| Juest | lol wut | 01:33 |
|---|---|---|
| Juest | using visual studio remote debugger is a network hazard | 01:34 |
| Juest | so instead i recommend using windbg | 01:34 |
| Juest | add a . to the symbol path env var to allow/make sure that pdb files in the same folder as the binary are read | 01:35 |
| blast007 | Dr. Memory didn't work there either | 01:35 |
| Juest | weird | 01:35 |
| Juest | i have hyper-v enabled and it works fine | 01:35 |
| Juest | so no idea whats up | 01:35 |
| Juest | controlled folders maybe? | 01:35 |
| blast007 | dunno what you mean | 01:36 |
| Juest | nvm, its a default in win 11 | 01:36 |
| Juest | but it exists in win10 too iirc | 01:36 |
| Juest | uhmmmmmmmmm | 01:36 |
| Juest | do you have windows sdk debuggers installed? | 01:36 |
| Juest | that should give you the needed dbghelp with support for downloading symbols online | 01:37 |
| blast007 | I have Visual C++ installed | 01:37 |
| Juest | windows sdk debugger is simpler in the sense it doesnt load everything all at once but only as its needed | 01:37 |
| Juest | its only for native code and drivers | 01:38 |
| Juest | anyway | 01:38 |
| Juest | i've been using windbg | 01:38 |
| blast007 | I don't need a debugger though, I need something that will find memory issues | 01:38 |
| Juest | oh okay right | 01:38 |
| Juest | idk why it works on my other computer but not on yours | 01:38 |
| Juest | my* | 01:38 |
| Juest | meh | 01:38 |
| *** Zehra <Zehra!~Yukari@user/yukari> has joined #bzflag | 01:54 | |
| Zehra | On the subject of scriptable languages for plug-ins. I think either some form of a functional language may perhaps work best. It would most closely match the "event" system which is used now. | 01:57 |
| Zehra | Adding to it, it would make perhaps likely sense for some form of "scriptable" world generating functions to be included, but not a high or important priority. (There's a small number of script generated maps, hence not a major consideration/issue.) | 02:00 |
| Juest | Zehra: angel script because why not then | 02:13 |
| Zehra | Juest: Taking a brief look, angel script looks almost the same to C++, so unless it doesn't need compiling, we're basically in the same place. | 02:17 |
| Juest | there are scriptable systems that are c++ism based but arent c++ | 02:17 |
| Zehra | what advantages you see to using something similar to c++? | 02:20 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has joined #bzflag | 02:20 | |
| Zehra | Because I'd see it easy from "if someone knows C++, they can easily use it". | 02:22 |
| Zehra | But by the same token, a lot of "ideas"/"concepts" aren't easy to express well/cleanly. | 02:23 |
| Zehra | I can't say it is entirely the fault of the language, the API could be better, or better tools could exist. | 02:24 |
| GEP | Juest: per client crash upon joining, and this latest commit https://github.com/BZFlag-Dev/bzflag/commit/63320f915abab8332187f95408b5747630506c79 what if you change that variable "world" to "_world"? | 02:25 |
| Zehra | To give the simplest example: It's not "easy" implementing flags which would apply/use custom effects. I'm aware it's just a "grievance" for me, because I just use a template now days. (I've worked out "most" of the bugs, except for one "possible" one.) | 02:31 |
| Zehra | It did cause confusion and bugs for some. But with quick sharing of information, it was resolved quickly. | 02:35 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has quit IRC (Ping timeout: 240 seconds) | 02:37 | |
| Zehra | In short: The principle was "grabbing" a flag toggles on an effect, and having the flag drop, it will deactivate it. Simple enough, but not intuitive enough. | 02:38 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has joined #bzflag | 02:44 | |
| GEP | never mind, what I will say about the bug as it stands.. while I can reproduce it, it seems a lot harder than on 2.4.26 | 02:47 |
| *** Tobbi is now away: Auto away at Fri Dec 13 02:48:18 2024 UTC | 02:48 | |
