| *** Optic_Delusion <Optic_Delusion!~Optic_Del@pool-96-235-40-180.pitbpa.fios.verizon.net> has joined #bzflag | 00:41 | |
| *** Optic_Delusion_ <Optic_Delusion_!~Optic_Del@96.235.40.180> has quit IRC (Ping timeout: 252 seconds) | 00:43 | |
| *** OkinaMatara <OkinaMatara!~Yukari@user/yukari> has joined #bzflag | 02:32 | |
| OkinaMatara | This may be a stupid point. What may be a future outlook for plug-ins, as distribution of them on non-*nix platforms becomes less easy. | 02:34 |
|---|---|---|
| OkinaMatara | In short: macOS and Windows plug-ins aren't as easy to provide. (Assuming most of the player base uses Windows & macOS) | 02:35 |
| blast007 | actually, Windows is the easiest one to provide them for | 02:41 |
| blast007 | You don't even need the full BZFlag source code if you wanted to build a plugin from source. We ship the API headers and the library, so you can compile directly against those. | 02:42 |
| blast007 | long term though, a script language is probably the way to go | 02:43 |
| OkinaMatara | sounds good, it answers all of my questions. | 02:44 |
| blast007 | tupone has been working on Python support, though I'm not sure if I'd consider that the best option. It might be interesting to have something that we could put inside of a BZW file, and python's whitespace requirements are a bit annoying. | 02:46 |
| OkinaMatara | ^Related to this on a tangent, I've been looking to "optimize" plug-in writing. I came to around 3 observations from this. | 02:49 |
| OkinaMatara | 1. On event stuff is easy. 2. It's tricky managing state. 3. The API could reduce a lot of "tedious" code. | 02:50 |
| OkinaMatara | In theory, you can have `OnCap, give [flagcode], send [msg], givepoints 5` by events alone, it's just passing around values or modifying states where the real "complexity" arises. | 02:53 |
| OkinaMatara | The point: Stuff like flag handling, game state...etc are difficult, since a lot of state needs to be managed, compare to the API handling it. | 02:55 |
| OkinaMatara | Probably one of the best examples: https://github.com/Zehra/Radar/blob/main/Radar.cpp | 02:57 |
| OkinaMatara | Die, Dropped, Transferred, Captured events "trigger" status change, and they're being tracked. | 02:59 |
| OkinaMatara | But on second thought, depending on how fully featured it would be, a scriptable/bzw plug-in language probably could replace "most" plug-ins to some extent. (If not all.) | 03:00 |
| OkinaMatara | IMO, I'd find Python very "interesting", but if a smaller/more efficient language is used with a good API, I'd say it's probably better. | 03:04 |
| Agatha | Lua is traditional for game plugins and mods, but from what I've seen it sucks, especially in comparison to Python. | 04:02 |
| OkinaMatara | IIRC, Scope and the "off by one", and the lack of "good" handling from data, are some of the annoyances. | 04:05 |
| OkinaMatara | of* instead of from* | 04:06 |
| Agatha | "from" is not incorrect, though if we're nitpicking my idle comments for some reason, I could have used slightly better commas. | 04:09 |
| OkinaMatara | it was in reference to my own. "the lack of 'good' handling from data", when instead "of" would have been a better choice. | 04:13 |
| *** OkinaMatara <OkinaMatara!~Yukari@user/yukari> has quit IRC (Quit: Leaving) | 04:31 | |
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-98-219-34-189.hsd1.ga.comcast.net> has quit IRC (Ping timeout: 244 seconds) | 04:34 | |
| Agatha | laaa I read your message 10 times and I didn't see it somehow | 04:58 |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has quit IRC (Read error: Connection reset by peer) | 08:09 | |
| *** FastLizard4 is back | 09:51 | |
| *** FastLizard4 is now away: AWAY from keyboard | 10:12 | |
| blast007 | http://www.squirrel-lang.org/ http://chaiscript.com/ https://www.angelcode.com/angelscript/ (and many others listed here https://github.com/dbohdan/embedded-scripting-languages ) | 10:47 |
| *** FastLizard4 is back | 11:18 | |
