Time Nick Message 00:32 hmmmm yes! let's release it 00:33 Exio :D 01:11 * VanessaE waits with her build script idling...come on, I'm burning gas here :) 01:49 iqualfragile waiting for the next day might be the better idea, as sfan5, pilzadam and kahrl will come back online then 06:31 Zeg9 lol ">>> FREEZE until 0.4.7 is released <<<" 07:14 celeron55 Zeg9: don't breathe! 07:16 Zeg9 I thought it was going to be released last night in fact 08:28 PilzAdam Im fine with releasing 08:29 PilzAdam everyone who doesnt know: http://dev.minetest.net/Releasing_Minetest 08:29 PilzAdam thats the plan 08:31 * PilzAdam wants to be the one who commits the new version number this time :-) 08:33 iqualfragile Kray: minetest is lgpl not gpl 08:35 PilzAdam is a protocoll version bump needed? 08:36 PilzAdam -l 08:47 celeron55 iqualfragile: kray does not maintain the arch package anymore 08:47 iqualfragile oh, ok 08:47 celeron55 iqualfragile: he has heard that like 1000000000000000 times already 08:47 celeron55 and it isn't probably funny anymore 08:47 iqualfragile yay 08:48 Anchakor celeron55: he didn't disown the package 08:49 celeron55 apparently not the git one 10:45 Kray err 10:45 Kray the stable package made it to the community repo 10:45 Kray and when that happened, the AUR package I maintained was deleted 11:05 celeron55_ Kray: but there's still minetest-git 11:09 Kray yeah 12:28 kahrl good morning 12:29 kahrl <+PilzAdam> is a protocol version bump needed? <-- as far as I remember, no 12:30 kahrl there's the new TOCLIENT_HUD_SET_PARAM but it shouldn't require a version bump 13:03 Exio good morning kahrl 16:01 PilzAdam Zeg9 found a bug in https://github.com/minetest/minetest/commit/649dca4d16786ed4e5b74878f56b86dea4f02542 16:01 PilzAdam maybe it should be reverted before releasing 0.4.7 16:13 Zeg9 PilzAdam: apparently it doesnt appear in all mods 16:14 Zeg9 "apparently it doesnt appear" << good sentence 16:17 PilzAdam maybe the mod where the bug appears is just using a "wrong" model? 16:18 Zeg9 Well, the model works on older versions. also these models are by Pavel_S, who made those for simple mobs (and simple mobs work perfectly) 16:21 Zeg9 I will try mobf 16:28 Zeg9 PilzAdam: mobf, simple mobs, boats, ufos, carts tested: only helicopter and flying machine from Pavel_S don't work 16:38 kahrl maybe the code in question should set all texture layers of getMaterial(i) to the new texture? 16:55 kahrl nvm, doesn't fix the problem with heli.x 17:03 sfan5 any comments about #760 ? 17:04 sfan5 *on 17:05 VanessaE sfan5: no way in hell 17:05 VanessaE sneak elevators are too useful 17:05 sfan5 but is a bug 17:05 PilzAdam sfan5, yes: 409 17:06 sfan5 PilzAdam: mine has an accordingly correct comment 17:06 PilzAdam well, rather comment on the existing one than creating a new one 17:06 hmmmm ! 17:06 hmmmm 0.4.7 still isn't out 17:06 hmmmm what's taking you so long 17:06 PilzAdam there is a bug 17:06 hmmmm what is the bug 17:06 PilzAdam we need to decide what to do 17:07 Zeg9 8D 17:07 PilzAdam read logs, something with models 17:07 kahrl there's been some talk about reverting 649dca4d167 17:08 hmmmm what seems to be the problem..? 17:08 PilzAdam it breaks one mod 17:08 PilzAdam textures are wron in the helicopter model AFAIK 17:08 PilzAdam +g 17:09 Zeg9 it breaks flyingmachine too 17:09 hmmmm you guys haven't figured out what the problem is? 17:10 hmmmm what i don't get is why this was committed in the first place, it wasn't really fixing a critical bug in the first place 17:10 hmmmm a normal priority bugfix broke something else right before a release 17:10 hmmmm think about that and tell me what you think you should do 17:10 PilzAdam git revert FTW 17:11 hmmmm yeah i think so 17:11 