Time Nick Message 10:17 seolfor darkrose, nore can i ask you question ? 10:17 seolfor can i use curl in version 7.36.0 instead of 7.33 ? 10:54 sapier As of adapting mainmenu for android devices I realized mainmenu design got quite strange while adding all those legacy features. As result some parts of it are quite hard to maintain. Doing some design cleanup does drasticaly improve this situation. Basicaly I took beginings of a formspec based toolkit some steps further. 10:55 sapier https://gist.github.com/sapier/10578873 shows a (early) suggestion for the fst (formspek tookit) API ... comments are welcome 10:55 seolfor sapier, i got question about curl lib 10:55 seolfor do i have to use version 7.33 ? 10:56 seolfor ther is new one 7.36 but i can't compile it ;( 10:58 sapier well if you want a quick solution take 7.33 ... but you could find out how to build 7.36 too and update the docs 10:58 seolfor hmy .. ok i will try 11:01 seolfor and i spend whole day on attempts to compile dependencies with VS2012 - it's to hard or infformations in wiki are not a accurate inive 11:02 sapier hmm at least the information I provided should be accurate I did it exactly this way 11:02 sapier on win7 64 bit 11:02 sapier but as I said windows build is tricky took me almost a week to do mingw + vs the first time 11:02 seolfor did you have to configure linker manulay for evry package ? 11:03 sapier there aren't many libs depending on each other? 11:03 sapier the only one depending on everything is minetest itself ... and yes you have to specify all those paths manually in minetests cmake gui 11:05 seolfor i been compiling using buildin engine in Visual Studio 2013 (Visual Studion 2012 x86) on win7 64bit 11:05 seolfor and not even 1 lib compiled without manual configuration 11:05 seolfor not even single* lib 11:06 sapier well that's pretty normal for win32 ... most likely that's reason why most developers scream and run away once you ask for working on windows ;-) 11:06 specing your problem is using windows 11:06 seolfor haha ye i known 11:06 sapier don't discourage him specing we need people developing on windows too 11:07 seolfor "Change to linux it will sove evrthing" <- best solution it can be made 11:08 seolfor right now im using mingw32 and it's look good 11:08 seolfor for now only problem i got with curl 11:09 sapier hmm depends on version current version of mingw (compiling all libs) doesn't build master 11:09 sapier there's a issue with zlib ... for some insane reason someone believed using msvc dll for mingw would be a good ide 11:09 sapier a 11:10 seolfor isn't mingw is compiler helper like Cmake or make ? 11:11 sapier no mingw is just a compiler suite 11:11 specing sapier: fuck windows ;P 11:11 seolfor specing, thx .. 11:11 specing I get angry a lot when I have to work on windows at school 11:11 sapier if there wasn't windows apple would rule the world ... and those guys are even more evil microsoft was in it's worst time 11:12 specing sapier: doubt it 11:12 specing apple is expensive, not many people can afford it ;P 11:12 sapier no doubt, microsoft never did take rigth to delete arbitrary data from users device 11:13 seolfor im not angry i just like windows desktop envirment in win 7 11:13 seolfor and i dont have space on my ssd for second OS like linux 11:13 seolfor :p 11:14 sapier well windows 7 shell is still one of best shells available ... not like that crappy gnome3, windows 8 or even ios or kde shells 11:14 sapier of course that's my personal opinion only 11:14 sapier and limited to gui shells 11:15 sapier bash is way better then windows command line 11:15 seolfor indeed gnome i crappy i preffer using cinnamon but it's still not win7 desktop 11:15 seolfor first thing i install in my win is bash shell 11:16 seolfor i just can't use cmd crap 11:32 specing seolfor: delete windows. That'll make room for Linux 11:32 specing seolfor: And learn to use a tiling WM 12:04 seolfor specing, great joke :p but what did you mean by WM ? 12:04 seolfor Virtual Machine ? 