Time Nick Message 11:05 kilbith https://github.com/minetest/minetest/pull/9517#issuecomment-983514479 11:05 kilbith I was dead sure they would cause a regression 11:06 kilbith everytime they propose a PR, it causes a regression 11:06 kilbith every. fucking. time. 11:07 kilbith Desour caused like 4 or 5 regressions in a row 11:07 kilbith they never extensively test their code 12:09 erlehmann kilbith_ to be fair, that thing got approved by rubenwardy and sfan5 without even testing instructions being provided. 12:09 erlehmann so they had to make up their own 12:09 erlehmann i wouldn't do this 12:10 erlehmann in mineclonia, the PR needs to have testing instructions. we regularly find bugs in *other* code via that rule. 12:10 erlehmann kilbith_ anyways, blaming ppl for being sloppy is not a good idea 12:11 erlehmann blame the process 12:12 erlehmann i find “Does it resolve any reported issue? Probably not.” very funny though 12:13 erlehmann kilbith_ btw the devs i can get along with least are the ones who think if the code looks ok it must work 13:46 kilbith I don't care of mineclonia, stop talking about it 13:55 erlehmann mineclonia doesn't matter. what matters is that simply insisting on the coder having a test plan improves quality. 13:55 erlehmann desour seems to be pretty good with regressions though :D 15:35 sfan5 man the client movement code really sucks 15:39 Lone_Wolf And how the server handles it :P 15:41 sfan5 the server is generally the more solid part of Minetest 15:41 sfan5 most of the awful mess happens on the client 15:51 rubenwardy I blame Irrlicht 15:54 Lone_Wolf Has savilli reported the segfault from modified client movement? 15:55 sfan5 doesn't ring any bells here 16:00 Lone_Wolf Oh, huh. I guess they're just going to make a PR then 16:00 rubenwardy they PM'd me for permission, but gave no details 16:01 erlehmann there is a griefer called Mr_Anderson who likes to kill players and crash things 16:02 erlehmann if you see that player around in a crash, keep the trace 16:02 erlehmann it probably won't be an accident 19:00 sfan5 v-rob: what solution do you propose for #11669 / how is that going? 19:00 ShadowBot https://github.com/minetest/minetest/issues/11669 -- formspecs aiming to be full screen are cut off 19:01 erlehmann sfan5, how can https://github.com/minetest/minetest/issues/11809 be a dupe of https://github.com/minetest/minetest/issues/11611 if it is possible to encounter it regardless of if https://github.com/minetest/minetest/pull/11767 is merged or not? 19:01 sfan5 it'd also be useful if someone could confirm that #11700 is a real issue 19:01 ShadowBot https://github.com/minetest/minetest/issues/11700 -- Clientside formspecs are broken 19:01 erlehmann sfan5, do i count if i try? 19:02 sfan5 sure 19:02 erlehmann ok wait pls 19:05 sfan5 erlehmann: whether non-ttf support is removed or added back, care should be taken about the error message in both cases 19:05 v-rob Yeah, padding[] is what I want for the fullscreen stuff; I have code that should theoretically work, but Visual Studio broke on me again (I've done most of my work for Minetest outside of Minetest's codebase for now, so it hasn't been a problem yet) and I kind of forgot about it. I think I'm going to try to get MinGW working (again) 19:05 sfan5 the first is already covered by 11611, keeping another issue open for a minor detail is not useful 19:05 erlehmann not true though 19:06 sfan5 compiling using VS works fine here fyi 19:06 Krock the CSM issue seems to be a false alert 19:06 erlehmann bc i complain specifically about the error message which will not be helpful if you convert the format to TSV 19:06 v-rob VS gives some weird error about cmd.exe failing to run something or other for me 19:06 sfan5 so you opened an issue for a hypothetical future scenario? 