Time Nick Message 04:59 hmmmm guys, when there's a deprecated field warning in the lua api, do you actually /want/ a stack backtrace??? or do you just want to know the file/line number it's at? 05:01 VanessaE probably just file/line is enough 05:02 VanessaE it's easy enough to grep one's mod tree to find references to the function/line/whatever that died. 05:02 VanessaE as long as you at least tell what mod made the bad call of course 05:02 VanessaE (error from foo: deprecated blah in bar/baz.lua line 1234 05:16 hmmmm damn shadowninja is on a roll 05:16 hmmmm look at all those commits 05:19 hmmmm ShadowNinja: https://github.com/minetest/minetest/pull/3250 this is the only aspect of the logging PR I did not merge 05:19 hmmmm I made a PR for this specific item on your behalf 05:33 hmmmm looking for some feedback on https://github.com/minetest/minetest/pull/3052 - merge or not merge? (will discard the .gitignore change) 05:36 VanessaE definitely merge, though he's wrong about there being performance impacts in minetest's case. 05:39 VanessaE (which reminds me, I need to do the same with HDX and my mods again, soon) 05:40 hmmmm thoughts on https://github.com/minetest/minetest/pull/3232 ??? 05:40 hmmmm I have no idea why this is staying open, something about.. android? 05:41 hmmmm android CI is working right now 05:41 VanessaE idk about android, but that setting IS also used on the PC version 05:47 hmmmm but is the fix itself good? 05:49 VanessaE I think so, but there's a related change that needs made to line 332 (as amended) -- don't do, `if foo == nil ...` rather, `if not foo ...`. 05:50 hmmmm why exactly is that needed? just curious 05:50 VanessaE my suggestion? 05:50 hmmmm my lua-fu isn't that great :) 05:50 VanessaE it's just bad style, that's all 05:50 hmmmm oh 05:50 hmmmm is the nil check needed at all though? 05:50 VanessaE nope,. 05:51 VanessaE if you're doing a check like this one, better to just check for a 'true' result than false-or-nil 05:51 hmmmm errr.. 05:51 hmmmm https://github.com/minetest/minetest/pull/3232/files#diff-b61e3e14dbe70423022f4ee065fa3f94R332 05:51 hmmmm am I wrong in saying this is duplicated work? 05:52 VanessaE you're not wrong 05:52 hmmmm viewing_range_auto is always == nil by virtue of the previous if statement and the fact this is an elseif 05:52 hmmmm so remove that entire clause entirely 05:52 hmmmm just make it into an else 05:52 VanessaE exactly. 05:52 VanessaE I didn't even catch that 05:52 hmmmm is any of this logic "correct"? 05:52 hmmmm I honestly don't know how the viewing range is supposed to work 05:53 VanessaE it doesn't :) 05:53 VanessaE auto view range is a clusterfuck :) but from the standpoint of the lua, you're correcty. 05:53 VanessaE -y 05:53 VanessaE if then/else/end is also that should have been needed here. 05:53 VanessaE is all* 05:54 VanessaE who wrote this anyway? 05:54 * VanessaE checks blame 05:54 hmmmm I have no idea 05:54 hmmmm but, man, a lot of the builtin lua code is crummy 05:54 hmmmm :) 05:54 VanessaE kilbith did. 05:54 hmmmm that exact piece of code? 05:54 VanessaE https://github.com/kilbith/minetest/commit/caf0df8924d4a64feee40170e7235eb2202ddde5 05:54 VanessaE yes 05:54 hmmmm blaaaaah 05:54 hmmmm how did this not get caught on review 05:55 VanessaE he probably just wasn't thinking about the behavior of the if/elseif/else maybe 05:55 hmmmm i'm not terribly interested in who the author is; i'm much more interested in the designated reviewers that let this slip through 05:55 hmmmm that's where the "blame" ought to be placed, no? 05:55 VanessaE after a fashion, I suppose 05:56 VanessaE though I don't like to blame est31 for stuff, as he's pretty quick to fix things also 05:58 hmmmm I think a "broken windows" theory is applicable to code quality 05:58 hmmmm if you let the small, obvious stuff slip, how well are you reallllly reviewing the rest of it? 05:58 VanessaE good point 05:59 hmmmm it's also tempting to let the rest of the stuff go, if the surrounding code