| Zehra | I can't be sure, but I nearly am sure I located that bug quite a while ago or something | 02:49 |
| Zehra | or a different one | 02:49 |
| *** Tobbi is back | 03:08 | |
| Agatha | C++ is great; the main pain point for plugins is just all the hassle of setting up builds and loading/unloading them. An in-game editor in COBOL would be easier to use than a restart-the-game-every-time plugin system in Go (or whatever the cool kids are lusting after nowadays). | 03:35 |
| Zehra | ^This. Especially with custom flags and similar. | 03:37 |
| *** Tobbi is now away: Auto away at Fri Dec 13 03:59:52 2024 UTC | 03:59 | |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has quit IRC (Ping timeout: 240 seconds) | 04:18 | |
| Zehra | I'm taking a wild guess, joinInternetGame(), joinGame()... Check "joinRequested" boolean, something gives me impression there. | 04:22 |
| BZNotify | bzflag: RighthandSon commented on issue #128 "Client crash when hitting Connect multiple times" by Juesto (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2540533709): at the head 2.4 I was able to reproduce and get a similar back trac... | 04:34 |
| Zehra | Do we know if it fails between joinInternetGame and joinInternetGame2? | 04:39 |
| Zehra | I'm guessing what happens is a loop of some sorts: | 04:48 |
| Zehra | JoinGame() -> JoinRequested=, JoinInternetGame() -> (Theory: JoinGame gets called again) -> JoinInternetGame2 -> JoinGame() -> JoinRequested=, JoinInternetGame() -> (joinGame might be called again) -> JoinInternetGame2 | 04:51 |
| Zehra | GEP's comment might indicate that the player is "interacting" with a world, while it's being modified. (Less likely of being run when the world is modified, reducing the bugs "probability".) | 04:53 |
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-98-219-34-189.hsd1.ga.comcast.net> has quit IRC (Ping timeout: 248 seconds) | 05:14 | |
| *** Zehra <Zehra!~Yukari@user/yukari> has quit IRC (Quit: Leaving) | 05:38 | |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has quit IRC (Read error: Connection reset by peer) | 07:20 | |
| *** Tobbi is back | 07:52 | |
| BZNotify | 2.4 @ bzflag: atupone pushed 1 commit (https://github.com/BZFlag-Dev/bzflag/compare/63320f915aba...f500b9400bd0): | 08:16 |
| BZNotify | 2.4 @ bzflag: atupone f500b9: playing.cxx: no world -> no tank (https://github.com/BZFlag-Dev/bzflag/commit/f500b9400bd03e91af3d70963b0f14be3e029829) | 08:16 |
| tupone | fixed last stack trace, but I'm expecting more | 08:16 |
| tupone | world seems null in this case | 08:17 |
| blast007 | tupone: I got it to crash on linux as well. I commented the two allow/deny lines in ~/.bzf/DownloadAccess.txt and joined the Planet Mofo server and hit connect again as soon as it started downloading textures. | 09:17 |
| blast007 | and to repeat the test, I delete ~/.bzf/cache/http/images.planetmofo.com/ | 09:18 |
| blast007 | here was the valgrind output: https://gist.github.com/blast007/5d72faf24b382a9fd39e4e8b78a65d6a | 09:20 |
| blast007 | sorry, I meant ~/.bzf/cache/http/images.planet-mofo.com/ (missed the hyphen) | 09:25 |
| tupone | can you do with sanitizers ? | 09:34 |
| tupone | if not, I will redo the step when at home | 09:37 |
| blast007 | k, building it again | 09:42 |
| tupone | -g -O1 or -O0 | 09:44 |
| blast007 | ==637850==ASan runtime does not come first in initial library list; you should either link runtime to your application or manually preload it with LD_PRELOAD. | 09:45 |
| blast007 | should I *not* run it in valgrind? | 09:46 |
| blast007 | sounds like valgrind doesn't support asan | 09:46 |
| blast007 | I configured with --enable-debug, so it would do that -g -O0 right? | 09:48 |
| blast007 | (seems so) | 09:49 |
| blast007 | https://gist.github.com/blast007/5d72faf24b382a9fd39e4e8b78a65d6a#file-asan-ubsan-txt | 09:49 |
| tupone | no need for valgrind | 10:12 |
| blast007 | I think the issue is that texture downloads don't get aborted when you "cancel" the join (and I also don't think we cancel the join completely even without downloads) | 10:30 |