12:04 PenguinDad seolfor: WM == Window Manager 12:06 specing regular window manages and desktop environments(huh) suck so much 12:06 specing I grief badly when I have to use em 12:06 specing s/grief/raaaaaaaaageeee/ 12:08 seolfor PenguinDad, thx. but i agree wigh specing ther isn't single one desktop env. for linux capable to hendle doual/triple screen well... 12:08 seolfor this is one of reasons i wond delete windows soon 12:08 seolfor won't* 12:12 specing awesomeWM handles N-screening extremely well 12:13 seolfor realy ? i will test it for shure ;] 12:14 specing As its name says, it is *AWESOME* 12:15 seolfor like I sad i will for sure test it as soon i will compile this freeking curl lib 13:10 seolfor i'm thinking what "run cmake and configure correctly" position mean in http://dev.minetest.net/Build_Win32_minetest_including_all_required_libraries 13:12 BlockMen seolfor, i guess the author ment that you get no configuration erros of cmake when pressing "configure" 13:13 seolfor thx BlockMen, but i come to great idea, becouse i have already all compiled dependencies i will try to use VS project to bild minetest ;D 13:19 seolfor why ther is cmake_config_githash.h.in instead of cmake_config_githash.h in sources of minetest ? 13:19 seolfor becouse of it i got compile errors ;( 13:23 BlockMen google reveals: "These files are usually header template to generate final .h files." 13:24 BlockMen you are sure that all if correct cofigured in cmake? 13:24 seolfor no i didn't used cmake 13:24 seolfor i tried compiling it trhought VS 2013 compailer 13:25 seolfor but it's looks like it's not capable of including .in files 13:25 BlockMen you should do to create correct vs project file. then you can use the vs compiler to build minetest 13:25 seolfor i corrected includes paths and linked libs 13:25 seolfor but still .. 13:26 BlockMen https://forum.minetest.net/viewtopic.php?pid=118990#p118990 13:26 BlockMen maybe this helps you 13:27 seolfor nop not a chance im using culr in version 1.36 and's its simply to old 13:30 BlockMen first: why are you using such old curl version? second: i think there is not easy way to built minetest without cmake (if there is at all), so you better use it 13:32 seolfor ups sory i mean curl 7.36* 13:33 seolfor right now i'm trying to use cmake becouse ther is not a way to bild *.in files in VS 15:07 seolfor what is the diffrence betwen VORIBS_DLL and VORBISFILE_DLL in cmake config ? 15:07 seolfor with file should i point for them ? 15:07 sapier you don't need to use curl nor vorbis 15:08 seolfor what o0 15:08 sapier maybe you should start with minimal library dependencys and once this works add the additional one 15:08 sapier s 15:08 seolfor great to known .... ;( 15:08 sapier well most people don't wanna play minetest without online or sound support but it is possible to compile this way 15:09 seolfor so how can i compile without vorbis libary ? 15:09 sapier yes just uncheck ENABLE_SOUND 15:09 seolfor and btw i copiled all dependences lited on wiki 15:10 seolfor but i simply dont know what file should i include as VORBIS_DLL 15:10 seolfor and VORBISFILE_DLL 15:12 sapier you can ignore vorbisfile_dll and vorbis_dll 15:13 sapier you only have to set the libraries correct 15:13 seolfor and Libraries are ... ? for vorbis and vorbisfile ? ^_^ 15:13 sapier most time they're dll.a (for mingw 15:14 seolfor ah .. ok thx 15:14 sapier luajit is .dll for me. I thought you're doing the vs2xxx build? 15:15 sfan5 seolfor: do you have all libraries you need yet? 15:15 sfan5 I have some 15:15 seolfor i didn't manage to make it .. 15:15 seolfor sfan5, i think so.. 15:16 seolfor i'v been bilding them since yestarday 15:18 seolfor omg i finally recived message form cmake "Generatin done" 15:18 seolfor ^_^ 15:31 seolfor bilding proces is running but i getting syntax erros form minetest 15:32 PenguinDad seolfor: real errors or only warnings? 15:32 BlockMen PilzAdam, set_bone is not broken, it works fine. You have disabled animation in singleplayer mode: https://github.com/PilzAdam/nodetopia/blob/master/mods/models/init.lua#L41 15:32 seolfor errors 15:32 seolfor Error 1 error C2059: syntax error : '{' 15:34 seolfor but it's looks good in cource... ;( 15:38 seolfor minetest is using c++ v11 standart ? 15:38 Calinou I don't think so 15:39 PenguinDad I've never heard of c++ v11 :/ 15:41 seolfor PenguinDad, you can read about it here http://www.cplusplus.com/info/history/ ;D 15:42 seolfor PenguinDad, and ex. changes : http://www.cplusplus.com/reference/stl/ 15:48 seolfor all sytax error im getting are form clientiface.h 15:49 seolfor for example form it : http://pastebin.com/pVGPLajE 15:53 BlockMen seolfor, use lastest version minetest code 15:54 BlockMen its fixed alreday 15:55 seolfor ah .. indeed it's not stable buld and 2 days old 15:55 seolfor thx for tip 15:57 BlockMen i ment use always lastest code from https://github.com/minetest/minetest, then report issues 15:58 BlockMen *coderelated issues 15:59 seolfor i known it but .. i been fighting almoust 2 day only with dependencies so I thought i still dooing somthing wrong 16:03 seolfor BlockMen, wow i got now: Error 2 error C2039: 'max' : is not a member of 'std' 16:03 seolfor it's only for me or it's a bug - any way is strange .. 