Time Nick Message 05:05 Zeno` Wayward_One, someone else was getting a 500 error for the public server list 05:29 Wayward_One Huh... Well, fwiw, i can access the server list just fine. Additionally, this issue has been going on since the 8th 05:49 Zeno` no I mean somebody's server when registering to the PSL 12:04 Zeno` what's blocking #4757? 12:04 ShadowBot https://github.com/minetest/minetest/issues/4757 -- Fix unexplained shader issue (glsl compiler bug??) by Rogier-5 12:05 Zeno` Rogier-5's change (using a constant) is the correct way 12:06 Zeno` and I disagree with paramat. Premature optimisation is the root of all evil 12:07 Zeno` Look, I'm the first person to push optimisations, I think you all know that 12:07 Zeno` but this is silly 12:07 Zeno` the bugfix is the more important of the two issues! 12:07 Zeno` by far 12:08 Zeno` there is no evidence that it even needs optimisation. I suggest it be merged 12:09 nore +1 12:13 Zeno` I'll merge in 15 minutes then if there are no objections 12:14 Zeno` every modern compiler will optimize that automagically anyway 12:15 Zeno` in this instance readability is the more important thing 12:16 Zeno` actually there is nothing to optimise. Because it's const float the compiler probably does it even with -O0 12:17 Zeno` and it certainly does it with -O1, O2 and O3 12:43 sfan5 Zeno`: you can't make any assumptions about glsl compilers 12:43 sfan5 if they weren't this stupid we wouldn't have this issue in the first place 12:44 sfan5 with glsl these changes can change "wtf i have 1fps" to "it run at 60 fps even on older devices" 12:45 Zeno` *sigh* 12:45 Zeno` I don't agree with optimising something when there is not evidence that there needs to be optimisation 12:46 Zeno` It's not considered good practice 12:47 sfan5 it isn't good practice for all other kinds of code 12:47 sfan5 but this does not necessarily apply to GLSL 12:47 Zeno` and yet, there is still no solid data that says that the current code is bad 12:48 Zeno` so until there is, then it should be merged 12:48 Zeno` and if evidence shows up that it's better to use magic values *then* they should be used 12:49 Zeno` this "oh it's gonna effect performance" is just conjecture 12:49 Zeno` and not a reason to :-1: something 12:50 Zeno` a few releases ago, how much time did I spend profiling stuff? 12:50 Zeno` maybe 100 hours 12:50 Zeno` maybe more 12:51 Zeno` I care about performance, a lot, but in this case there is nothing to care about 12:52 Zeno` sfan5, just to be clear... you're objecting to it being merged as-is? 12:52 Zeno` if so, then I won't merge. But I think it's silly 12:52 red-001 could someone review #4472 ? 12:52 ShadowBot https://github.com/minetest/minetest/issues/4472 -- Make serverlist searchable. by red-001 12:52 sfan5 dunno 12:53 sfan5 i'm just saying that doing what paramat suggested could be useful 12:53 Zeno` sfan5, "premature optimisation is the root of all evil" 12:53 Zeno` code can change and comments can be forgotten to be updated 12:54 sfan5 but this does not necessarily apply to GLSL 12:54 Zeno` but nobody has shown that it *does* apply 12:55 Zeno` and doing something just because it /might/ affect something is well within the definition of "premature optimisation" 12:56 Zeno` clear code wins every time unless there are solid reasons to do otherwise 12:56 Zeno` the const float is clear, readable, and maintainable 12:57 Zeno` And I'm kinda shocked we're debating this, heheh 12:57 Zeno` I won't merge because it seems there is still objection. But I think it should be 13:05 Zeno` Zeno`: you can't make any assumptions about glsl compilers 13:06 Zeno` and yet, you and paramat are making an assumption that it's slower? 14:37 red-001 why is https://github.com/minetest/minetest/blob/master/src/game.cpp#L1142 needed? 14:40 est31 red-001: its just some unification to have the same size for builtin formspecs 14:42 red-001 why does android not use diffrent settings? 14:42 red-001 why does android use diffrent settings? 