| *** FastLizard4 is now away: AWAY from keyboard | 11:38 | |
| *** FastLizard4 is now away: GONE - Screen Detached and Disconnected from IRC (I'm probably asleep, at work, or doing something in real life) | 12:03 | |
| blast007 | Juest: here's a signed and notarized build that should work on 10.13 or newer. Would you want to test that? https://static.bzexcess.com/downloads/BZFlag-2.4.27.20241210-macOS-universal.zip | 12:25 |
| *** Cobra_Fast is now away: offline | 12:31 | |
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-98-219-34-189.hsd1.ga.comcast.net> has joined #bzflag | 12:31 | |
| *** Cobra_Fast is back | 12:35 | |
| *** Cobra_Fast is now away: offline | 14:37 | |
| *** Cobra_Fast is back | 14:37 | |
| *** Cobra_Fast is now away: offline | 15:14 | |
| *** Cobra_Fast is back | 15:14 | |
| *** Juest0 <Juest0!juest@user/Juest> has joined #bzflag | 16:37 | |
| Juest0 | hi, what did i miss? i should get on my znc but im currently booted to another os | 16:38 |
| blast007 | I had a link to a macOS test build. Do you want that again? | 16:40 |
| *** SpringTank is now away: No clients connected to this core right now. | 17:09 | |
| *** SpringTank is back | 17:10 | |
| Juest0 | i got it on discord thanks | 17:30 |
| blast007 | k | 17:40 |
| *** Cobra_Fast is now away: offline | 17:49 | |
| *** Cobra_Fast is back | 17:49 | |
| Juest0 | anything else? | 18:03 |
| Juest0 | i can see later when i get on znc but for now.... | 18:03 |
| tupone | blast007: do you want a fix for high altitude? I have an idea on how to fix without protocol change, but I will take a week or more | 18:05 |
| *** TimRiker <TimRiker!TimRiker@2607:fa18:9ffe::196> has joined #bzflag | 18:16 | |
| blast007 | I'm assuming that fix would change how clients behave? | 18:42 |
| tupone | only for high frame rate client | 18:52 |
| tupone | would not change dead reckon | 18:52 |
| blast007 | I probably won't hold up the 2.4.28 release for it, but I'd be happy to help try out the change. We could always turn around and push a 2.4.30 quick if it's something we want to include there. | 19:00 |
| Juest0 | hmm interesti | 19:00 |
| Juest0 | oh hey TimRiker :D | 19:00 |
| Juest0 | i feel the ipv6 work kinda died down for the time being | 19:00 |
| blast007 | TimRiker: and if you're curious about what we're talking about: "< tupone> blast007: do you want a fix for high altitude? I have an idea on how to fix without protocol change, but I will take a week or more" | 19:01 |
| blast007 | Juest0: it's something we'll work on again, but since it has to be in 2.5 it's taken the back seat for a while. The bzfls3 web code will be able to handle IPv6 with the REST API, which will need to be fleshed out more. | 19:02 |
| Juest0 | understood | 19:05 |
| Juest0 | priority is getting a 2.4 release currently, yes? | 19:06 |
| blast007 | 2.4.28, bzfls3, new server for the web services (Debian 11 -> 12, no weird load balancer setup) | 19:23 |
| blast007 | I have an updated replacement for the texture hosting site too that I need to finish up | 19:23 |
| blast007 | I'll want to have us go through changes that are already in 2.5 and see if there's any we want to rip out so that we don't end up in another 2.99 situation | 19:25 |
| Juest0 | tank skins are feature creep maybe? | 19:32 |
| Juest0 | hmmmm | 19:33 |
| Juest0 | i do wonder, are there any 2.5 things that can be implemented in 2.4 easily or it has to wait? | 19:33 |
| blast007 | no, those are something that are easy to implement and finish | 19:33 |
| blast007 | things I'm wondering about are larger things, like how we split out flag effects from flag types | 19:33 |
| Juest0 | oh that | 19:33 |
| Juest0 | i see | 19:34 |
| blast007 | I don't recall how "done" that is | 19:34 |
| Juest0 | got a checklist? | 19:34 |
| blast007 | [ ] Make checklist | 19:35 |
| Juest0 | lol oop | 19:35 |
| blast007 | :) | 19:36 |
| Juest0 | wait what happened with the bzflag org? | 19:36 |
| Juest0 | its empty | 19:36 |
| blast007 | BZFlag-Dev is the org | 19:37 |