| blast007 | so textures still finish downloading and then call joinInternetGame2() | 10:30 |
| tupone | But can you run without valgrind but with asan and ubsan? | 10:32 |
| *** Lantizia_ <Lantizia_!~Lantizia@user/lantizia> has quit IRC (Ping timeout: 252 seconds) | 10:33 | |
| tupone | The problem that I found, when it crashed at me was related only to the world | 10:33 |
| tupone | But there are other problem that I was not finding, maybe the texture download | 10:34 |
| *** FastLizard4 is back | 10:57 | |
| blast007 | that was without valgrind. | 11:10 |
| blast007 | the asan-ubsan.txt is the output I get when I crash it without valgrind | 11:10 |
| tupone | ok. so both asan & ubsan. that trace is easy to fix (seems). | 11:18 |
| *** FastLizard4 is now away: AWAY from keyboard | 11:18 | |
| tupone | just to check the world at those locations | 11:19 |
| tupone | probably there is a more elegant solution like deleting myTank, but I cannot find it | 11:21 |
| *** Lantizia <Lantizia!~Lantizia@user/lantizia> has joined #bzflag | 11:31 | |
| blast007 | It's probably not worth the time/effort to fix this right now. A more major refactor should happen for our join code at some point. | 11:37 |
| *** Tobbi is now away: Auto away at Fri Dec 13 11:52:09 2024 UTC | 11:52 | |
| *** FastLizard4 is back | 11:55 | |
| *** Tobbi is back | 12:15 | |
| *** I_Died_Once <I_Died_Once!~I_Died_On@c-98-219-34-189.hsd1.ga.comcast.net> has joined #bzflag | 12:20 | |
| *** FastLizard4 is now away: AWAY from keyboard | 12:23 | |
| *** FastLizard4 is now away: GONE - Screen Detached and Disconnected from IRC (I'm probably asleep, at work, or doing something in real life) | 12:31 | |
| Juest | ooooooooooooooooooooooo | 12:40 |
| Juest | blast007: so it enters a weird state related with textures and kinda why it seems like i cant repro without textures to download? | 12:41 |
| Juest | lol gep, tell that to tupone | 12:41 |
| BZNotify | master @ bzfls3: blast007 pushed 1 commit (https://github.com/BZFlag-Dev/bzfls3/compare/e0ee7931967b...9051ccafabab): | 12:42 |
| BZNotify | master @ bzfls3: blast007 9051cc: Add error page for 404 and 405. (https://github.com/BZFlag-Dev/bzfls3/commit/9051ccafabab87fe4ddb37b50d60170ec54de343) | 12:42 |
| Juest | blast007: *nothing* gets aborted when you retry a connect, i think | 12:43 |
| Juest | the texture download does claim to get aborted lol | 12:44 |
| blast007 | hitting 'Connect' calls joinGame() and joingame does handle cancelling the *world* download, but not texture downloads. https://github.com/BZFlag-Dev/bzflag/blob/2.4/src/bzflag/playing.cxx#L518-L536 | 12:45 |
| Juest | okay, so in short: lots of issues with joingame() abortion? | 12:45 |
| blast007 | the joinRequested = true at the bottom does mean that it will instantly start connecting again instead of *just* cancelling the world download | 12:46 |
| Juest | at least this is good progress, right? | 12:46 |
| blast007 | really we just need to clean up the entire join process | 12:46 |
| Juest | can we do it for 2.4 or it has to come in 2.6? | 12:47 |
| blast007 | what I want would require breaking changes | 12:47 |
| Juest | can you make it good enough without breaking changes or...? | 12:47 |
| Juest | are the breaking changes a hard dependency/requirement? | 12:48 |
| blast007 | this is one of those rabbit hole bugs where the more you dig the deeper you get | 12:48 |
| Juest | Oh right | 12:48 |
| Juest | seems so, yeah, i did notice | 12:48 |
| Juest | hmmmm | 12:49 |
| Juest | you think the stacktraces are vaporware and the problem lies in joinGame() entirely? | 12:49 |
| blast007 | I did also get Dr. Memory to work. I used 2.5.0 instead of 2.6.0, but the output I got didn't make any sense, so I gave up with that program :P | 12:49 |
| Juest | oh yeah weird | 12:50 |
| blast007 | no, the problem isn't in joinGame at all, really | 12:50 |
