IRC log for #bzflag on 20190704

00:16.34*** join/#bzflag infobot (ibot@c-174-52-60-165.hsd1.ut.comcast.net)
00:16.34*** topic/#bzflag is https://BZFlag.org || https://www.openhub.net/p/bzflag || https://bzflag.org/help || Channel Logs: http://infobot.rikers.org/%23bzflag/ || Current version: 2.4.18 || New shiny website! https://forums.bzflag.org/viewtopic.php?f=8&t=20201
00:16.34*** mode/#bzflag [+v infobot] by ChanServ
00:17.26ZehraWill 'Metal' be supported by BZFlag?
00:17.35blast007Zehra:
00:17.42blast007whoops, bumped enter
00:18.11ZehraHappens to the best.
00:18.13blast007Zehra: I was looking at bgfx a while ago as a possible route forwards, which would let us support Direct3D, Metal, and OpenGL with a single API
00:18.31blast007it uses a GLSL-like shader language
00:18.51blast007https://github.com/bkaradzic/bgfx
00:20.33blast007likely more feasible using that then it would be for us to include and maintain support for multiple graphics APIs ourself
00:22.42blast007it would potientially limit our use of third-party graphics libs, such as a user interface library, as it would have to have a bgfx renderer
00:25.34blast007I was messing around with bgfx, trying stuff like toggling between window and fullscreen, taking screenshots, toggling vsync, etc (stuff other than just drawing, since I can assume that works well enough)   https://github.com/blast007/BGFlagFX
00:28.33ZehraSeems reasonable, if everything works properly it would mean more up to date API's would be used at the possible expense of third-party library usage.
00:29.04blast007there were some performance issues with screenshots in certain renderers because it did some pixel format conversion in the main thread, inside the bgfx code itself (so I can't easily move it to a thread)
00:31.59blast007think it was one of the older OpenGL versions and Direct3D 9 that was slow for screenshots
00:32.54blast007ah, in *debug* builds it was slow for OpenGL, but it was okay in release builds
00:36.06ZehraSeems good so far, thankfully the performance issues were within the debug, but not in the release builds.
00:37.22ZehraWhat would be some of the major downsides if the switch is made?
00:37.31*** join/#bzflag cods (~fred@tuxee.net)
00:43.35blast007dunno
00:44.41blast007tupone has been working on moving to OpenGL 2.0, gradually replacing the old fixed function graphics code and moving to shaders, which would ease moving to newer OpenGL versions (or bgfx)
00:50.56ZehraI'm thinking that there may be some bugs which could be difficult to solve if bgfx is used, such as where the implementation causes a bug to appear based on the specific graphics API used.
00:53.37Zehra(sample: Direct3D3, Direct3D12, Metal, OpenGL 3.1 work, other versions have some inconsistency.)
01:17.42*** join/#bzflag blast007_ (~blast007@bzflag/developer/Blast)
01:19.01*** join/#bzflag Zehra (~Zehra@unaffiliated/zehra)
01:25.31*** mode/#bzflag [-v short_circuit] by ChanServ
01:27.12*** join/#bzflag shamdango (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)
04:38.29*** join/#bzflag TheWalker (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)
04:42.59Flashwhat do you think about #pragma once vs include guards?
05:10.11*** join/#bzflag TheWalker (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)
05:11.19DTRemenakI use #pragma once on any project that doesn't have to be too widely portable (vs/gcc/clang all support it, but there were holes outside that last time I checked, particularly sun cc).
05:11.59DTRemenakit has the potential to be substantially faster and I find it's less error-prone in practice (although that assumes you are not involving crazy filesystem shenanigans)
05:12.01FlashWhen was the last time you used Sun CC
05:12.06DTRemenaklike 2014 ;)
05:12.17Flashthey are converging on gcc
05:12.31DTRemenakdon't have any idea what oracle has done with it
05:12.43FlashI think as a separate compiler, it's dying. Oracle has their own Linux
05:12.59Flasheverything Linux is gcc
05:13.25FlashSolaris and Linux will ultimately converge, I would wager
05:13.40DTRemenakand I assume we don't target mipspro anymore?