19:06 erlehmann no, i encountered it today when trying to upgrade from 5.4.1 19:07 erlehmann i am just saying it is not a duplicate 19:07 sfan5 11611 covers that 19:07 Krock will merge #11796 and #11642 in 15 minutes. the mod loading stuff is better done after 5.5.0 19:07 ShadowBot https://github.com/minetest/minetest/issues/11796 -- Improve error message if using "/help --" by Wuzzy2 19:07 ShadowBot https://github.com/minetest/minetest/issues/11642 -- Network: Delete copy constructor and use std::move instead by SmallJoker 19:08 erlehmann sfan5 the error mesage is trying to be helpful but not covering all cases and the pixel font thing just exposes that, that's all. 19:08 sfan5 v-rob: do you compile from the command line or within VS? 19:08 v-rob Within VS 19:08 Krock why would you care about non-issues? 19:08 sfan5 the little development I did was with np++ and running "cmake --build . --config Release" in git bash 19:08 MTDiscord if a core dev has some time could https://github.com/minetest/minetest/pull/8448 get a review? it already has one approval 19:08 v-rob Hang on a sec, I'll try to get the exact error message 19:09 sfan5 @Jonathon not a priority for me sorry but if not already done someone should mention concrete usecases in the PR 19:10 sfan5 also I couldn't confirm #11753 yet, possibly driver specific or whatever 19:10 Krock tools with specific modes would be one such case, where you'd need to change the item otherwise. 19:10 ShadowBot https://github.com/minetest/minetest/issues/11753 -- Images in formspecs use "smooth" linear instead of nearest neighbor filtering 19:13 Krock testing 11767.... 19:21 Krock merging (2) 19:22 Krock done 19:29 erlehmann sfan5 why is this spam? https://github.com/minetest/minetest/issues/11812 19:29 sfan5 who cares 19:29 sfan5 seriously 19:30 erlehmann then write that 19:30 erlehmann i do care, bc the build takes basically forever 19:30 sfan5 any time spent responding to useless issues is better spent on something else 19:30 MTDiscord wouldn't that actually be a cmake issue? 19:30 sfan5 (it's also not even a Minetest bug) 19:30 v-rob Oh carp, I found the issue--I didn't have Zstd, although VS's error message was less than helpful in figuring that out. 19:30 erlehmann Sublayer plank not sure it might be minetest misconfiguring cmake 19:31 v-rob OK, it compiles now, and I'll test and get a PR out later today 19:31 sfan5 v-rob: thanks 19:34 erlehmann sfan5, what bugs do you also not want me to report besides security issues, build system issues and wrong error messages? 19:34 erlehmann i have the feeling that “spam” is a broad category 19:34 erlehmann and no intention of spamming further 19:34 Krock actual game problems 19:35 Krock visual issues, wrong behaviour. performance issues. that sort of stuff. for such trivial stuff it would be much more efficient to collect them and propose a general cleanup PR sometime 19:36 erlehmann Krock if most of the stuff i reported was trivial, i doubt it would exist in minetest. 19:37 erlehmann Krock, this is trivial. how to continue with it? https://github.com/minetest/minetest/issues/11797 19:38 Krock did you find any other, similar issues in the meantime? 19:48 sfan5 rubenwardy: can CDB packages have spaces in their name? 19:48 sfan5 guess not 19:48 rubenwardy no 19:50 erlehmann sfan5 i don't know why you would think that “if freetype is disabled at runtime you get an error message that complains about missing or malformed font files” has anything to do with pixel font support. have you even tried to reproduce that thing? 19:50 sfan5 if you don't want your issue to be considered in the context of the xml font regression then don't mention it 19:51 erlehmann maybe i need to spell it out here: if you break pixel fonts you have to remove/ignore any possiblity to not have freetype, otherwise you can get to a state where you have no usable font format left 19:52 erlehmann i don't know what is unclear about that 19:52 erlehmann or what is spam about complaining that the error message is wrong 19:52 sfan5 freetype is currently in a non-working but not properly removed state 19:52 sfan5 an issue already exists to track that 19:53 erlehmann also, it needs to be considered in the context of pixel fonts because the error message will be wrong too (but in a different way) if you restore pixel font support with your PR 19:53 sfan5 s/freetype/xml fonts/ 19:53 Krock and for the rescue, there's still bmp fonts 19:53 erlehmann which by the way has 4 