| *** OkinaMatara <OkinaMatara!~Yukari@user/yukari> has joined #bzflag | 00:15 | |
| *** OkinaMatara <OkinaMatara!~Yukari@user/yukari> has quit IRC (Quit: Quit.) | 04:07 | |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has quit IRC (Read error: Connection reset by peer) | 07:08 | |
| *** BulletCatcher <BulletCatcher!~bc@user/bulletcatcher> has quit IRC (Ping timeout: 256 seconds) | 07:49 | |
| tupone | I would like to share some work I have done on python plugin server side. I have ported the first 4 (alfabetically) test plugin. I want to do a PR on 2.4 without the intention to merge now, just to share | 09:07 |
|---|---|---|
| BZNotify | bzflag: atupone opened pull request #306 "Rewrite some plugins in python" (https://github.com/BZFlag-Dev/bzflag/pull/306) | 09:12 |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has joined #bzflag | 14:12 | |
| *** bboles_ is now known as bboles | 17:32 | |
| macsforme | this is the first bad SDL commit of the macOS issue where the window resizes and goes halfway offscreen when you press F1 the first time: https://github.com/libsdl-org/SDL/commit/69518b9ecc1b987ce2962216289318fdd44ece46 | 17:35 |
| blast007 | might be because we destroy and recreate the window when we toggle fullscreen | 17:40 |
| blast007 | is this something we want to fix in 2.4? and if so, do we want to rip out all the non-SDL2 platform code and redesign our platform wrapper and main game code around the capabilities of SDL2? | 17:42 |
| blast007 | that, or we could move some of the logic in the game code (such as MainWindow::toggleFullscreen) into the platform code, and then fix the SDL2 code to not re-create the window | 17:43 |
| blast007 | I did have that start of new SDL2 platform code that was designed more around the game code telling the platform code what it was wanting and the platform code then figuring out how to get it to that state | 17:45 |
| blast007 | I'm personally fine with removing non-SDL2 code in 2.4 | 17:54 |
| blast007 | we've already been defaulting to SDL 2 for a while, and more recently requiring passing an extra configure flag to use something other than SDL2 | 17:54 |
| blast007 | somewhere I have some code that added haptics (rumble feedback) to the SDL2 platform code | 17:55 |
| macsforme | my first thought was just to revert that patch, re-build SDL 2, and see if that makes it work for now | 17:55 |
| blast007 | does it only affect macos? | 17:56 |
| macsforme | but yeah, if there's a possibility of further 2.4 releases, we should probably try to mitigate it in 2.4 | 17:56 |
| macsforme | I'm not sure if other platforms are affected | 17:57 |
| blast007 | I have some time off starting tomorrow and I was planning to work on 2.6 next week | 17:57 |
| macsforme | what was your linux window issue? you said something about two monitors? | 17:58 |
| blast007 | yeah, when using Wayland, at least, it's trying to span across both windows, but only shows up on the larger one | 17:58 |
| macsforme | that should be fun to diagnose | 18:00 |
| blast007 | is there any reason to keep the client -view options other than normal and anaglyph? | 18:01 |
| blast007 | getting rid of those other view mode could simplify a lot of other code.. | 18:02 |
| blast007 | (the other modes being stuff for old outdated VR headsets and multi-projector rooms, and GL stereo that doesn't even work on most GPUs) | 18:03 |
| blast007 | and I guess at that point, is it even worth it to keep the anaglyph mode? | 18:04 |
| blast007 | anaglyph doesn't require rendering multiple scenes, at least, so it's much simpler than the others | 18:05 |
| blast007 | I'll play around with gutting all and the other platform code in a 2.6 fork | 18:06 |
| macsforme | yeah those could all go away as far as I'm concerned... I wouldn't mind someday supporting multiple 3D views on multiple monitors (maybe), but I'm sure that would all have to be redone anyway | 18:08 |
| blast007 | my newPlatform code supports multiple windows and monitors | 18:08 |
| macsforme | nice | 18:09 |