16:04 seolfor becouse i got problems only with max function form std and it shoudl be there .. 16:04 BlockMen what file? 16:05 seolfor profiler.h 16:08 BlockMen try adding header 16:08 BlockMen *to 16:10 BlockMen and seems to be a vs2013 special thing 16:11 seolfor indeed im bilding with vs2013 16:21 seolfor BlockMen, it's seems to solved problem with std::max 16:47 seolfor i updated wiki page http://dev.minetest.net/Build_Win32_minetest_including_all_required_libraries for never versions of dependancies (changes allways can be removed) 16:55 BlockMen i thought you use vs2013, why do you wrote the informations under Mingw32 then? 16:55 seolfor becouse i conpiled dependencies under mingw32 16:56 ShadowNinja sapier: Formspecs need to be redesigned, not have a fancy string parsing wrapper. I made a mockup here: http://sprunge.us/XDUS?lua 16:56 seolfor any way i sill didn't compile enything on vs2013 ;( 16:57 PenguinDad this looks much better than what we have at the moment 16:58 seolfor mayby somebody can explain me this error form mingw32-make ? 16:58 seolfor mingw32-make: *** No targets specified and no makefile found. Stop. 16:58 BlockMen ShadowNinja, is this '"Here's a table:\n",' supposed to replace labels? 16:59 ShadowNinja BlockMen: Yes. 16:59 BlockMen how do you set position then? 16:59 ShadowNinja BlockMen: Elements are either a table describing the element or a string, which is equivalent to {type="label", text="..."} 17:00 BlockMen ah ok, thats fine then 17:00 ShadowNinja You can use x= and y=, or use HTML-like auto-positioning. 17:00 PenguinDad seolfor: it tells you exactly what the problem is "no makefile found" 17:00 BlockMen looks like a good concept to me 17:02 BlockMen size should support percent values to in rewrite 17:02 BlockMen *+o 17:02 seolfor ;( but already run cmake and press "generate" - it's finished without errors .. 17:03 seolfor should not i have make file already ? 17:03 ShadowNinja celeron55: I find CMake's syntax very convoluted. I've used SCons in my projects, which uses Python scripts. It's cross-platform too. 17:07 celeron55 i have never used scons (except built something made by someone else using it) 17:07 celeron55 is there a good comparison available somewhere? 17:07 celeron55 syntax, platforms and stuff like that 17:09 ShadowNinja celeron55: Well there's this, which is fairly complete but possibly biased: http://www.scons.org/wiki/SconsVsOtherBuildTools 17:09 ShadowNinja celeron55: And here's a (very) basic example: https://github.com/ShadowNinja/NBT-CPP/blob/master/SConstruct 17:11 ShadowNinja celeron55: This is slightly more advanced: https://github.com/kaeza/minetest-irc/blob/master/SConstruct 17:12 celeron55 doesn't seem to differ much 17:12 celeron55 it's just the same thing, with python 17:12 celeron55 would you really be willing to rewrite all the configuration options in minetest's build system in that? 17:12 celeron55 and then deal with all the regressions for months to come 17:14 ShadowNinja Um, maybe. I'd rather start with a smaller project like minetestmapper though. 17:14 celeron55 i do hate cmake's macro language; but this isn't going to magically fix itself by just saying "let's use scons" 18:53 sapier ShadowNinja: that redesign wouldn't solve any of the issues I'm addressing with my toolkit 18:55 sapier my toolkit is meant to be a evolutionary addon to formspecs with quite calculatable effort to be finished while your (current) suggestion is just replacing current formspec but not adding any of the features I need for the menu 18:56 sapier if you provide this replacement and rewrite menu this way you're welcome. I'm glad if I don't have to maintain it any longer but I wont gonna write this total conversion with no obvious major benefit to our current version for sure. 19:02 sapier especially as I consider your suggesion way to underambitioned to justify that much work 19:04 ShadowNinja http://pastebin.ubuntu.com/7246313/ 19:05 sapier scons is neither standard nor anything close to it ... what's wrong with cmake? 