14:43 est31 because android devices usually have smaller screens 14:44 sfan5 red-001: in that specific case, on pc it's no problem to fix the size of the menu to something 14:44 sfan5 but on android the screen sizes vary so much that it makes zero sense to set a fixed size 14:44 sfan5 also a fixed size would make the menu unusable on small screens 15:18 Zeno` ok that'sit 15:18 Zeno` as much as I like the guy, paramat's latest comment goes way too far 15:18 Zeno` https://github.com/minetest/minetest/pull/4757#issuecomment-260945387 15:19 Zeno` that attack upon me is not only misrepresentation of what happened, it's WRONG 15:19 Zeno` he should not modify his original messages 15:20 Zeno` because doing so seems to be a deliberate attempt to obfuscate things 15:20 Zeno` for the record it has TWO approvals 15:20 Zeno` my approval and nore's 15:22 Zeno` here is his comment, because he'll probably change it: "Zeno this is not trivial and has only one approval, you almost impulsively merged against the rules again :)" 16:55 Zeno` merging #4757 16:55 ShadowBot https://github.com/minetest/minetest/issues/4757 -- Fix unexplained shader issue (glsl compiler bug??) by Rogier-5 17:04 Zeno` merging simple stuff like this should not be so difficult 17:26 Zeno` video of paramat: https://youtu.be/Q93SysOr8WU?t=1273 17:28 Zeno` with all due respect of course 17:28 Zeno` but it's funny! 17:39 Zeno` paramat approving a merge: https://www.youtube.com/watch?v=Q93SysOr8WU 17:40 Zeno` ooops: https://youtu.be/Q93SysOr8WU?t=1946 17:44 est31 I want to merge #4780 17:44 ShadowBot https://github.com/minetest/minetest/issues/4780 -- Introduce builtin_shared and use it to fix #4778 by est31 17:44 est31 I've improved the commit message like paramat requested 17:45 est31 will do the merge in 10 minutes if nobody objects 17:45 Zeno` I object! 17:45 est31 :/ 17:45 Zeno` ok, I don't 17:45 est31 yeey 17:46 Zeno` is it necessary? 17:47 Zeno` also, can builtin_shared be a constant? 17:47 est31 its a bugfix 17:47 est31 constant? 17:47 Zeno` yes, for performance 17:48 Zeno` lol, I'm sorry 17:48 Zeno` the PR is fine :P 18:06 kaeza you are comitting global genocide 18:07 kaeza good 18:07 Krock heh, git commit "genocide" 18:15 sofar zenocide 18:19 sfan5 Krock: you forgot the -m 18:20 est31 git commit -m "MY HANDS ARE TYPING WORDS" 18:20 est31 git commit -m "HAAAAAANDS" 18:25 Zeno` zenocide? roflol 18:25 Wayward_One xD 18:30 Krock sfan5, sorry. I don't use the command line :3 18:30 sfan5 you should 18:31 Krock est31, xkcd knows how you should do it properly ;) 18:38 Zeno` command line is evil 18:39 Zeno` 'tis black magic voodoo 18:41 Zeno` you can set the consoles 18:41 Zeno` you can set the console's background colour to something other than black, but it's still black magic 18:43 Zeno` I think I've gone insane 18:43 Zeno` every comment I make I imagine as a chess move and update the view of the board in my mind 18:44 Zeno` I'm not really sure if that's normal 19:01 sofar sounds like sleep deprevation to me... I'd have nightmares as a child like that 19:01 * twoelk wonders who won Zeno´'s game of chess 19:01 sofar semi-lucid 19:03 SpawnBugs i've seen this "HAAAAAANDS" commit somewhere for real 19:05 twoelk I remember when playing early low pixel games with more jitter than fps my dreams where sort of - pixilated with low fps :-D 19:13 Krock SpawnBugs, it's a quite common nonsense commit message ^^ 19:13 SpawnBugs Krock: i think I've seen it in some minetest mod 19:14 Krock I've used it once too. No idea in what project :3 20:02 SpawnBugs Krock: could be you 20:03 kaeza would anybody find it useful to have a string.split_iter() or such function? 20:04 kaeza split() creates a table which is unneeded in 99% of cases 20:04 Krock really? I sometimes run through the array with the for-like loop 20:04 kaeza Krock, yes, that's the table I'm referring to 20:05 kaeza for i, part in ipairs(x:split(...) do ... end 20:05 Krock an iterrator is not really required IMO 20:05 kaeza vs: for part in x:split_iter(...) do ... end 20:05 Krock going to check out the logs tomorrow 20:07 kaeza it should be a trivial change in builtin 20:15 sfan5 kaeza: sounds useful 20:53 kaeza sfan5, http://codepad.org/gFUY08lF 20:53 kaeza also simplified it a bit in the process 20:53 sfan5 looks complicated 20:53 sfan5 post a pr if you want it reviewed also 20:55 kaeza it's just a simmplified form of the string.split currently in the codebase, modified to return an iterator function instead of a table 20:56 kaeza string.split now uses split_iter to build a table 21:05 SpawnBugs found it! /media/sf_Exchange/minetest/mods/bitchange : 20fcf0a - SmallJoker, 2 years ago : HAAAANDS 21:09 SpawnBugs /media/sf_Exchange/minetest/mods/names_per_ip: 61fd749 - SmallJoker, 2 years, 3 months ago : More haaaaaands 23:25 red-001 !seen troller 23:25 ShadowBot red-001: I haven't seen troller in #minetest-dev. 23:28 proller__ ?