upvotes and 2 downvotes 19:53 erlehmann Krock, wtf? 19:53 erlehmann show me bmp fonts :D 19:54 sfan5 the irrlicht built-in font is bmp, no xml 19:54 sfan5 don't think Minetest allows loading such ones from disk tho 19:55 Krock erlehmann: https://postimg.cc/7GfGrtvG 19:55 Krock I thought it would still work 19:56 Krock goddamn it. the new irrlicht build completely destroyed my Minetest binary archive 19:57 Krock anyway. the idea was that you could specify "font_path = /path/to/font.png" and use that as pixel font 19:57 Krock and those could be generated using https://github.com/kahrl/irrtum if you'd want to import some TTF 19:58 erlehmann sfan5 this might be rude, but i am beginning to think that you do not actually read some of my issues but reflexively react to keywords if it is some topic. please stop doing that, it sucks. 19:59 sfan5 ok 20:01 erlehmann thx 20:06 sfan5 rubenwardy: how does the contentstore detect that packages are installed? I touched package.id and that no longer seems to work 20:07 sfan5 nvm found it I think 20:10 rubenwardy (id and release existing) 20:14 sfan5 rubenwardy: #11814 20:14 ShadowBot https://github.com/minetest/minetest/issues/11814 -- Fix URL escaping in content store by sfan5 20:18 erlehmann lib/irrlichtmt/source/Irrlicht/CMakeFiles/IRRVIDEOOBJ.dir/CImageWriterJPG.cpp.o 20:18 erlehmann how exactly can i write jpeg from minetest? 20:18 sfan5 screenshot I guess 20:19 erlehmann ah! 20:19 erlehmann thx 20:20 sfan5 " id is used to track packages locally, and should never be used in a URL as case info is lost " 20:20 sfan5 uh huh 20:20 sfan5 it works fine though and there's one place where the old code already does this 20:35 erlehmann sfan5 Krock can i assume that https://github.com/minetest/minetest/issues/11700 is enough or do you need me testing? i was occupied 20:35 erlehmann i mean https://github.com/minetest/minetest/issues/11700#issuecomment-983967585 20:35 erlehmann bc i realized i can't test it 20:36 erlehmann > The above (working) screenshot comes from my Linux pc, and it still works atm. The screenshot (not working) from the first post comes from a windows one. 20:36 erlehmann but my friend has windows 10, i can ask him 20:37 sfan5 having a windows only regression would be weird 20:37 sfan5 but it's okay anyway 20:38 MTDiscord speaking of windows, has someone made a issue for sp connections port not being available? 20:38 sfan5 no 20:38 sfan5 make one 20:41 MTDiscord when i have a chance i will then 20:42 sfan5 don't forget that issues not reported don't get fixed so it's better to do it sooner than later 21:03 erlehmann does anyone of you have experience with crosscompiling minetest? 21:03 erlehmann i.e. compile on one architecture for another one? 21:55 erlehmann closing issue #11808 as “duplicate of #11808”, i guess that's what happens in a pointer overflow 21:55 ShadowBot https://github.com/minetest/minetest/issues/11808 -- Pointer overflow in src/util/srp.cpp (Secure Remote Password 6a implementation) 21:55 ShadowBot https://github.com/minetest/minetest/issues/11808 -- Pointer overflow in src/util/srp.cpp (Secure Remote Password 6a implementation) 21:56 erlehmann sfan5 what's the actual duplicate? or did you just have two tabs open 21:56 MTDiscord lol 21:57 sfan5 either that or github autocomplete helpfully changed the number after I finished typing 21:58 erlehmann not the first time someone accidentally duped something in minetest ;) 22:10 erlehmann sfan5, how are you so sure that https://github.com/minetest/minetest/issues/11808 is a dupe of https://github.com/minetest/minetest/issues/10238 – same root causes? it bugs out differently 22:11 sfan5 warning on the same line is good enough to me 23:46 erlehmann after having compiled the newest minetest, i must say: congratulations, despite what everyone keeps saying about render performance, i can still get 25 to 45 fps on a thinkpad that's older than some of the regulars on the server i play on 23:46 erlehmann and the water render bug finally seems to be fixed 23:46 erlehmann thanks to whoever worked that out 23:46 erlehmann it was there like forever 23:47 erlehmann (i might be wrong)