19:05 ShadowNinja http://ix.io/bEx/diff 11 files changed, 106 insertions(+), 164 deletions(-) This makes minetestmapper use SCons. 19:06 ShadowNinja sapier: What build system standard is there to conform to? 19:06 sapier yes and another build system to install with don't know how much dependencys and don't know how many additional bugs and another syntax to handle 19:06 ShadowNinja It uses Python, so it's syntax is more standard. 19:06 ShadowNinja sapier: I cannot parse that sentence. 19:07 sapier I've never written python and I know quite a lot developers never ever having heared about python ... of course if you know any language you know (almost) all ... but the subtile difference still are annoying 19:08 sapier and you haven't told about a real benefit by now ;-) 19:08 seolfor i finly manage to compile in vs2013 but aftrer execution attempt i got missing error zibwapi.dll, enybody known wher can i find it ? 19:08 iqualfragile sapier: you do not really need to know python to write python 19:08 ShadowNinja sapier: Python is much easier to learn than CMake's language. 19:08 ShadowNinja celeron55: ^ Thoughts? 19:08 iqualfragile ShadowNinja: ~ he complains about scons having bugs and dependencies 19:09 sapier I don't wanna discuss about languages thats nonsense I wanna know what scons claims to handle that much better to justify changeing a working buildsystem without major flaws 19:09 seolfor ShadowNinja, i agree python is pice of a cake compare to cmake 19:09 iqualfragile was kinda direct translation from german 19:10 ShadowNinja iqualfragile: Ah, the only dependency that SCons has is Python. And I have only found one thing that could be called a bug, but really indicates a problem in your code. 19:10 sapier WHATS SCONS BETTER THEN CMAKE ... NO SYNTACTICAL VRAP BUT REAL FEATURES ... will build be done in half time for example? 19:10 ShadowNinja sapier: The reason is simply that CMake is harder to use 19:10 iqualfragile well, understanding your build system seems like a bonus for me 19:10 ShadowNinja And no need to get worked up. 19:11 sapier sorry but if you're unable to typecmake on console you most likely shouldn do development 19:11 iqualfragile ShadowNinja: but he allready asked that question without getting an answer 19:11 ShadowNinja SCons *Might* have better implict dependency handling. But it's have to try it on a bigger project to know. 19:12 seolfor btw can i get an answer meen while fight python vs cmake ? 19:12 ShadowNinja i'd* 19:12 sapier and I don't expect windows developers to be any more successfull with installing python and scons then with installing cmake only 19:12 ShadowNinja sapier: SCons and Python are available for Windows. 19:12 seolfor for shure i will not test it after finly executing build in cmake 19:12 ShadowNinja You can even bundle SCons with the project to remove that dependency. 19:13 sapier yes vs2012 is available too excel too .... wait excel is spread quite a lot why not build with excel? 19:13 iqualfragile sapier: ' '.join(["windows", "developers"]) 19:13 sapier I've still not become a answer what bug scons is supposed to fix 19:13 iqualfragile IncompatibleDatatypesError 19:14 ShadowNinja sapier: Um, are you proller?... 19:14 seolfor iqualfragile, it's not so bad :p 19:14 ShadowNinja I told you. It's easier to use and has more familiar syntax. 19:14 iqualfragile https://github.com/minetest/minetest_game/commit/a5dcf123e0d9b568569118d1a05e24f56e7b5d69 but why? 19:14 sapier no I just hate switching tools for sake of switching, I've got better things to do then learning and fixing bugs in new tools because someone felt this to be more cool then another one 19:15 sapier especially if I don't have any problem with the current tool 19:15 ShadowNinja sapier: As for formspecs: A DOM-like library like you're proposing sounds good, but it would be much more natural with a table-based formspec. 19:16 seolfor point for sapier :p 19:16 iqualfragile sapier: have not read the discussion but implementing something dom-like instead of giving up and using html/xml seems wrong to me 19:17 sapier As I said ShadowNinja imho your suggestion is underambitioned for that less benefit doing that much work isn't worth it 19:17 ShadowNinja Otherwise it would need a formspec parser and writer. 