05:14.09Flashin any event, I do know that the C++11 version of the Studio compiler uses gnu name mangling and ABI compatibility
05:14.20DTRemenakthat's convenient for sure
05:14.25Flashis mipspro even a thing anymore?
05:14.40DTRemenakI'm sure it died with SGI unless they open-sourced it
05:14.44FlashSilicon Graphics abandoned the best OS ever in favor of linux
05:18.03Flashanyway, I'm with you, although we are currently using both, as I nudge my team into modern C++
05:21.24BZNotifybzflag: 03jose1711 commented on issue #201 "bzflag is not working properly with multiple monitors" (06https://git.io/fj6cj): > What Linux distribution and distribution version are you running? ...
05:22.39FlashArch? we support Arch?
06:11.10*** join/#bzflag Sgeo_ (~Sgeo@ool-18b98455.dyn.optonline.net)
06:49.43BZNotifybzflag: 03atupone 11synchronized pull request #148 "Use of Vertex Buffer Objects instead of Client Arrays" (06https://git.io/vhCKr)
07:00.47BZNotifybzflag: 03atupone commented on pull request #199 "Missile guidance math rewritten" (06https://git.io/fj6CP): I had two pull request is sequence. Most of your comments apply to ...
12:06.09*** join/#bzflag I_Died_Once (~I_Died_On@unaffiliated/idiedonce/x-1828535)
14:27.14*** join/#bzflag Dracos85 (~dracos85@24-119-188-49.cpe.cableone.net)
16:43.56*** join/#bzflag TheWalker (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)
17:44.46*** join/#bzflag TheWalker50 (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)
18:03.13moriah~Delusional++
18:08.57BZNotifybzflag: 03atupone 11synchronized pull request #194 "Use glm instead of our implementation (vectors.h)" (06https://git.io/fjYFq)
18:23.14BZNotifybzflag: 03atupone 11synchronized pull request #199 "Missile guidance math rewritten" (06https://git.io/fj0lw)
18:47.59tuponejwmelto? who is here ?
18:52.48BZNotifybzflag: 03atupone 11synchronized pull request #148 "Use of Vertex Buffer Objects instead of Client Arrays" (06https://git.io/vhCKr)
19:05.41macsformeI had an issue with triple monitors in Linux as well... I forget the exact issue, but I think it tried to create one window that was like 5K pixels wide (the width of my three monitors put together), which failed, so I couldn't launch the program fullscreen
19:38.00allejojwmelto == flash
19:40.24blast007I'm wondering if it's just trying to create a window at a resolution not supported on the primary screen.  might have another resolution stored in the config.
19:41.20blast007I've had the flashing window issue before, I think..
19:43.40tuponeallejo: ty
19:49.50*** join/#bzflag spldart (~james@2601:2c5:c600:2365::7c1e)
19:49.51*** join/#bzflag spldart (~james@bzflag/contributor/spldart)
19:49.51*** mode/#bzflag [+v spldart] by ChanServ
20:02.28macsformethere was a flashing window when we had the issue with scaled retina MacBook displays, where it tried to create a window with a given resolution and it was failing
20:45.27BZNotifybzflag: 03atupone 11synchronized pull request #199 "Missile guidance math rewritten" (06https://git.io/fj0lw)
20:49.54BZNotifybzflag: 03atupone 11synchronized pull request #148 "Use of Vertex Buffer Objects instead of Client Arrays" (06https://git.io/vhCKr)
20:52.01*** join/#bzflag the_map (~the_map@unaffiliated/the-map/x-1795707)
22:34.22*** join/#bzflag dngor (abuse@p3m/dngor)
22:45.03*** join/#bzflag dngor (abuse@p3m/dngor)
22:50.22*** join/#bzflag dngor (abuse@p3m/dngor)
23:22.16*** join/#bzflag TheWalker (ac0d4a9c@gateway/web/cgi-irc/kiwiirc.com/ip.172.13.74.156)

Generated by irclog2html.pl Modified by Tim Riker to work with infobot.