| blast007 | I was just showing that we *do* cancel part of the join | 12:50 |
| Juest | well, for me it worked but bzflag crashed | 12:50 |
| Juest | on my other computer dr. memory 2.6.0 works but the computer is sooo slow | 12:50 |
| blast007 | we could just remove the ability to cancel joins :) | 12:50 |
| Juest | sure that'd be good actually, disable the connect button | 12:51 |
| Juest | OR add a cancel button that is not in the connect button | 12:51 |
| Juest | that bypasses the bug entirely | 12:51 |
| blast007 | if we could make a cancel button, we'd not have this problem | 12:51 |
| Juest | yep | 12:51 |
| Juest | it should've been done long ago | 12:51 |
| blast007 | (meaning, if we could cancel, we wouldn't need a cancel button - the issue isn't the lack of a button) | 12:51 |
| Juest | or i guess we could disable the connect button in first place | 12:52 |
| blast007 | just always disable the connect button, then we can never join and thus never trigger any problem | 12:52 |
| Juest | i think that'll be healthier to avoid the treadmill rabbit hole | 12:52 |
| blast007 | :) | 12:52 |
| Juest | oh come on blast007, that's too much | 12:52 |
| blast007 | lol | 12:52 |
| blast007 | honestly, just preventing cancelling during *texture* downloads might be good enough | 12:53 |
| Juest | i get its a joke but realistically.... it does feel like too much :P | 12:53 |
| blast007 | texture downloads are pretty fast, unlike the world download from bzfs sometimes | 12:53 |
| Juest | hm, is the "mesh" download broken now or something? | 12:53 |
| Juest | the progress is (0/0) and then goes to world downloading | 12:54 |
| Juest | yeah cuz its http(s) | 12:54 |
| blast007 | dunno what you mean by mesh download. world download? | 12:54 |
| Juest | the first stage before the world download | 12:54 |
| Juest | you know there are like maps that bypass bzw commands or something | 12:54 |
| Juest | for more complex structures | 12:54 |
| blast007 | no, I have no idea what you mean | 12:55 |
| Juest | sorry | 12:55 |
| blast007 | the world download can be over HTTP or over BZFlag protocol | 12:55 |
| blast007 | the progress (0/0) might be HTTP | 12:55 |
| Juest | im just talking about the ability to make complex geometry without geometry bzw | 12:55 |
| Juest | OH | 12:55 |
| Juest | gothca | 12:55 |
| Juest | gotcha | 12:55 |
| Juest | that can be cancelled too actually | 12:55 |
| blast007 | either way, the client never actually gets BZW | 12:55 |
| Juest | so the issue is with http specifically | 12:55 |
| blast007 | it gets a binary representation of the world data | 12:55 |
| Juest | and save world converts that to a bzw however it can? | 12:56 |
| blast007 | yeah | 12:56 |
| blast007 | it's like taking a photo of a printout :) | 12:56 |
| Juest | oh i see | 12:56 |
| Juest | but its more like a decompile instead | 12:56 |
| Juest | feels like so considering its missing stuff for instance | 12:56 |
| blast007 | it's a lossy process because the server doesn't send *all* of the map content | 12:56 |
| Juest | stuff like that | 12:56 |
| Juest | hence why it feels more like a decompile :) | 12:57 |
| Juest | is it not possible to actually request the bzw from the server? | 12:57 |
| blast007 | nope, not possible with the current code | 12:57 |
| Juest | okay | 12:58 |
| Juest | hmm | 12:58 |
| Juest | when you hit host server | 12:58 |
| Juest | it should automatically connect as part of the start server button | 12:58 |
| Juest | imo | 12:58 |
| blast007 | that could do a lot better in general :) | 12:58 |
| Juest | how so? | 12:58 |
| Juest | okay | 12:58 |
| Juest | let's sort this through: | 12:58 |
| blast007 | it auto fills in the hostname, but that doesn't work on all platforms | 12:58 |
| blast007 | anyway, gotta go to work | 12:59 |
| Juest | oh okay | 12:59 |
| Juest | take care blast007 ttyl | 12:59 |
| Juest | 127.0.0.1 and ::1 does exist in all platforms | 12:59 |
| Juest | also localhost | 12:59 |
| Juest | unless dns is broken | 12:59 |
