Time Nick Message 01:13 MTDiscord Maybe more coders need to be inspired to do competitive programming. That's where I learned git. 11:34 MTDiscord We would be more inspired if the challenges were more fun, they're always super technical 11:39 erlehmann programming is not fun 11:40 erlehmann getting results is 12:06 MTDiscord programming is fun 12:23 MTDiscord programming is fun 12:23 MTDiscord otherwise there would be EVEN LESS software developers haha 12:42 celeron55_ about irrlichtmt as a submodule: what actually are the arguments against it? clearly it would allow keeping track of compatible commits in an automated way and also clearly it doesn't pollute the the minetest repo with "unstripped" parts of irrlicht as it's basically a mapping of git refs 12:42 celeron55_ and also clearly you don't have to use it if you don't want to, you can use the current workflow if you like 12:44 celeron55_ github doesn't fully support submodules (eg. the zip downloads), but that part would just stay unchanged from now 12:46 sfan5 one is that you'd have to keep making commits to update the revision 12:47 sfan5 but if there's an alternative for the user to choose (and you'd only bump on every tagged release) that's not an issue 12:52 MTDiscord Well aren't we merging the two code bases in a few months anyways? 12:57 MTDiscord hello, can someone tell me why is BS defined as 10.0f? i don't get the comment: https://github.com/minetest/minetest/blob/master/src/constants.h#L74 13:01 sfan5 fundamentally it could be any value like 5.0 or 20.0 13:02 MTDiscord well then, perhaps a better question would be why BS exists at all? 13:03 MTDiscord i'm not clear what purpose it has other than being a magic number 13:06 sfan5 like the comment says it was supposed make people remember to not cast between number types but use a special helpers 13:07 sfan5 it has come with a number of kludges however so in hindsight BS can be called a mistake 13:08 erlehmann “but if there's an alternative for the user to choose” → sfan5 if you do not use “git submodule init”, all will be like before (except additional commits, but if your commit does not cleanly compile without updating the dependency, that's on you and you should amend it) 13:08 sfan5 erlehmann: I know 13:09 sfan5 just that "either you use the submodule or the project won't build" would not be an alternative (this is not the case here) 13:09 MTDiscord i see. thanks for clarifying, i'm pretty new to C++ overall 13:24 erlehmann sfan5 how about “"either you use the submodule or you do it the old way and find out by yourself with which version of irrlichtmt this compiles” 13:25 erlehmann i mean the current state is pretty much the worst i have ever seen in any project 13:25 erlehmann (that maintains its own library) 13:25 erlehmann there is literally no way to figure out what might compile 13:26 erlehmann and that's why there are the complaints 13:26 sfan5 factually incorrect 13:26 erlehmann “worst i have ever seen”, have i seen worse? 13:26 erlehmann btw, i'd love to hear from someone who is actually against submodules. maybe those ppl can speak out. 13:27 erlehmann bc i think we all know second-hand what arguments exist 13:27 erlehmann if no one is making them, the decision may be easier 13:27 sfan5 a given commit works with the most recent irrmt tag (at that time) and the most recent irrmt commit (at that time) 13:27 sfan5 not convenient or easy but not "literally no way" 13:28 erlehmann it may compile, but there are commits that don't work correctly depending on the version of irrmt 13:29 erlehmann sfan5 i retract my statement and replace it with “no way other than brute-forcing it” 13:29 erlehmann (starting with commits around the same time) 13:29 sfan5 that is also factually incorrect 13:31 sfan5 CI literally builds every commit with an irrlichtmt version defined in its scripts 13:32 sfan5 which is as close to 'officially supported combination' you can get 13:35 sfan5 anyway when it was first proposed I was against submodules because then the end result would have been that checking out the submodule is the only option to get it to build 13:40 MTDiscord I'm for submodules in general, though it does make dev slightly more error prone 13:41 MTDiscord But I'm not a core dev, so shrug 13:42 erlehmann does there exist an issue for submodules? 13:44 sfan5 don't think so 13:44 sfan5 or actually, yes but it'll be closed 13:45 sfan5 #11264 sort of 13:45 ShadowBot https://github.com/minetest/minetest/issues/11264 -- Add IrrlichtMt to minetest repo 13:48 erlehmann actually merging irrlichtmt would be the worst you could do (due to the -200k lines commit) 13:49 sfan5 It will be committed without history when it's ready 13:50 erlehmann breaking git blame 13:51 erlehmann well i hope there will be a PR 13:51 erlehmann so about the thing with the white/black textures, anyone willing to look into it? 13:51 erlehmann should i link the actual files here? https://github.com/minetest/minetest/issues/11146#issuecomment-962294459 13:52 erlehmann about the same model of a boat – one works, one trigger the bug 14:02 MTDiscord Nice, it will be brought in, that's good news. We could leave up irrlichtMT for git blame reasons 16:29 MTDiscord There are way too many confirmed unconfirmed bugs: https://github.com/minetest/minetest/issues?q=is%3Aissue+is%3Aopen+label%3A%22Unconfirmed+bug%22 16:30 erlehmann luatic schrödinger's bug 16:31 MTDiscord Take https://github.com/minetest/minetest/issues/10383, the earliest "Unconfirmed bug", for instance 16:31 MTDiscord It's a bug everybody who ever worked with semitransparent particles can confirm 16:31 MTDiscord There even exists a bugfix PR for it, yet it's still tagged as "unconfirmed" 16:33 sfan5 the better question is why the reporter (who is a coredev) didn't tag it as bug when they even provided clear proof of it being a bug 16:34 MTDiscord Good question. Either way, way too many confirmed bugs are still mistagged as "unconfirmed". 16:34 MTDiscord Probably the majority of them. 16:42 rubenwardy Core devs are still suseptible to silly mistakes 16:42 rubenwardy that's an obvious actual bug though 17:04 erlehmann i think stuff is tagged unconfirmed automatically? 17:05 rubenwardy yeah, it's that by default 17:20 erlehmann can i tag it confirmed by default? 17:20 erlehmann hehe 17:20 erlehmann can anyone tag issues? 17:21 MTDiscord no