kahrl correct textures: http://i.imgur.com/PNZRXsT.png, currently: http://i.imgur.com/kE6zCQy.png 17:12 PilzAdam so, reverting that commit now, release 0.4.7, then think about a proper fix? 17:12 kahrl PilzAdam: I agree with that 17:13 PilzAdam ok, Ill do it now 17:13 kahrl brb 17:16 hmmmm okay great 17:16 hmmmm now we're moving 17:16 hmmmm it was already ridiculous a week ago 17:18 PilzAdam ok, tags created, stable-0.4 branch updated 17:19 PilzAdam thexyz, can you build the windows version now? 17:22 PilzAdam Ill update launchpad 17:24 thexyz PilzAdam: http://buildbot.minetest.net/build/s4pfK6D39NfwsbDkGmJWe4 17:25 PilzAdam launchpad builds will be availible soon 17:26 VanessaE my linux build is already up :D 17:26 PilzAdam Ill write something in News section 17:28 thexyz ..\..\src\minetest\src\util\numeric.cpp(185): error C2065: 'M_PI' : undeclared identifier 17:30 hmmmm hmmmmm 17:31 hmmmm that's a new one... just #define M_PI 3.141592 and compile that, we'll figure out a real solution later 17:32 thexyz okay 17:34 PilzAdam launchpad builds up 17:39 thexyz content_sao.obj : error LNK2001: unresolved external symbol "public: void __thiscall ScriptApiEntity::luaentity_Punch(unsigned short,class ServerActiveObject *,float,struct ToolCapabilities const *,class irr::core::vector3d)" (?luaentity_Punch@ScriptApiEntity@@QAEXGPAVServerActiveObject@@MPBUToolCapabilities@@V?$vector3d@M@core@irr@@@Z) [C:\minetest\realtime\build\src\minetest.vcxproj] 17:39 thexyz s_entity.obj : error LNK2001: unresolved external symbol "void __cdecl push_tool_capabilities(struct lua_State *,class ToolCapabilities const &)" (?push_tool_capabilities@@YAXPAUlua_State@@ABVToolCapabilities@@@Z) [C:\minetest\realtime\build\src\minetest.vcxproj] 17:39 thexyz l_craft.obj : error LNK2001: unresolved external symbol "class std::vector > __cdecl read_items(struct lua_State *,int,class Server *)" (?read_items@@YA?AV?$vector@UItemStack@@V?$allocator@UItemStack@@@std@@@std@@PAUlua_State@@HPAVServer@@@Z) [C:\minetest\realtime\build\src\minetest.vcxproj] 17:39 thexyz l_craft.obj : error LNK2001: unresolved external symbol "void __cdecl push_items(struct lua_State *,class std::vector > const &)" (?push_items@@YAXPAUlua_State@@ABV?$vector@UItemStack@@V?$allocator@UItemStack@@@std@@@std@@@Z) [C:\minetest\realtime\build\src\minetest.vcxproj] 17:39 thexyz C:\minetest\realtime\src\minetest\bin\Release\minetest.exe : fatal error LNK1120: 4 unresolved externals [C:\minetest\realtime\build\src\minetest.vcxproj] 17:39 thexyz I guess class-struct issue strikes again 17:39 thexyz oh well 17:39 thexyz https://travis-ci.org/minetest/minetest/jobs/7848327#L440 17:41 thexyz it's everywhere 17:41 thexyz hm, maybe put -Werror to travis gcc/clang flags 17:42 PilzAdam well, launchpad needs 3 more hours for the builds :-/ 17:42 thexyz PilzAdam: what's our launchpad url? 17:42 rubenwardy Is there an official 0.4.7 build yet? 17:43 PilzAdam https://code.launchpad.net/~minetestdevs/+recipe/minetest-c55-stable 17:43 PilzAdam thexyz, do you mean that ^ ? 17:43 thexyz yes 17:44 thexyz it's not minetest-c55 though, can we rename it? 17:44 celeron55_ lol nobody tested MSVC builds before tagging a release? 17:44 celeron55_ that'll be fun for anyone downloading the stable and thinking he can just compile it on MSVC 8) 17:45 celeron55_ (not that it actually is a viable thing to do at all anymore anyway...) 