| Juest0 | there's a bzflag org | 19:37 |
| blast007 | someone else had already created the bzflag org before us and had paid private repos on it | 19:37 |
| Juest0 | oh, but it says bzflag information on the description | 19:37 |
| blast007 | yeah, but it's not ours | 19:38 |
| Juest0 | #79 seems easily doable? | 19:38 |
| Juest0 | hmmm, there are better options for dns than c-ares indeed | 19:39 |
| blast007 | yeah, that one is mostly done and is already functionaly, just a bit noisy in the logs | 19:39 |
| Juest0 | unrealircd uses something different | 19:39 |
| blast007 | c-ares is probably fine to keep now | 19:39 |
| blast007 | here was a list of 2.5 changes: https://docs.google.com/document/d/e/2PACX-1vTyJ9QlufMACveMmH3cyKr8zERuma1h_gsJkrLV6--y2FrMJVPH5d-z8w9dBo5nkTSQGMMp2pwqrwij/pub | 19:40 |
| Juest0 | has it been considered using lua or squirrel? | 19:51 |
| Juest0 | for scripting | 19:52 |
| blast007 | lua I don't like | 19:59 |
| tupone | python is a well known language :) | 20:01 |
| allejo | managing Python runtimes are a pain, no? | 20:04 |
| tupone | I don't seem. I have ported some plugin | 20:05 |
| allejo | would BZFlag come bundled with a python runtime? or would BZFlag be reliant on the system installation? | 20:06 |
| Juest | great question | 20:06 |
| tupone | In my tree bzfs link python library. | 20:07 |
| Juest | statically? | 20:07 |
| tupone | system library | 20:07 |
| blast007 | Windows doesn't have a system library | 20:07 |
| blast007 | system installation* | 20:08 |
| Juest | python can be installed "system-wide" on windows | 20:08 |
| Juest | hexchat calls python from PATH as you'd expect | 20:08 |
| blast007 | yeah, but then are we doing that, or requiring that the use does it? or would we just have to bundle it into our install directory? | 20:08 |
| blast007 | user* | 20:08 |
| Juest | hexchat handles lua and perl out of the box in windows | 20:09 |
| Juest | well, it seems so | 20:09 |
| Juest | but python plugin failed to load because it seems like python is not installed | 20:09 |
| Juest | lua is built-in | 20:09 |
| Juest | perl and python require system installation | 20:10 |
| Juest | lua doesn't necessarily require it | 20:10 |
| Juest | or at least it seems the case with software | 20:10 |
| Juest | znc also wants system installation for perl and python modules | 20:10 |
| * Juest shrugs | 20:11 | |
| Juest | just some examples :) | 20:12 |
| Juest | blast007: any words about the person who already got the bzflag org? interests in giving it? | 20:15 |
| BZNotify | 2.4 @ bzflag: atupone pushed 1 commit (https://github.com/BZFlag-Dev/bzflag/compare/6089084130c2...c71893fb592e): | 20:17 |
| BZNotify | 2.4 @ bzflag: atupone c71893: WorldInfo: addWall is no more a method (https://github.com/BZFlag-Dev/bzflag/commit/c71893fb592e013b9bc29253e76a57831ba6711c) | 20:17 |
| BZNotify | bzflag: atupone synchronized pull request #334 "Use glm" (https://github.com/BZFlag-Dev/bzflag/pull/334) | 20:18 |
| Juest | would be great to fix that spam connect crash | 20:19 |
| tupone | Juest: are you able to reproduce ? | 20:19 |
| Juest | tupone: https://github.com/BZFlag-Dev/bzflag/issues/128 | 20:20 |
| tupone | it is still happening with 2.4 head ? | 20:21 |
| Juest | great question | 20:21 |
| Juest | let me try | 20:21 |
| Juest | huh seems its fixed \o/ | 20:30 |
| tupone | I saw the stack trace in the comment and I guessed it is | 20:31 |
| Juest | fun | 20:32 |
| Juest | good to know | 20:33 |
| Juest | feel free to close it then :) | 20:35 |
| Agatha | Sidenote: the main consideration for plugins in my book would be hotswap debugging. Dream would be a single hotkey that reloads all plugins (and rejoins you as necessary, it shouldn't be for many plugins), without having to recompile and reboot bz or run shell commands. | 20:36 |
| Juest | hope you thought on the security of that | 20:40 |
| blast007 | tupone: whoops, thanks, forgot to remove that from the .h | 20:47 |
| Juest | hmmmmmm | 20:47 |