quality is crap because "meh it's all a bunch of junk anyway" 05:59 hmmmm "i'm sure they'll fix all of it at the same time later on" 05:59 VanessaE heh 05:59 VanessaE that, frankly, sounds like something I might say in a moment of frustration 06:00 * VanessaE slumps down into her seat just a little, trying to hide 06:00 VanessaE :) 06:00 hmmmm I feel sorta bad modifying peoples' commits without their knowledge 06:00 hmmmm but it's so much quicker 06:00 VanessaE nah 06:00 VanessaE modify, inform, push. 06:01 VanessaE the way I figure it, if they have a problem with the resultant code -- and it's not a problem with the functionality of the code -- then it's their problem, not yours. 06:01 VanessaE (style issues aside) 06:04 hmmmm dafuck 06:04 hmmmm I can't cherry pick kilbith's commit 06:05 VanessaE forgot to fetch? 06:08 hmmmm oh 06:08 hmmmm i accidentally copied the commit id from the link YOU pasted 06:08 VanessaE hah 06:08 VanessaE sorry :) 06:08 VanessaE derp? :) 06:13 hmmmm the viewing range thing still doesn't quite work 06:15 hmmmm chalking it up to formspec crappiness 06:18 hmmmm you know what, there's more wrong with this PR 06:18 hmmmm i'm going to leave it to kilbith to fix it 06:19 hmmmm although i could, quite trivially, i don't want to do other peoples' work 08:54 est31 hmmmm, it would have been good to get agreement on #3013 08:54 ShadowBot https://github.com/minetest/minetest/issues/3013 -- 08:54 est31 https://github.com/minetest/minetest/pull/3013 08:59 nrzkt +1 09:32 est31 still wondering whether hmmm really can design something better than formspecs. 09:33 est31 well, lets see 11:03 VanessaE sfan5: question -- why wasn't the index.mth method set up to use human-readable hashes? 11:04 sfan5 because index.mth is not supposed to be read by humans probably 11:04 VanessaE just wondered; binary isn't a problem, just a curiosity. 16:08 luizrpgluiz what they will do in the new version of minetest in six months, in version 0.4.14? 19:35 ShadowNinja hmmmm: Hah, I apparently implemented that twice. It's in my multi-socket PR too. 19:36 ShadowNinja hmmmm: Seems you're missing the face_position_cache change too though. 19:37 hmmmm oh, make a separate PR for that 19:37 hmmmm let some others comment on it first, that one I'm not toooo sure about to be honest... 19:43 ShadowNinja Done. 19:47 hmmmm to be honest, I'm not even sure if FacePositionCache is needed 19:48 hmmmm might be best to remove it entirely 20:05 ShadowNinja when there's a deprecated field warning in the lua api, do you want a stack backtrace? -- Yes actually. Consider something like register_door("foo", {deprecated_field=true}). The dors mod would be reported as offending in that case. 20:08 hmmmm okay, fair enough 20:09 hmmmm I left it in as a backtrace anyway 20:13 celeron55 somebody is emailing me about concerns about minecraft skin support 20:14 Calinou do LibreOffice developers get concerns about .doc and .docx support too? 20:14 celeron55 apparently something has been changed around there and new versions of the skins aren't supported or something 20:14 Calinou yes, that's right 20:14 Calinou 1.8 has a new skin format, allowing for separate arm/leg texturing, overlaying the whole body and not just head 20:14 celeron55 do we plan to support the new ones, or not? 20:14 Calinou I thought you were talking about legal stuff at first :P 20:15 celeron55 how is the skin format detected 20:15 Calinou the new one is 64×64, old one is 64×32 20:15 celeron55 ok, well, i'm leaving this to whoever has been involved with the player model 20:16 Calinou Jordach has, he's in #minetest 20:22 celeron55 i figure if this is to be supported, then there needs to be two player model files, one for each skin format, and the engine has to provide some way of detecting the size (or aspect ratio) of the image 21:49 est31 https://github.com/est31/minetest/commit/b600bc30a9fbd92800dabad14a38f28f0ccdd28e 21:50 est31 ^ pushing in 10 mins 21:50 est31 tested it works 22:46 hmmmm oops.