| blast007 | I think it still ran into some issues with SDL2 though even with it being a fresh start, so some of our workarounds would have to carry over to that | 18:09 |
| blast007 | I'm planning to work on some of the broken changes or half finished features in 2.6 | 18:12 |
| blast007 | I've had the idea of making three tiers of goals for 2.6. Tier 1 would be features/changes we deem required for a release of 2.6 and would be developed in-tree. Tier 2 would be features/changes we'd really like to have, and if we were otherwise ready to release we might delay the release for a while to finish these, and would be developed in-tree. If it was going to cause too much of a ... | 18:16 |
| blast007 | ... delay, however, we'd discuss and possibly pull the changes. Tier 3 would be nice to haves that we'd develop in a branch and only merge when they were done. Does that seem reasonable? | 18:16 |
| macsforme | sounds reasonable... I'm personally good with moving to more of a rigid PR model for any substantial change... I'm thinking that would reduce the amount of half-completed stuff that gets committed | 18:27 |
| macsforme | it looks like the -window option no longer causes the game to launch in windowed mode... I'm trying to figure out since when | 18:28 |
| blast007 | on macos? | 18:34 |
| macsforme | at least, yeah... trying linux now | 18:34 |
| blast007 | it worked for me on linux the other day | 18:35 |
| macsforme | looks like throwing out my config fixed it... odd | 19:05 |
| macsforme | oh, I guess -view persists when you try it out :-P | 19:07 |
| macsforme | another good reason to get rid of it | 19:09 |
| blast007 | yeah... :) | 19:27 |
| blast007 | I think it's also a big user of glScissor, so getting rid of that might also help me fix our glScissor issues | 19:28 |
| tupone | what is the status of ipv6 ? | 20:02 |
| blast007 | nothing in the code itself | 20:19 |
| blast007 | I've made a few attempts at it, but never got far enough for it to be functional | 20:19 |
| tupone | the problem is in bzfs or in the web/forum/authenticator ... | 20:21 |
| tupone | is it in the 2.6 1st tier goal ? | 20:23 |
| tupone | i see that servers are sell now with ipv4 as an option | 20:24 |
| blast007 | bzfs, game client, web | 20:32 |
| blast007 | and yes, IPv6 I feel should be a tier 1 feature | 20:32 |
| blast007 | I'll compile a list of areas that need work for IPv6 support | 20:33 |
| macsforme | blast007: I re-read your earlier comments... I would be very much in favor of refactoring our platform code around the SDL2 paradigm versus continuing to try to shoehorn SDL2 fit into our twenty-year-old platform model... that should definitely reduce our headaches around SDL2 | 21:50 |
| macsforme | I think the whoooooole application/display/window models and the associated platform code could all be replaced with a much simpler model | 21:52 |
| *** BulletCatcher <BulletCatcher!~bc@user/bulletcatcher> has joined #bzflag | 21:53 | |
| macsforme | if there's a model that will work consistently with SDL2 plus any other platform libraries we may be interested in transitioning to, all the better | 21:55 |
| blast007 | the newPlatform code I was writing had both SDL2 and GLFW | 22:03 |
| blast007 | GLFW lacks audio code though | 22:03 |
| blast007 | hmm, and rumble feedback support | 22:06 |
| blast007 | GLFW does use SDL2's SDL_GameControllerDB, which we don't currently use | 22:08 |
| blast007 | https://wiki.libsdl.org/CategoryGameController#introduction | 22:09 |
| blast007 | I guess we should also document how to add the Let's Encrypt root certificate on our site, at least for macOS | 22:25 |
| BZNotify | bzflag.org: blast007 opened issue #48 "Document how to install the Let's Encrypt ISRG Root X1 certificate on older macOS" (https://github.com/BZFlag-Dev/bzflag.org/issues/48) | 22:30 |
| BZNotify | bzflag.org: blast007 labeled issue #48 as documentation | 22:31 |
| *** Zehra <Zehra!~Yukari@user/yukari> has joined #bzflag | 23:30 | |
| *** Zehra <Zehra!~Yukari@user/yukari> has quit IRC (Quit: Quit.) | 23:52 | |
Generated by irclog2html.py 2.17.3.dev0 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!