17:45 celeron55_ thexyz: why not start charging $1000 per build, nobody else can do them anyway 17:45 celeron55_ 8) 17:46 thexyz wait, there's a guy who does amd64 msvc builds 17:46 celeron55_ thexyz: renaming it on launchpad is a... umm... fucking pain in the ass 17:47 celeron55_ it's cross-referenced internally so much by using the full name 17:47 PilzAdam launchpad generally is a PITA, Im happy I manged it to update at all 17:47 celeron55_ (in the recipes and package) 17:48 celeron55_ of course everyone would be thankful if somebody is crazy enough to pull that off 17:48 Jordach I'm* <- PilzAdam 17:49 PilzAdam should I wait with the forum topic until we manage to get a windows build? 17:49 thexyz Jordach: your grammar nazi 17:49 Jordach in front of a german too 17:50 thexyz okay, let's assume we want classes 17:50 thexyz everywhere 17:50 PilzAdam "218 changed files with 20,682 additions and 9,873 deletions." 17:55 thexyz oh, wait, we totally want structs everywhere 18:09 thexyz oh shit 18:09 thexyz somebody should definitely fix all those clang warnings 18:10 Calinou there are lots of warnings in gcc too 18:10 Calinou gcc has a bigger market share, so... do it 18:10 PilzAdam there are almost no warnings in gcc 18:10 Calinou I see about 10 lines 18:10 PilzAdam have you seen a clang build log? 18:10 PilzAdam it has almost 100 lines of warnings 18:12 sfan5 why are there all these // and \\ in the dokuwiki pages? 18:12 kahrl back again 18:12 kahrl what is this mess :D 18:12 PilzAdam \\ == newline 18:12 thexyz most of it comes from hides overloaded virtual function [-Woverloaded-virtual] 18:12 rubenwardy / == itallic 18:12 rubenwardy *// 18:13 thexyz PilzAdam: more like 7k lines 18:13 thexyz is it so hard to not mess with class/struct? 18:13 kahrl the wiki lists tagging the release as the second step, maybe that should be moved to below building and testing the windows version? 18:14 thexyz and what's the point of virtuals if they cannot be called 18:14 PilzAdam also "Update source downloads on minetest.net" is not needed since there are no version specific source downloads 18:16 celeron55_ Calinou: that's not trivial; different gcc versions give different warnings 18:16 celeron55_ Calinou: and everyone uses a different version 18:18 sfan5 is there a windows build yet? 18:18 rubenwardy no 18:19 rubenwardy How do you apply a stub notice to a page? 18:19 rubenwardy on dev.minetest.net 18:22 thexyz rubenwardy: {{Stub}} 18:23 rubenwardy http://dev.minetest.net/Visual_style_guidelines 18:23 rubenwardy thanks 18:24 PilzAdam rubenwardy, what is x in 16^x=32? :-p 18:24 rubenwardy 2? 18:24 rubenwardy oh 18:24 PilzAdam 16*16=256 18:24 PilzAdam also, only 16 pixel textures are allowed 18:25 rubenwardy ok 18:25 rubenwardy fixed 18:26 thexyz I think we should create stubs for all missing articles 18:26 thexyz especially for things documented in lua-api.txt 18:28 kahrl why do people put up with this shit called MSVC 18:29 kahrl its name mangling is violating the standard and this has been known for years 18:29 Calinou MSVC is faster than mingw, the end 18:29 Calinou it also compiles... twice as fast for same program 18:29 Calinou really, it does 18:29 rubenwardy that is what I am doing, thexyz 18:30 thexyz kahrl: okay, let's not support it then 18:33 rubenwardy http://dev.minetest.net/How_to_start_contributing 18:36 celeron55_ kahrl: if it is to be dropped, it has to be done properly or not at all 18:36 celeron55_ so, either decide to drop it or not; if so, announce it and if not, keep proper compatibility 18:37 hmmmm woah woah let's not drop MSVC 18:37 kahrl uh oh I know we can't really drop msvc support, but why can't microsoft fix their shit? 18:37 celeron55_ also it is to be kept in mind that if we drop it, somebody else might start making a patchset that makes it buildable on it 18:37 celeron55_ i know we *can*; i just think dropping is not worth it 18:39 celeron55_ supporting it is hard when nobody makes regular builds of it 18:39 celeron55_ that part really should be fixed 18:42 kahrl so what's the plan now? fix the MSVC troubles and release 0.4.8? 