| Juest | blast007: can you place the pdbs for the dependencies and the game in the bin folder? :D | 13:27 |
| Juest | i manually have done that | 13:27 |
| Juest | for debugging purposes | 13:27 |
| Juest | repro'd | 13:43 |
| BZNotify | bzflag: Juesto commented on issue #128 "Client crash when hitting Connect multiple times" (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2541499718): we determined that reproduction is when aborting downloading files ... | 13:46 |
| blast007 | is there still an option to turn off downloads? if so, can you get it to crash with that disabled? | 13:46 |
| BZNotify | bzflag: Juesto commented on issue #128 "Client crash when hitting Connect multiple times" (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2541499718): we determined that reproduction is when aborting downloading files ... | 13:47 |
| Juest | hmmmmmmm | 13:53 |
| Juest | blast007: i think i got to the point its fixed and its just memory issues on my computer and probably others' end | 13:54 |
| Juest | can we build bzflag with sanitizer in vs 2017???? | 13:54 |
| Juest | also hi work blast007 :) | 13:54 |
| Juest | you seem to not be far from work computer? | 13:54 |
| Juest | do you vnc or use a irc bouncer? | 13:55 |
| Juest | :D | 13:55 |
| Juest | im having a freeze when the player finishes joining | 13:58 |
| Juest | its over windows remote desktop | 13:58 |
| Juest | idk how to debug this further | 14:01 |
| Juest | but yeah | 14:01 |
| Juest | that's pretty much it | 14:01 |
| Juest | im glad its over | 14:01 |
| BZNotify | bzflag: Juesto edited issue #128 "Client crash when hitting Connect multiple times at downloading files stage" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 14:03 |
| BZNotify | bzflag: Juesto edited issue #128 "Client crash when hitting Connect multiple times at downloading files stage" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 14:04 |
| Juest | edited OP ^ | 14:05 |
| BZNotify | bzflag: Juesto opened issue #369 "Unable to connect after aborting world download" (https://github.com/BZFlag-Dev/bzflag/issues/369) | 14:07 |
| BZNotify | bzflag: Juesto commented on issue #128 "Client crash when hitting Connect multiple times at downloading files stage" (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2541544337): I suggest removing the ability to click on connect when a connectio... | 14:09 |
| BZNotify | bzflag: Juesto closed issue #128 "Client crash when hitting Connect multiple times at downloading files stage" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 14:09 |
| blast007 | I SSH in to a system where irssi runs inside of a screen session | 14:10 |
| Juest | ah i see | 14:10 |
| Juest | ssh bouncing to a irssi on screen | 14:10 |
| blast007 | :) | 14:10 |
| Juest | what do you think about the status of the issue? | 14:17 |
| Juest | it also tbf, doesn't make much sense | 14:17 |
| Juest | like, what did we really fix with the stacktraces? | 14:17 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has joined #bzflag | 14:40 | |
| Juest | lol gep, tell that comment to tupone :P | 14:53 |
| Juest | hi | 14:53 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has quit IRC (Ping timeout: 240 seconds) | 14:56 | |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has joined #bzflag | 15:17 | |
| GEP | The is a reason I don't often join irc :) | 15:17 |
| GEP | blast007: saw your suggestion pertaining to turning off automatic downloads. I was able to get a couple different stack traces having to do with loadingCachedWorlds where is the best place to post them. | 15:25 |
| blast007 | I haven't looked at why github issue #128 was closed, so I don't know if that's actually resolved. But that was/is the one about the client crashing when hitting connect multiple times. | 15:28 |
| blast007 | 2.4.28 is going to release without a fix for this, but I can see about something after that. | 15:30 |
| GEP | yea it seems like way too deep a problem | 15:33 |
