IRC log for #bzflag on 20170503

00:27.20*** join/#bzflag infobot (ibot@rikers.org)
00:27.20*** topic/#bzflag is https://BZFlag.org || http://ohloh.net/p/bzflag || https://bzflag.org/help || Channel Logs: http://infobot.rikers.org/%23bzflag/ || 2.4.10 released! https://forums.bzflag.org/viewtopic.php?f=8&t=19401
00:27.20*** mode/#bzflag [+v infobot] by ChanServ
02:29.51*** join/#bzflag BZnotify (~BZnotify@192.30.252.34)
02:29.51BZnotify[13bzflag] 15blast007 pushed 1 new commit to 06master: 02https://git.io/v9Rw8
02:29.51BZnotify13bzflag/06master 14371b9d1 15Scott Wichser: Merge remote-tracking branch 'origin/2.4'...
02:29.51*** part/#bzflag BZnotify (~BZnotify@192.30.252.34)
03:05.55*** join/#bzflag nitroxis (n@nxs.re)
03:13.31*** join/#bzflag nitroxis (n@nxs.re)
03:32.05*** join/#bzflag BZnotify (~BZnotify@192.30.252.45)
03:32.05BZnotify[13bzflag] 15blast007 pushed 2 new commits to 06master: 02https://git.io/v9RKE
03:32.05BZnotify13bzflag/06master 146fb0b62 15Scott Wichser: Revert dialog system commits:...
03:32.05BZnotify13bzflag/06master 14c20aa8e 15Scott Wichser: Fix warning about loss of precision.
03:32.05*** part/#bzflag BZnotify (~BZnotify@192.30.252.45)
13:50.36*** join/#bzflag infobot (ibot@rikers.org)
13:50.36*** topic/#bzflag is https://BZFlag.org || http://ohloh.net/p/bzflag || https://bzflag.org/help || Channel Logs: http://infobot.rikers.org/%23bzflag/ || 2.4.10 released! https://forums.bzflag.org/viewtopic.php?f=8&t=19401
13:50.36*** mode/#bzflag [+v infobot] by ChanServ
14:17.29*** join/#bzflag spldart (~spldart@bzflag/contributor/spldart)
14:17.29*** mode/#bzflag [+v spldart] by ChanServ
14:27.44spldartAnyone else play'n hookie?
14:28.19the_mapI don't usually fish during the week
14:29.07spldartWhat about sit on the porch swing smoking a cigar and drinking a beer during the week?
14:29.32spldart... or a glass of Merlot and pipe.
14:31.03spldartStorms coming later today and I'm excited.
14:38.17the_mapoh, tell Storm I said hi
14:38.33spldartRoger that.
14:53.47blast007Hulk smash storm!
14:54.12spldartXD
15:03.40blast007Another interesting thing might be a boolean to define a server as running a custom game mode that a plugin can set to true.
15:04.20blast007So, if there's an HTF or AHOD plugin, the server won't identify as "CTF" or "FFA" but instead "Custom"
15:05.25spldartOr hulk just fly over turbulence https://3.bp.blogspot.com/-qG8wVx4Th7Q/V6iIQiJtE5I/AAAAAAAAYU4/AlfcRdVRTz8trNgIHUUhliybXanG58UFgCLcB/s1600/hulk-quinjet.jpg
15:08.39blast007Also part of 2.6 should be a new auth system so that IPv6 support can be possible.  Whether or not we end up getting IPv6 support in the game for 2.6, it'd be nice to have one less roadblock in its way.
15:14.28allejodoes the client know what CTF and FFA mean on the server list or does it just blindly output info from the server list?
15:15.07blast007it knows
15:15.30blast007the list provides a numeric value, so it could probably just be value 255 of that
15:15.48blast007though I still want it to behave like CTF/FFA/Rabbit/OFFA, so maybe not
15:16.31blast007I also want to fix the hostban issue for 2.6 :P
15:17.02blast007at least I assume it's still a problem that some clients caught in a hostban hang during the join instead of being rejected
15:57.10*** join/#bzflag Zehra (~Zehra@unaffiliated/zehra)
16:17.52JeffMallejo: yes you can just add a find method
16:18.26JeffMor provide a free api function bzapi_stringFind()
16:31.26JeffMyou could swap the networking system over to somethign that supports ipV6 already,  I think raknet supports it
16:34.46JeffMwould not need to make bzfs use all the fancy features of the library, just need to get the data blob across
16:41.05*** join/#bzflag bertman (~bert@unaffiliated/bertman)
16:48.26blast007I don't know which raknet fork we'd use though :P
16:48.46blast007the one oculus dumped hasn't been touched even though people have posted issues
16:49.13JeffMis there an active fork?
16:49.14blast007and I'd think our networking system is the least of the troubles with moving to IPv6
16:49.18blast007I have no idea
16:49.32JeffMyou think services is the major hurdle?