18:43 kahrl or delete the 0.4.7 tags 18:43 celeron55_ just reset the tags; it's not really released yet (there are zero announcements or anything) 18:43 NakedFury choose something like 100 bugs to fix, MSVC troubles, that helicopter texture stuff, add like 10 new stuff then release 18:44 PilzAdam can you just git push upstream :0.4.7 ? 18:44 celeron55_ thexyz is the only one who can do it though 18:44 celeron55_ (fix it, that is) 18:44 kahrl it's been kind of announced in #minetest but I hope they will understand 18:44 PilzAdam we can just say "Windows build takes a while" 18:44 PilzAdam and release it for Linux 18:45 PilzAdam (the launchpad builds are almost done) 18:45 celeron55_ but the windows version of 0.4.7 can't really be a different git hash than linux 18:45 celeron55_ that way it's just messy and inconvenient 18:47 PilzAdam VanessaE and sfan5 already released 0.4.7 builds 18:47 celeron55_ also people will think it's odd if 0.4.8 is made only for fixing the windows build, altough i think we should eventually move to less "remarkable" y.y.x releases 18:47 celeron55_ so that'd be a move to that direction 18:48 VanessaE it's not like you have to build 0.4.7-release more than once anyway. just fix the build issues in the next git commit or somethibng 18:48 kahrl PilzAdam: those are linux builds, right? should not be a real problem 18:48 VanessaE something* 18:49 PilzAdam kahrl, sfan5's is a win build 18:49 kahrl with mingw? 18:49 celeron55_ VanessaE: people expect to download a stable version and build it on any of supported platforms 18:49 PilzAdam yep 18:49 VanessaE how many people do you really expect are gonna try to build it for windows? 18:49 VanessaE your average user ain't even gonna try that 18:50 kahrl the commit that fixes the msvc build wouldn't affect the behaviour of the existing builds unless somebody really screws up 18:50 kahrl so it's still not a problem in my eyes 18:50 celeron55_ it isn't by itself, but if that is taken as a habit, it's baaaaaaaad 18:50 PilzAdam I aggree to kahrl 18:50 celeron55_ if the next version is similarly broken, i'll punch everyone in the face 18:51 celeron55_ deal? 18:51 PilzAdam I like my face... 18:52 PilzAdam Ill update the dev wiki page about releasing 18:54 kahrl https://www.kernel.org/pub/software/scm/git/docs/git-tag.html#_on_re_tagging 18:54 kahrl I'd say it would be best to leave the 0.4.7 tag alone and push a 0.4.7.1 tag 18:55 kahrl if somebody comes to the forum and asks why 0.4.7 can't be built with msvc tell them to try 0.4.7.1 18:55 celeron55_ fine i guess 18:56 PilzAdam or 0.4.7-MSVC ? 18:56 kahrl fine too if that is a valid tag name 18:57 celeron55_ https://www.kernel.org/pub/software/scm/git/docs/git-check-ref-format.html 19:00 kahrl git check-ref-format refs/tags/0.4.7-MSVC <-- returns success 19:15 sfan5 seems like there isn't an official windows build yet.. 19:15 thexyz woa, what a slowpoke! 19:16 sfan5 just wanted to check if I can add the DL to minetest.net/downloads 19:16 sfan5 -s 19:26 thexyz sfan5: no idea 19:27 sfan5 if "ERROR: An unhandled exception occurred: Player::deSerialize(): PlayerArgsEnd not found" occurs , why can't it just tell me who it is trying to deserialize 19:31 kahrl because the exception is unhandled? 