19:17 seolfor well now sapier, ShadowNinja it dosn't metter what it is cmake or scons but plz explain me why in zlib source isn't zlibwapi.dll 19:19 sapier zlib is a very special case, if you use mingw you shouldn't use the prebuilt ones as mingw provides it's own version, if you mix up those two versions it's not gonna work. Sadly master is buggy and doesn't use mingw zlib you need to patch it manually 19:19 seolfor i made freking build of minetest i linked zlibwapi.lib but i can't wind enywhere zlibwapi.dll 19:19 ShadowNinja seolfor: Huh? That sounds like a Zlib windoes shared library. 19:19 ShadowNinja seolfor: You linked a static library and it's looking for a shared one? 19:20 sapier but fixing this bug causes buildbot to fail 19:20 seolfor sapier, i made build of zlib with vs2012 so it should be compatiblie 19:20 ShadowNinja sfan5: You should look at ^ too. 19:20 sapier nope seolfor as mingw compiles everything else against it's built in zlib 19:22 iqualfragile hmmmm: we had a discussion about map generators some days ago 19:22 iqualfragile i build something in python, which outputs gnuplottable stuff, but i mainly have problems with getting the wind right 19:22 sapier ShadowNinja: if we didn't have our current formspec I'd agree to your syntax. But I don't feel this syntax to be advanced enough to be the final (or at least a long term) solution. And I don't have any interest in spending that much time to fix regressions and do maintenance for a interim format ... that has to be maintained with the final solution too. 19:23 iqualfragile if you want to take a look at it tell me and i will publish the repo somewhere 19:23 ShadowNinja sapier: Suggest a good alternative them. 19:23 seolfor sapier, im don't understand, so i have to bild evrything in vs2012 ? 19:23 iqualfragile seolfor: 1st step: install debian 19:23 seolfor or am i have problems only with zlib ? 19:23 sapier I don't have enough time to implement a really good solution. That's a task for someone else. 19:24 sapier yes either build in mingw or vs2012 seolfor 19:24 iqualfragile the cammera possition/mode stuff seems to raise new waring during compile 19:24 iqualfragile s/mm/m 19:24 sapier hmm pulls with warnings shouldn't be merged 19:25 seolfor sapier, thx .. 1 more try if it wont work i will install linux >.> 19:25 ShadowNinja sapier: Oh, I added experimental thread name support for Windows, BSD, and OSX here: http://ix.io/bBp/diff PilzAdam, sfan5: Can one of you make a Windows build? 19:25 sapier pthread_setname_np isn't generally available on linux 19:26 PilzAdam ShadowNinja, no, not now 19:26 PilzAdam I havent done builds in a while 19:26 ShadowNinja PilzAdam: Alright. 19:26 iqualfragile sapier: http://bpaste.net/show/mBI4gTLT3E4DdzacvakL/ 19:26 PilzAdam ShadowNinja, BlockMen does win builds too 19:27 ShadowNinja sapier: prctl could be used too, but that's been in glibc since 2.12 or so. 19:27 ShadowNinja BlockMen: ^ 19:27 sapier windows does have threadnames? 19:27 ShadowNinja Er, well, you hilighted him already. 19:27 ShadowNinja sapier: Seems so: http://msdn.microsoft.com/en-us/library/xcb2z8hs.aspx 19:28 sapier well I guess I'll be way more successfull to find a linux without pthread_setname_np then you're gonna find a linux 2.12 19:29 ShadowNinja sapier: I don't know what you mean, but I was refering to GlibC v 2.12, not Linux v 2.12 19:29 sfan5 ShadowNinja: how would you cross-compile with scons? 19:30 ShadowNinja sfan5: Hmmm, how do you with CMake? Just add -march=win32 or some such? 19:30 sfan5 look into buildbot.sh 19:31 proller sapier, but why you merge pulls with lot of warnings? 19:31 sapier ahhhm I'm on wheezy and pthread_set_name_np isn't available? 19:31 ShadowNinja sapier: Hmmm, glibc 2.12 seems over 10 years old... 19:32 sapier proller I usually check my pulls to not add any warning on gcc and by if someone tells me about a warning on another compiler I'll fix it. But I can't fix things I don't know about. 19:32 proller try to use clang 19:33 sfan5 ^ 19:33 sapier keep on talking proller 19:33 sapier If I didn check anything anyone ever asked to check I'd have to run 1 week of testing prior commiting a 1 line comment fix 19:34 sfan5 ShadowNinja: could you make a diff -Ru for thread names? 19:34 sfan5 oh, nvm 19:34 celeron55 >Python is much easier to learn than CMake's language. 