| BZNotify | bzflag: RighthandSon commented on issue #128 "Client crash when hitting Connect multiple times at downloading files stage" by Juesto (https://github.com/BZFlag-Dev/bzflag/issues/128#issuecomment-2541725747): A couple more back traces.... | 15:40 |
| blast007 | Our join process is rather convoluted. | 15:41 |
| BZNotify | bzflag.org: RighthandSon synchronized pull request #54 "Add documentation of basic client commands" (https://github.com/BZFlag-Dev/bzflag.org/pull/54) | 15:50 |
| *** Tobbi is now away: Auto away at Fri Dec 13 16:02:16 2024 UTC | 16:02 | |
| *** Sgeo <Sgeo!~Sgeo@user/sgeo> has joined #bzflag | 16:16 | |
| Juest | blast007: oh, i closed it because i think we're done with it, but im probably wrong | 16:18 |
| Juest | whoops | 16:18 |
| BZNotify | bzflag: Juesto reopened issue #128 "Client crash when hitting Connect multiple times at downloading files stage" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 16:18 |
| Juest | GEP: is the reproduction scenario the same as issue #128? | 16:19 |
| blast007 | Juest: a cancel button isn't a fix. the problem is that we don't have a way TO correctly cancel a join. | 16:19 |
| Juest | OH OK | 16:19 |
| GEP | Pretty much | 16:19 |
| Juest | well post it i guess | 16:19 |
| BZNotify | bzflag: Juesto edited issue #128 "Client crash when hitting Connect multiple times" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 16:20 |
| *** Tobbi is back | 16:20 | |
| BZNotify | bzflag: Juesto edited issue #128 "Client crash when hitting Connect multiple times" (https://github.com/BZFlag-Dev/bzflag/issues/128) | 16:20 |
| Juest | updated | 16:20 |
| GEP | spamming connect in the later stages of connection to a server | 16:20 |
| Juest | yes | 16:21 |
| BZNotify | bzflag.org: RighthandSon synchronized pull request #54 "Add documentation of basic client commands" (https://github.com/BZFlag-Dev/bzflag.org/pull/54) | 16:39 |
| *** GEP <GEP!~GEP@c-76-21-169-178.hsd1.va.comcast.net> has quit IRC (Quit: Client closed) | 17:31 | |
| tupone | I put the network to 100Mb/s and I was able to kill it, 2.5G is too much to get the texture window | 17:55 |
| Juest | lol | 17:58 |
| Juest | fancy | 17:58 |
| tupone | a reminder: to get stack trace with ubsan: export UBSAN_OPTIONS="print_stacktrace=1:halt_on_error=1" | 18:01 |
| blast007 | so what you're saying is that my powerline ethernet connection that gets only around 30 to 50 Mbps is ideal for this test ;) | 18:03 |
| Juest | lmfao | 18:05 |
| *** FastLizard4 is back | 19:06 | |
| tupone | :) | 19:07 |
| *** FastLizard4 is now away: AWAY from keyboard | 19:24 | |
| BZNotify | 2.4 @ bzflag: atupone pushed 2 commits (https://github.com/BZFlag-Dev/bzflag/compare/f500b9400bd0...070d498b7a9f): | 19:39 |
| BZNotify | 2.4 @ bzflag: atupone d7d374: LocalPlayer: remove an unused getHitBuilding (https://github.com/BZFlag-Dev/bzflag/commit/d7d3742843d04bf8dab6006d7e40880ed2f8efac) | 19:39 |
| BZNotify | 2.4 @ bzflag: atupone 070d49: Fix crashes when world is null after connect spam (https://github.com/BZFlag-Dev/bzflag/commit/070d498b7a9f50926f213a8362503ddd296054d5) | 19:39 |
| tupone | hope it fixes all what is happening at connect spawm | 19:39 |
| tupone | put back net at 2.5G | 19:39 |
| Juest | tupone: wanna bet its fixed for my computer or im going to face the issue because the memory is fucked? | 20:59 |
| Juest | i have a windows 7 computer to test on too | 20:59 |
| Juest | but it has some issues | 20:59 |
| Juest | so meh | 20:59 |
| tupone | Juest: i think is not your memory. your stack trace was always the same | 21:02 |
| tupone | the fix is not perfect. But it does not crash anymore | 21:04 |
| Juest | oh okay | 21:13 |
| Juest | tupone: i did not try more than once, but i had varying stacktraces so i guess it was multiple issues manifesting "randomly" | 21:14 |
| Juest | gotcha then my memory is mostly fine, yes | 21:14 |
Generated by irclog2html.py 2.17.3 by Marius Gedminas - find it at https://mg.pov.lt/irclog2html/!