16:49.49blast007auth and services, and bans
16:50.04blast007and bans isn't even that terrible
16:50.07JeffMauth/services are the same to me :)
16:50.17blast007:)
16:50.23blast007to me it's a specific part of services
16:50.30JeffMsure
16:50.34blast007we'd need to change the list as well
16:50.37blast007and weblogin
16:50.59blast007probably move weblogin to OAuth
16:51.20blast007make a proper group system too
16:52.27blast007the new forums can use OAuth, though I'm unsure what caveats there are with that (like, I'm not sure if we tie existing accounts in to an external OAuth, or how it handles name changes)
16:52.45JeffMgmail accounts for everyone!
16:53.06blast007the ideal way would probably be using an email address as the login username and then having callsigns assigned to that
16:53.26blast007and then you log into the game client and it could do fun stuff like syncing your settings and whatnot
16:53.32JeffMthat could let you allow multiple callisigns per user
16:53.37JeffMsomething that has been discussed
16:53.41blast007yup
16:53.54JeffMcharacter select :)
16:54.06JeffMand none of this "I lost my name because I changed it" BS
16:54.53blast007yeah, I'd have grace periods for name changes (at least for the last X names)
16:55.08JeffMI'd let people keep 2-3 names
16:55.15JeffMand buy extras :)
16:55.21blast007heh
16:56.18JeffMthen you'd also have a nice account to attach hats too
16:56.25blast007the official raknet repo "Latest commit 1a16989  on Feb 27, 2015", and has 45 issues and 42 pull requests.
16:56.40JeffMthen it's meh
16:56.45JeffMI'm sure there is something else out there
16:56.49blast007enet
16:56.55JeffMsure
16:56.58JeffMforgot about them
16:57.00JeffMthey were good too
16:58.01blast007https://github.com/OculusVR/RakNet/issues/3
16:58.30blast007escaped SQL... mmm... tasty
16:58.38blast007apparently it's still 1995
17:00.52JeffMheh
17:01.24JeffMIIRC enet didn't have as many fancy features, but it'd be a pretty decent dropin for what bzflag does
17:02.00blast007http://enet.bespin.org/Features.html
17:03.11blast007the "Channels" part would let us send low priority stuff like chat on a separate channel so that resends of a chat message don't delay stuff like flag grabs/drops
17:04.14JeffMyeah
17:04.23JeffMthere is some stuff that would make things easier
17:04.26blast007what things from RakNet would be things we might have wanted that enet doesn't have?
17:05.49JeffMthe RPC stuff was nice
17:05.53JeffMbut not required
17:06.02JeffMshared objects and stuff
17:06.13blast007http://www.grpc.io/
17:06.14JeffMprobalby woudl have made the code too hard to understand anyway
17:06.19blast007https://capnproto.org/
17:06.27JeffMthere ya go then
17:06.58JeffMso use enet, get the basic transport working using the current message system, then look into RPC messages later
17:07.01JeffMeven simpler
17:07.58blast007I'd probably push a network library change off to 2.8 though
17:08.33JeffMyeah, you have a lot to manage
17:08.41JeffMfor one person
17:10.48blast007main networky stuff for 2.6 would be making the built-in map sending code fast enough that we don't need HTTP map loading
17:11.10JeffMa perfect use for channels :)
17:11.23JeffMbut yeah, that could be fixed pretty easy
17:11.42blast007cuz I want to remove the HTTP server as well and switch back to a single proto port
17:11.54JeffMsure, it's time has passed
18:52.14spldartAntique question. If I'm on a server with 6 player and 4 observer and I'm getting hundreds of fps but I'm on a server with 12 + player and a couple observer and all of a sudden it SEEMS like my framerate has gone to poop that's likely my bandwidth. Right?
18:53.47spldart*meh* should have done /lagstats on both too.
18:54.14Zehrathis seems to happen to me too =P
18:54.53Zehraonly difference is i seem to have jitter
18:56.04spldartYeah.. My problem kinda looks like jitter. Server didn't warn me though. But I didn't stay long since game play sucked a little.
18:57.03Zehraiirc
18:57.22Zehrait's something with the amount of data sent and players
18:57.47spldartThat's what I was thinking. Cuz I don't have a great connection all the time.
18:58.14spldartMore players. More updates on their position, shot position, my position etc
18:58.57Zehrathat's what i remember hearing which affects things a lot
18:59.07spldart:/ maybe their server is in europe too.. I forgot to ask
18:59.20spldartoh well...
19:00.28Zehrai happen to pay far too much attention to things related to server details
19:00.45Zehrahas quick keys on /lagstats and /idlestats
19:01.11spldartYeah. Shoulda /lagstats my situation. Meh
19:02.20Zehrait also depends on type of maps too iirc
19:03.01Zehrawas it a simple one or more complex like LouMan's or ahs3's maps?
19:03.49Zehrasince it seems maps which are very large create some jitter/lag
19:04.13spldartGood question however. I have lots of experience in game play in Louman maps and I even designed some of my own maps using wings3d export to bzw for some of my dumb ideas. Never had a problem with them.
19:05.27spldartI did a map called War Towers and another that was a giant air hockey table.
