 (0%)
    AC for Linux
    -=-
    5 (2%)
    Gog version for Mac
    -=-
    9 (3%)
    No patch
    -=-
    10 (4%)
    Total Members Voted: 229
    AC2 Wiki Logo

    * Random quote

    You are the children of a dead planet, earthdeirdre, and this death we do not comprehend. We shall take you in, but may we ask this question?will we too catch the planetdeath disease?
    ~Lady Deirdre Skye 'Conversations With Planet'

    * Select your theme

    *

    Facebook Comments