19:32 celeron55_ just like a segfaulting code can't work 19:32 celeron55_ just because someone hasn't fixed it 8) 19:34 sfan5 why not just add an exception handler 19:34 sfan5 most commonly the playerfile is just 0 bytes 19:34 sfan5 it could be ignored if an exception occurs while deserializing 19:34 VanessaE I had that error about an hour and a half ago today in fact. 19:35 kahrl player file saving should be fixed, then the exception won't occur anymore 19:35 VanessaE didn't notice whether the player file was null or broken though 19:35 kahrl well, it might still occur on windows machines which don't have atomic file updates 19:35 sfan5 my server had a big downtime, because I didn't notice the player file didn't load :/ 19:36 kahrl ;q 19:36 kahrl oops 19:36 kahrl TIL irssi is not vim 19:37 celeron55_ kahrl: it won't be noticeable as windows machines don't really work at other parts either! 19:37 thexyz VanessaE: broken ram, again? 19:38 VanessaE thexyz: doubt it. Been through that once already. 19:38 VanessaE what is surely needed is to tell me which fricken file it was that was in error 19:40 * ShadowNinja works on a script for that. 19:42 VanessaE I had to nuke everyone's player files again today because of that. 19:43 VanessaE (I don't have time to go through dozens of files looking for broken ones) 19:43 celeron55_ protip: you can just script that 19:43 thexyz VanessaE: that's not very user-friendly 19:44 VanessaE thexyz: I know, but it's a creative server so there's little harm 19:44 VanessaE celeron55_: I don't know the exact process that MT server goes through to validate a player file. 19:45 thexyz okay, so who's going to fix those class/struct errors? 19:46 celeron55_ VanessaE: try "find . -print | xargs grep -iL PlayerArgsEnd" 19:46 celeron55_ actually "grep -IRiL PlayerArgsEnd" 19:46 celeron55_ that'll do it prefectly 19:46 VanessaE that's simple enough. 19:47 sfan5 -I equivalent to --binary-files=without-match 19:47 sfan5 ^ why that? 19:47 celeron55_ because i copied it from stackoverflow 19:47 * ShadowNinja found it too late. 19:47 sfan5 :D 19:48 celeron55_ in this world you never need to know any command line flag combinations, just google it goddamnit 19:49 VanessaE still, not bothering to tell in the log which file was broken was, frankly, stupid. 19:49 VanessaE it's like having a "Check engine" light that flashes or lights up for any fault, including being out of fuel. 19:49 kahrl if you need to do anything more fancy modify this awk script: awk 'BEGINFILE {a=1;} /PlayerArgsEnd/ {a=0;} ENDFILE { if (a) print FILENAME; }' worlds/*/players/* 19:50 celeron55_ VanessaE: if you think somebody ever implemented printing of that error, you're wrong 19:50 VanessaE ? 19:51 celeron55_ it seems you think somebody in some way wrote some check to print that error without the filename 19:51 VanessaE in all the times I've either seen or experienced that error, there's never been a clear indication in the log for which file was the cause. 19:52 VanessaE (seen as in someone else posts about it) 19:52 celeron55_ anyway it could be just fixed, whatever 19:52 celeron55_ (or more like added) 19:52 celeron55_ (there's zero code for handling the error as it's just a runaway C++ exception) 20:10 sapier where do I find the luamainmenu changes to fix minetest path check? 20:39 kahrl class vs. struct normalization: https://gist.github.com/kahrl/5724740 20:40 kahrl I hope my script found all differing declarations 20:41 Exio lets hope msvc can even compile that ;P 20:51 kahrl thexyz: what's the status of the win32 build? 20:53 kahrl should I push the struct-class patch to see if that fixes anything? 20:55 thexyz yes, do that 20:57 kahrl done 21:00 sapier I'd like to fix win32 mainmenu build but I'd prefere to merge the pathcheck changes first 21:02 kahrl sapier, I might add the path functions I wrote to master (after 0.4.7) if the others agree 21:02 kahrl or do you mean something different? 21:02 sapier feature enhandements in feature freeze? 