19:34 celeron55 that's not true 19:34 celeron55 cmake is basically shell scripting 19:35 celeron55 ...but simpler 19:35 celeron55 (which is why it sucks for complicated things) 19:36 sapier ahh ok its pthread_setname_np on linux ... who did steal that "_" 19:36 celeron55 (eg. it doesn't have support for any kind of arrays; everything is a string) 19:37 ShadowNinja sfan5: http://ix.io/bEA 19:37 ShadowNinja Er, NVM too. :-P 19:38 sapier And proller if you're that eager about fixing warnings why didn't you agree to any of those two attempts to fix them? someone tried half a year ago, I tried 2 months ago 19:38 ShadowNinja There aren't many, if any, warnings with GCC. 19:39 sfan5 ShadowNinja: Hunk #1 FAILED at 267., make a diff -ru (that works) please 19:39 sapier ShadowNinja: there are quite a lot on clang 19:39 ShadowNinja sfan5: You said -Ru! 19:40 sfan5 -R is not a valid option 19:40 ShadowNinja sfan5: It is, it means reverse. 19:40 ShadowNinja http://ix.io/bBp 19:40 ShadowNinja -R -- Swap two inputs; that is, show differences from index or on-disk file to tree contents. 19:41 sfan5 my diff does not know that option 19:41 sapier ShadowNinja: glibc 2.12 is 3. Aug 2010 ... far from 10 years old 19:41 sfan5 thanks 19:42 sapier debian squeeze has glibc 2.11 in ... and isn't outdated yet 19:42 ShadowNinja sapier: Yes. I think we can safely rely and people having a <14 year old glibc. 19:43 sapier 4 YEARS ShadowNinja NOT 14 19:43 sapier you miss a full decade 19:43 ShadowNinja Er, right. 19:43 * ShadowNinja bangs his head against the wall. 19:43 sapier no we can't I still use squeeze on most of my servers and I guess quite a lot people do too 19:44 sfan5 ShadowNinja: that diff still fails to apply 19:45 sapier sfan5: do it manually it's minor 19:45 ShadowNinja Well, I guess we can't then. 19:45 sfan5 but I'm to lazy :s 19:46 sapier mid term I agree switching to pthread for all (used) os but until squeeze support has ended I'd prefere not break minetest on good old standard debian :) 19:47 ShadowNinja sfan5: Seems like cross compiling will only require env.tools.append("mingw") or some such. 19:47 sapier maybe we should check if debian backported that feature. I have doubts ... but maybe 19:47 sfan5 what if I want to specify a compiler? 19:47 ShadowNinja sapier: It's trivial to change that to prctl with a note that it should be done with pthread later. 19:47 sapier export GCC=clang 19:48 sapier export CC=clang 19:48 sapier without g 19:48 sapier and export CXX=c.lang 19:48 sfan5 does scons have a GUI? 19:48 sapier yes I think so too ShadowNinja in general that patch is quite good 19:48 ShadowNinja sfan5: Load the appropriate Tool(), or manually set env["CC"] and the like if your compiler isn't available in the default Tools. 19:48 sfan5 a-la cmake-gui 19:49 sfan5 /home/stefan/mt-build/build/c55/minetest-minetest-8b02a01/src/porting.h:283: error: ‘SetThreadName’ was not declared in this scope 19:49 ShadowNinja sfan5: Nope. 19:49 ShadowNinja sfan5: Hmmm, what version of Windows? 19:49 sfan5 dunno 19:49 sapier that patch is based uppon a outdated (broken) version 19:50 sfan5 i586-mingw32msvc-gcc (GCC) 4.2.1-sjlj (mingw32-2) 19:50 sapier I fixed windows build yesteday because of that bug 19:50 sapier most likely that's reason why it doesn#t apply too 19:51 ShadowNinja Oh. 19:51 sapier gcc 4.2????? that's ancient 19:52 sfan5 sapier: complain to the debian/ubuntu maintainers 19:52 sapier what os? 19:52 sfan5 mine? xubuntu 19:53 sapier version ;-) 19:53 sapier even squeeze has gcc 4.4 19:53 sapier 4.2 ... etch? 19:54 sfan5 13.10 19:54 sapier wait .. mingw on linux? cross compiler? 19:54 sfan5 yes 19:54 sapier why not use a vm? 19:54 sfan5 too much work? 19:54 sapier compared to work required to get that ancient versions work? 19:55 sfan5 way less 19:55 sfan5 gcc 4.2 compiles minetest just fine 19:55 sapier ok :) if you think so :) I've installed win7 on a vm and latest mingw within a few minutes ;-) 19:56 sapier especially as I feel it's not gonna be much time till ppl demanding c11 standard could become successfull 19:56 ShadowNinja Here's an updated diff: http://ix.io/bEB 19:57 sapier why didn't you fix the prct thing as you already touched it? ;-) 19:57 ShadowNinja I don't know when SetThreadName was added, but it should be in windows.h. 19:57 ShadowNinja sapier: Er, right... 19:57 sfan5 *compiles new mingw* 19:58 sapier It's been added last week when the threadname support was pushed 19:58 iqualfragile debian maintainser seem to love to patch 19:58 iqualfragile oh w8 19:58 * iqualfragile was stupid 20:01 ShadowNinja CMake seems to *always* rebuild the ScriptAPI. 