19:05.54spldartPlayed alot on Loumans race course and his churchyard/cemetery map too.
19:06.10spldartYah know... This has to be bandwidth.
19:07.08spldartI wonder how bzflag gameplay compares to full hd streaming video in terms of bandwidth...
19:07.42Zehrait would be an interesting test
19:08.04Zehrabtw this does remind me of a few maps which i'd like to see again
19:08.43Zehraahs3 made a racing series of maps too and it would be nice to see them hosted again :)
19:09.45Zehradrawinfo seems to make things run nicely too
19:09.48spldart*strokes beard and considers a hidden server at work*
19:10.20Zehrathinks about asking ahs3 for the racing maps to be hosted again
19:18.10Zehrai'm thinking of a few ideas for some maps too, but they'd require a few plug-ins and configurations to work
19:23.54JeffMDo they have the same map?
19:24.12JeffMTank count does not really affect fps
19:24.54spldartSo... More likely my connection to that particular server then
19:25.46JeffMThat won't affect client fps
19:25.52JeffMNot by much
19:26.28JeffMLow fps can add lag, lag does not lower fps
19:27.52JeffMThe number of polygons in the map is the largest factor in fps
19:30.18JeffMthe game does not "wait" for a packet before it draws a frame
19:30.29JeffMbut it does wait for it to draw a frame before it sends/receives a packet
19:30.48JeffMso there is no way to blame low FPS on any aspect of networking
19:31.13JeffMnetworking can make people seem like they jump around, because of low packet rates, but your FPS shoudl be constant.
19:31.26spldartSo I need to jump into a non busy map with tons of polygons and bounce around a bit after pressing the 'T' key.
19:31.43JeffMdid it seem like your motion was laggy? or just everyone else?
19:31.55JeffMyour local motion is based off your FPS
19:32.04JeffMother player motion is based off bandwith
19:32.09spldartI did recently install deb 9 some months ago. Perhaps that's an issue
19:32.28JeffMcheck your driver
19:32.36spldartmy rendered motion seemed to become an issue when things got real busy on the server
19:32.42spldartdriver... Just thinking that.
19:32.53JeffMwhat card do you have?
19:33.56spldartCuz I'm default whatever deb 9 autoplaced. But I have all build dependencies also taken care of so I know 3d is at least going... gimme a sec on the card.. It's old but it worked fine for bz in the past. Did ok with Unreal Torney too ;)
19:34.38JeffMa game from 2004?
19:35.11spldartMy point being.... BZFlag hasn't changed that much
19:35.35spldartGeforce 9800 gt
19:35.48JeffMthat should be a fine card
19:36.01spldartHad no problems in deb 7. Skipped 8 on this machine. Running deb 9 unstable now
19:36.05JeffMI'd bet you got toggled back to the open source driver
19:36.15spldartyeah
19:36.37JeffMit may do blending on the CPU, and with a lot of shots that can slow it down
19:36.48spldartI'll hafta check my drivers
19:37.27spldartAh! Even more reason to look for my old 'lsmod' and compare to new 'lsmod'
22:15.20TD-Linuxnouveau should be pretty feature complete on that card, but it may be stuck on the lowest performance level.
22:18.00JeffMI've never had good luck with the open source drivers, and the closed source ones have always worked better
22:19.16JeffMit's cool that there is an open source one, but to be competitive, it needs to be just as good as the proprietary driver, not "close"
22:19.52TD-Linuxthey have gotten vastly better as of late. the intel and amd ones are actually competitive and sometimes beat the proprietary driver
22:20.01TD-Linuxnot so much for nouveau due to reclocking.
22:20.12JeffMwell it's good they are getting closer
22:20.19JeffMbut in the long run it just has to work :)
22:20.56TD-Linuxyeah unfortunately the situation for nvidia is getting worse. they now require signed fw and only provide it in their blob driver
22:23.35TD-Linuxat least said blob driver is pretty good, if you are willing to run it.
22:26.37JeffMwhatever works
22:26.42JeffMthe blob is easy to install
22:28.48*** join/#bzflag brad (~brad@2001:41d0:a:14be::bad)
22:28.48*** join/#bzflag brad (~brad@unaffiliated/brad)
22:29.37blast007TD-Linux: isn't that somewhat the same for AMD and Intel?
22:30.02blast007there's firmware for some of Intel's newish stuff (Skylake and up, I believe) for some parts of the core
22:30.22blast007unless you're just talking about the method of distribution
22:30.44blast007(AMD also requires a firmware blob to get 3D working)
22:31.01JeffMand your facebook account
22:31.43TD-Linuxblast007, nope because it's distributable in linux-firmware
22:31.54blast007non-free
22:33.34TD-Linuxblast007, yes. but at least it can be shipped and doesn't require peeling apart the blob.
22:33.52TD-Linuxfor nouveau, they had actually rewritten the firmware from scratch
22:33.57blast007does nvidia's license prohibit distribution?
22:35.39TD-Linuxit allows distribution, but only in whole and without modification.
22:36.48blast007fun

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