21:03 kahrl after 0.4.7 21:03 sapier sorry ... didn't read exactly 21:03 sapier there are some typos in win32 only code in main menu 21:04 kahrl win32 only code should be moved to filesys.cpp or porting.cpp if possible 21:04 sapier it is in filesys.cpp 21:05 sapier at leat two stupid mistakes 21:09 thexyz http://buildbot.minetest.net/build/NtiTkBmFcBSaBmAzdHG2CB 21:19 kahrl looks good :) 21:40 thexyz why the fuck is gettext disabled 21:50 thexyz okay 21:50 thexyz it has to work now 22:01 thexyz woo, it's done 22:01 PilzAdam whoever cares: amd64 builds of launchpad done 22:01 Exio 0 warnings? 22:01 thexyz http://minetest.ru/builds/minetest-0.4.7.zip 22:01 thexyz Exio: dunno, probably not 22:01 thexyz it seems no one likes the idea 22:01 thexyz of having 0 warnings, that is 22:01 thexyz please, test 22:01 PilzAdam should I create the topic in the forum now? 22:02 thexyz PilzAdam: okay, do that 22:02 PilzAdam http://forum.minetest.net/viewtopic.php?pid=93622#p93622 22:02 sapier I like 0 warnings :-) 22:02 thexyz I hope it works 22:02 thexyz common is dopped 22:03 PilzAdam ..\..\src\minetest\src\client.cpp:1746: Client::ProcessData: Assertion '!m_mesh_update_thread.IsRunning()' failed. 22:03 thexyz so, anyone to check the package? 22:03 Exio thexyz: oh! can you check 'about' the utf8 thing? it seems to work without patches/with upstream, on windows and hmmm's bsd; added a 'workaroud' for linux 22:04 PilzAdam seems like a random bug, works now in wine 22:04 sapier "workaround" bääääähhhhh 22:04 Exio blame irrlicht devs 22:05 thexyz PilzAdam: did I include everything I should? 22:05 thexyz like, minimal and minetest_game only 22:05 PilzAdam yes 22:06 PilzAdam Ill throw the link on minetest.net 22:06 PilzAdam oO someone already did 22:06 PilzAdam oh, "Not yet available" 22:08 PilzAdam done 22:09 ShadowNinja PilzAdam: Common is "dopped". 22:09 PilzAdam fixed 22:10 PilzAdam any way to comment out lines in dokuwiki? 22:10 thexyz ? 22:10 thexyz also, shouldn't the topic be named "Minetest 0.4.7 released!"? 22:11 PilzAdam nope 22:12 PilzAdam (nope = ? <- doenst work) 22:13 thexyz PilzAdam: uh oh 22:13 thexyz there's a plugin for that 22:13 thexyz https://www.dokuwiki.org/plugin:comment 22:15 PilzAdam well, whatever 22:15 PilzAdam added a note that a installer is not available for 0.4.7 currently 22:17 PilzAdam do we really need to set up a new tag for the one commit kahrl pushed? 22:17 PilzAdam or do we just push it into the stable branch? 22:35 kahrl set up the tag 22:35 kahrl you never know if someone might want to retroactively build 0.4.7 on msvc later 22:36 PilzAdam do you want to set it up or should I? 22:38 kahrl do it, I've never used tags so I might mess something up :) 22:38 thexyz I guess just moving a tag should be ok 22:38 kahrl moving? 22:38 PilzAdam moving a tag is rewriting the history, we have a 10 minutes rule for that 22:39 thexyz = delete the old one, create the new one 22:39 thexyz how's that? 22:39 thexyz you mean, deleting a tag is rewriting the history? 22:39 proller PilzAdam, tag needed fof freebsd port 22:40 proller i hope it will be updated soon 22:40 PilzAdam I bet freebsd doesnt compile with MSVC ;-) 22:40 thexyz PilzAdam: http://www.nax.cz/2011/01/15/move-a-remote-git-tag/ 22:40 kahrl if people have already pulled the 0.4.7 tag and continue pulling the minetest repo it won't update the tag 22:40 thexyz enjoy eyecancer 22:40 thexyz kahrl: oh 22:40 thexyz really? 22:41 kahrl I linked that earlier, let me see... 22:41 kahrl https://www.kernel.org/pub/software/scm/git/docs/git-tag.html#_on_re_tagging 22:42 thexyz well, okay 22:43 PilzAdam so, just tag 0.4.7-1 and push to stable-0.4 again? 22:44 kahrl I like your 0.4.7-MSVC suggestion better 22:44 kahrl makes it clear why there is a second tag 22:45 PilzAdam well, when I link the MSVC thing in the topic then people think "hey, why is there MSVC? isnt that for Linux?" 