20:01 ShadowNinja Unless nothing's changed at all. 20:03 sapier my squeeze pthread.h doesn't contain pthread_setname_np so at least debian didn't patch glibc-2.11 ... prctl will have to do till sequeeze support is ended. As jessie doesn't seem to have a release date that could be anything from a couple of weeks to about a year 20:03 sapier at least the last years debian had a 2 year cycle if they keep it jessie is about to be released spring 2015 20:06 iqualfragile its nice to have minetest in debian 20:12 ShadowNinja Comments? Reorders initialization to make GCC happy: http://ix.io/bEG 20:13 ShadowNinja sfan5: Progress? 20:14 sfan5 progress with what? 20:16 ShadowNinja sfan5: The Windows build. Does a newer MinGW work? 20:16 sfan5 I'm currently compiling MinGW 20:18 ShadowNinja Ah, I think I found out how to do it, but it will be MSVC only. 20:18 ShadowNinja (And ridiculously ugly) 20:32 sapier how to show the threadnames on windows? 20:32 ShadowNinja This should work for Linux, BSD, MSVC, and OSX. Other windows compilers are ignored: http://sprunge.us/RgRS?diff 20:33 ShadowNinja sapier: Look at that. It's a rediculously ugly hack, but it should work, and Microsoft seems to support it. 20:34 sapier by using an exception? :-) very strange way to do it :-) 20:34 seolfor sapier, i'v copiled all dependencies with vs2012, made proper config with cmake, succesfully build maintest and in runing attempt i reciving error like Kernel32.dll not found 20:34 sapier one of the most uglyest hack's I've ever seen 20:35 sapier you did compile minetest with vs2012 too? 20:35 seolfor yep 20:36 sapier can you try a tool like dependencywalker on your resulting binary? kernel32.dll is windows itself, I don't believe you're relly missing it 20:36 BlockMen ShadowNinja, your patch was not compiling at all, so i found this solution in inet https://gist.github.com/anonymous/2f6e2440976fcc969f32 20:37 seolfor i will try to find this "dependencywalker" 20:37 BlockMen oh, you posted a new one. 20:37 ShadowNinja BlockMen: ^ My latest one should work. 20:37 sapier http://www.dependencywalker.com/ seolfor 20:38 BlockMen ShadowNinja, will try 20:38 seolfor i thought it's build in vindows 20:38 seolfor windows* 20:38 ShadowNinja I thought that SetThreadName was a function in the Windows API, because you couldn't have possibly designed something THAT hacky, right?... 20:39 sapier there's no design it's just a function to be called in portab layer 20:40 ShadowNinja sapier: I mean, I didn't think Microsoft would have thought up something so bad and included it in their OS. 20:40 sapier actually that code is about half a year old and was linux only by that time 20:41 sapier I don't even have an idea where to see threadnames on windows? on linux I can lookup in proc filesystem and using top ... how to see them on windows? 20:42 seolfor sapier, only thing what's red in kernel32.dll is check sum http://snag.gy/3pXQg.jpg 20:42 seolfor what can i do with it ? 20:43 ShadowNinja sapier: VC debugger: http://www.codeproject.com/Articles/8549/Name-your-threads-in-the-VC-debugger-thread-list 20:43 sapier any way to see them without debugger? ;-) 20:44 sapier usually I don't have a debugger attached to a minetest server/client going crazy ... it's often a release build too ... ok that wouldn't be a problem for this situation 20:45 sapier but it's better having it then not having anything, of course having a tool to read that information would be even better 20:46 ShadowNinja sapier: Um, the task manager might be able to show them. 20:46 sapier shadow can you create this diff as pull request? 20:46 ShadowNinja sapier: Sure. 20:47 sapier if noone complains till tomorrow I guess it's fine to merge 20:49 ShadowNinja https://github.com/minetest/minetest/pull/1222 20:49 ShadowNinja BlockMen: Can you confirm that that works on Windows? 20:51 BlockMen ShadowNinja, wait a minute... 20:53 BlockMen sapier, https://forum.minetest.net/viewtopic.php?pid=137514#p137514 forgot to report, i noticed after switching to formspec pause menu 20:54 BlockMen ShadowNinja, if #1222 is like your patch, then yes, i can confirm 20:54 ShadowBot https://github.com/minetest/minetest/issues/1222 -- Add support for threadnames on BSD, Windows (MSVC-only), and OSX by ShadowNinja 20:54 BlockMen sapier, and it does only happen sometimes to me. maybe some threading problem? 20:55 sapier is there a way to reproduce it BlockMen? 