Jump to content

sanderd17

WFG Retired
  • Posts

    2.225
  • Joined

  • Last visited

  • Days Won

    77

Posts posted by sanderd17

  1. The checksums match, so your download is correct. If you say it's your first installation, and you already tried reinstalling 0 A.D. and rebooting the computer, then I can only guess it's something 0 A.D. can't handle w.r.t. your system.

    Can you check what log files you have (I hope the game at least got time to write some logs before crashing)? See http://trac.wildfiregames.com/wiki/GameDataPaths on where to find the logs.

    Certainly the system-info.txt would help, and also the crashlogs.

    If there's no system-info.txt, perhaps you can give us some information on how your system looks like (what CPU, what graphics card, if you have experimented with some drivers, ...)

  2. Yes, you can cycle through multiple markets, but afaik, it doesn't give you extra gain.

    Here's the formula used: http://trac.wildfiregames.com/browser/ps/trunk/binaries/data/mods/public/simulation/helpers/TraderGain.js

    So it depends on the square of the straight-line distance, with some bonuses like for trading with a foreign market.

    It is planned however to also take into account the number of markets you pass.

    • Thanks 1
  3. We currently get quite some help from the distro guys wrt testing on different platforms, taking care of the dependencies, and distributing releases.

    If we'd move to AppImages, we would most likely have to do that testing ourselves, taking care of the dependencies by ourselves, and also host the linux release packages (which isn't that cheap to do for such a big binary with so many downloads).

    So yes, it's possible, but I don't see the advantages in it for us (as we are already in most distros).

  4. Hmm, now there are holes in the antlers. Also note that the game only renders tris from one side. So it's better to make the lantlers round, as else you'll have to duplicate the tris anyway.

    Also, AFAIK, it's best to make sure the tris are more-or-less symmetric. It should make it easier to model and also easier to animate.

    I'm no artist however, so I can't give advise on how to achieve these things.

    • Like 1
  5. Hi, thanks for wanting to work on some models.

    First of all, a technical notice: our engine can only handle triangles. While most 3D tools can "triangulate" models, it often doesn't make optimal use of the triangles. Certainly for more natural objects, like animals. Buildings are more square by nature, there a triangulate operation can give good enough results. So for animals, it's better to design in triangles already.

    Now, for the actual model, I find the body way too rough. It should be made a lot rounder and more organic. But you also should keep it low-poly. So there are a lot of triangles in the antlers that aren't needed (those won't be seen), and there are also strange triangles on the bottom of the animal.

    Try to keep the model as organic as possible with the least tris possible.

    But in any case, thanks for wanting to work on this.

     

    • Like 2
  6. I can understand it's hard work when you're making a mod that doesn't only add features (like adding a new civ), but also modifies features. Certainly modifying templates currently means that you have to copy a lot of data over, and keep the things you want unchanged synched with the main game.

    I hope we can find some solution for #4076 soonish, which would certainly help maintainability of mods by reducing the amount of data that has to be copied from the start.

    So I agree that modders need to know what they're getting into, but it's only thanks to modders that we know what's still lacking in our game wrt moddability. So we really appreciate your effort and the effort of other modders.

    PS. if a single commit sets you back an hour or more, it's likely someone on our side also worked an hour or more on related data changes. Ofc, our team is a bit bigger, so I agree it's harder for modders.

    • Like 7
  7. Glad you like it.

    But just to save you from disappointment, Alpha 21 most likely won't come out this month. That date was set when we released the previous version, and was just a rough estimate then (as trac needs some date). But it looks like we'll now aim for somewhere in August (though we also have a tendency to seriously overshoot our aims). We prefer to release a working game over releasing quickly.

    • Like 4
  8. @Mah-Dry-Bread, thanks a lot for mentioning it on your channel. Looking forward to seeing your streams.

    Btw, most devs usually gather on IRC ( https://kiwiirc.com/client/irc.quakenet.org#0ad-dev ), we don't play that often (as devs tend to have a broken version of the game most of the time), but if you give a shout there, I guess some of us could follow the stream and join your chat if we have time.

    • Like 2
  9. That's probably related to one of the formation bugs. If the formation entity spawns on something impassable, it can't move and gets stuck. And as the entity spawns just in the geometrical center of the units, this can happen randomly. The problem isn't in the spawning of the formation controller, but in the pathfinder that shouldn't care about the controller being on an impassable spot (as the old pathfinder didn't care about it).

    The other possibility is that the units forget their tasks when put in a formation. This is because the formation controller remembers the tasks of all units, and formations were never designed to be optional (i.e. in the original design, units were always in formation). So when a formation controller gets created now, all tasks (also movement tasks) are forgotten.

    Both issues should indeed be solved sometime.

    Even with single-digit FPS, you should still be able to see units move after 20-30s, so the problem is likely one of the issues above, and not a strict performance problem.

  10. Our mods are also sandboxed, and shouldn't be able to get access to random files, or get access to system memory. So at most they would be able to crash the game.

    But it all depends on how good the sandboxing is. It's easy to have a bug that suddenly does expose parts of the filesystem or memory. That said, SpiderMonkey is made to execute scripts from unknown origin (scripts from webpages), so the sandboxing on their side should be done quite well.

    • Like 1
  11. Yes, there's a crude outline to what should happen:

    • One of the first things is making it possible to install mods by just opening a *.pyromod file with 0 A.D. (instead of copying the files manually to the right place, which causes some confusion).
    • Then the lobby should also display the installed mods, and possibly link to a page defined in the metadata.
    • And a final step would probably be to enable direct downloads (if that doesn't give a security issue).
    • Like 1
  12. Why would we limit great mod functionality to just maps?

    From our perspective, it's about as hard to send files on the fly for generic mods as it would be for maps alone (I'd even argue that generic mods would be easier by now, given the parts of the infrastructure we already have in-place). So I see no reason to limit the functionality to maps alone.

    And to turn your map into a mod, it just has to be zipped with a bit of meta data. This could very well be done as an "export" option in Atlas.

  13. Catapults have splash damage, but they have been nerfed a bit, as it was sometimes annoying that a catapult would pick a new target, and fire at some spot that was very close the the own units, and thus kill the own units with it.

    For catapults to work better, we should probably first have focus-fire, so a way to command catapults and other units to fire at an area rather than at a single target.

×
×
  • Create New...