22:45 kahrl hmm 22:45 kahrl or link both? 22:46 PilzAdam yea, "Fixed source for MSVC: " 22:46 kahrl sounds good 22:47 PilzAdam done 22:48 PilzAdam topic updated 22:48 PilzAdam http://forum.minetest.net/viewtopic.php?pid=93622#p93622 22:49 PilzAdam btw: I suggest removing all branches from minetest/minetest except stable-* and master 22:50 PilzAdam c55 has the other ones in his repo 22:50 PilzAdam and many of them are already merged 22:51 kahrl guess that can be done, yea 22:53 kahrl and models_rebase2.master in minetest_game 22:53 PilzAdam ok, will do that now 22:54 PilzAdam oO "PilzAdam deleted branch *" spam at the dashboard 22:55 VanessaE is anyone else getting "builtin/item.lua:339: attempt to call field 'ha 22:55 VanessaE ndle_node_drops' (a nil value)" ? 22:56 PilzAdam nope 22:56 PilzAdam I still think its one of your mods messing with it 22:57 VanessaE I dunno 22:58 VanessaE only one of my mods that I know of that even messes with item drops is technic's item_drop code, and afaik that's been stable for weeks; this started a few days ago. 22:59 PilzAdam from what I heard technic is constantly broken 22:59 PilzAdam ;-) 23:02 VanessaE heh 23:02 Exio so cute 23:02 VanessaE a user on my server reports that technic's chainsaw is what triggers that. 23:02 VanessaE but as the error clearly occurs within builtin/item.lua, that's you guys' territory 23:03 PilzAdam well handle_node_drops is defined in item.lua; if you redefine it and it breaks then its your fault 23:03 VanessaE ah, looks like technic does redefine it. 23:04 PilzAdam of course it does, every item_drop mod does that 23:04 PilzAdam (although the very old versions of my item_drop mod have a really hacky way without redefining it...) 23:05 PilzAdam (mainly because this function didnt exist to that time) 23:05 Exio what did you use? :P 23:05 PilzAdam a funny mixture of drop = "" and on_dignode() IIRC 23:05 PilzAdam but dont ask 23:05 Exio k 23:06 PilzAdam (look at the git commit log) 23:06 Exio i can't at this moment; that is why i asked 23:06 VanessaE figures. and RBA is nowhere to be found for a while (his internet is down) 23:07 Exio he is probably writing very 'l33t' stuff with/for shaderz now :P 23:08 VanessaE nope 23:08 PilzAdam Exio, lol, its even better :D I havent used the global on_dignode() but the on_dig() of each single node! 23:08 VanessaE he's moved to a new apartment over the weekend and won't have 'net access for a few days, so he said. 23:09 Exio PilzAdam: omg 23:09 Exio that is worse than hacky 23:10 Exio VanessaE: that is why i say that, :P 23:12 * PilzAdam spotted first bug report: http://forum.minetest.net/viewtopic.php?pid=93624#p93624 23:13 PilzAdam http://devopsreactions.tumblr.com/post/44776341195/bug-reports-after-a-new-release 23:13 Exio that was fast 23:15 Exio lol PilzAdam (for the link) 23:15 thexyz reproduces in wine 23:15 VanessaE lol 23:15 thexyz PANIC: unprotected error in call to Lua API (bad FPU precision (use D3DCREATE_FPU_PRESERVE with DirectX)) 23:16 thexyz can't reproduce in virtualbox (win xp) 23:17 * PilzAdam heards winedbg can pipe to gdb... 23:17 PilzAdam *heard 23:18 thexyz better pipe this output to google 23:18 thexyz http://irrlicht.sourceforge.net/forum/viewtopic.php?t=8773 23:34 PilzAdam "Remove supoort for common mods after 0.4.7 or so https://github.com/minetest/minetest/commit/306d1ab866a3ce820e95f4faf805684cd4122ae4" 23:35 PilzAdam ^ found in TODO 23:36 PilzAdam I would say just remove the common loading and leave the getGameConfig() stuff in 23:46 PilzAdam like this: https://github.com/PilzAdam/minetest/commit/c4d22d6e666c89658434b9636519560e956473e8