| tupone | yw | 20:48 |
| Juest | ah i remember the 2.6 changes doc, what is outdated and what is still valid? | 20:48 |
| *** FastLizard4 is back | 20:59 | |
| blast007 | Juest: that doc should still be pretty accurate | 21:04 |
| Juest | irix is long dead, right? | 21:05 |
| blast007 | yep | 21:05 |
| Juest | how about haiku or os/2? | 21:05 |
| blast007 | haiku might work, but I don't know if anyone tests it | 21:05 |
| tupone | blast007: My idea was to fix the bumping at high altitude, but does not fix the movement when far from center. Now I only imagine to go double, but that change protocol. So no way for now | 21:07 |
| Juest | im taking my time to upgrade my software on this windows 10 installation, while i defrag my other disk | 21:07 |
| Juest | cant really test debug bzf because of ucrtd missing | 21:08 |
| blast007 | tupone: the network protocol could remain using 32-bit floats even if the local simulation calcuations use 64-bit floats | 21:15 |
| tupone | all solution to have a good physic in a game is to have fixed timeslot for all players | 21:15 |
| blast007 | but... might still be better to wait for a proto break | 21:15 |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has joined #bzflag | 21:16 | |
| blast007 | yeah, I believe 2.99 had some work towards that goal. I can't recall if that only split up the simulation steps in low FPS situations or if it also held off on simulation updates until dt grew large enough. | 21:16 |
| tupone | all players should compute physics using the same time interval | 21:17 |
| blast007 | that might be *less* important if the server was authoritative and the client would update it's own state based on what the server told it | 21:18 |
| tupone | or we could compute physic only when enough time passed | 21:19 |
| tupone | this last I can think more on it | 21:21 |
| blast007 | a hacky fix for 2.4 would just be to limit the max FPS to about 180 :) | 21:22 |
| tupone | and also need to limit how far you can go | 21:23 |
| blast007 | I noticed physics issues starting at around 200 FPS, specifically where your tank seems to bounce off buildings when you jump at them | 21:24 |
| blast007 | the high altitude one started around 500, IIRC, at least at the altitude I was testing | 21:25 |
| blast007 | https://www.gafferongames.com/post/fix_your_timestep/ | 21:29 |
| Juest | lol i triggered a c++ EH exception by reproducing the issue, tupone :) | 21:32 |
| Juest | its the bug | 21:32 |
| tupone | do you have a core ? | 21:33 |
| Juest | no, im live on windbg | 21:33 |
| Juest | unless you are familiar with the commands there | 21:33 |
| Juest | to dump it | 21:33 |
| Juest | its windows | 21:33 |
| tupone | going to sleep. But if you have a stack trace, add to the issue I can see it tomorrow | 21:34 |
| tupone | 2.4 head ? | 21:34 |
| Juest | its the same as it was discussed on irc back then | 21:34 |
| blast007 | Juest: try the 'k' command | 21:34 |
| Juest | Almost tip of the head but yes it is HEAD as of 8 december 2024 | 21:35 |
| Juest | blast007: i already did KP on raiseexception | 21:35 |
| Juest | https://hastebin.skyra.pw/raw/olareqotas | 21:35 |
| tupone | I saved that stack trace locally. I will check asap | 21:38 |
| BZNotify | bzflag: Juesto commented on issue #128 "Client crash when hitting Connect multiple times" (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2532948811): reproduced on HEAD as of dec 8 2024, i think it's still very incons... | 21:40 |
| Juest | good night tupone | 21:43 |
| *** FastLizard4 is now away: AWAY from keyboard | 22:00 | |
| BZNotify | bzflag: ODelusion commented on issue #128 "Client crash when hitting Connect multiple times" by Juesto (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2533125284): ... | 22:50 |
| *** BulletCatcher <BulletCatcher!~bc@user/bulletcatcher> has joined #bzflag | 23:04 | |
| *** Tobbi <Tobbi!~Tobbi@SuperTux/Tobbi> has quit IRC (Ping timeout: 248 seconds) | 23:14 | |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!