20:57 BlockMen start a game, go into pause menu and then quit menu again (except exit application). but since its not always happening i cant reproduce it reliable 20:58 BlockMen and for some reason i cant debug atm 20:58 sapier should be fixed prior next release 20:59 sapier but the error doesn't help very much 20:59 BlockMen thats the problem...i will try to fix the debug mode 20:59 BlockMen *s/the/my 21:18 PenguinDad I don't like how Sam's arm turns when digging https://mediacru.sh/4XOIJxdpNIUE :( 21:25 PenguinDad ^ Taoki this is related to https://github.com/minetest/minetest_game/pull/251 21:26 Taoki PenguinDad: Jordach corrected that, but it should be fine? 21:26 Taoki Need to check it better, I didn't notice any changes to animations 21:29 PenguinDad Taoki: this happened after I applied #251 21:29 ShadowBot https://github.com/minetest/minetest/issues/251 -- Add #include guards to base64.h by matttpt 21:32 Taoki https://forum.minetest.net/viewtopic.php?pid=137530#p137530 Opinions? 21:35 sapier quite interesting 21:36 Taoki I initially wanted to make a simple player list, showing players and their ping. But I realized we could have more than Minecraft, such as icons and profiles / descriptions :) 21:36 Taoki I think this will motivate online play a bit more. Since players can get something that offers them more personality 21:37 Taoki Not sure how and if to use profile info and avatars from minetest_game by default, but even if it won't be used there other games / mods / servers can do so. Will go nicely with player skin selection mods :) 21:48 seolfor sapier, it's seems i made it runable but i geting some test's error can you look at this log http://pastebin.com/42iY16GD 21:49 seolfor or anybody else ? 21:49 ShadowNinja Pushing in a minute if there are no objections: http://ix.io/bEJ 21:50 sapier seolfor add "--disable-unittests" to command line 21:50 ShadowNinja This too: http://ix.io/bEK 21:50 sapier some of the tests are quite timing criticak 21:51 sapier http://ix.io/bEJ is fine and I'd make this http://ix.io/bEK a error message only 21:51 sapier wait that's configuring a timer? 21:52 sapier if yes you should check "func" to be a function too 21:52 ShadowNinja sapier: minetest.after(1, nil) is a real error. 21:52 seolfor now i got "ERROR FROM LUA WHILE INITIALIZING ASYNC ENVIRONMENT" 21:52 ShadowNinja Func is checked. 21:52 sapier but only to be not nil 21:53 sapier minetest.after(1,"someone") would still work 21:53 ShadowNinja True. 21:53 sapier interesting seolfor what machine do you have? did you compile luajit too? 21:54 seolfor yep but i didn include them in cmake config 21:54 seolfor so mayby that is the case .. 21:54 seolfor from cmake gui console : LuaJIT not found, using bundled Lua. 21:55 sapier no I ask because we've got some race condition on debian luajit with quite fast machines ... but as you don't have luajit this must be something different 21:56 seolfor here is whole log form minetest 21:56 seolfor http://pastebin.com/zj3r5dbb 21:57 seolfor and log form cmake: http://pastebin.com/Lqrft5y6 21:57 sapier ah that file isn't there ... can you look for it and check it is there? 21:57 sapier C:\Users\Art\Documents\projects\minetest\bin\Debug\..\builtin\async_env.lua 21:58 seolfor yep the path is wrong it should have ../../ instead of one ../ 21:59 seolfor who can i set it to proper path ? or should i move builtin ? 21:59 seolfor where* 22:01 sapier start minetest from base folder not from bin 22:01 sapier bin\minetest 22:03 ShadowNinja sapier: http://sprunge.us/OBNf 22:03 seolfor if i do it i have other path problems >.> like ../../src/script not foud 22:03 seolfor i copied builtin in to bin and it worked 22:04 sapier ok will do for the moment seolfor 22:05 ShadowNinja Also, I need this http://ix.io/bcp to get Minetest to load a world now. 22:05 sapier ShadowNinja: what happens if you call type of nil? 22:06 ShadowNinja sapier: "Assertation failed" followed by a traceback. 22:06 seolfor it's great it's wokring at least ;] 22:06 ShadowNinja Or rather "Invalid minetest.after invocation". 22:08 sapier ShadowNinja: that's a workaround fix the invalid nlx nly or nlz value 22:08 ShadowNinja hmmmm: Any ideas why I need that hack? 22:08 ShadowNinja sapier: Well, duh... 22:08 sapier don't push it it's hiding bugs 22:10 ShadowNinja Hmmm, this seems odd: http://pastebin.ubuntu.com/7247067/ 22:10 ShadowNinja sapier: Yes, I won't push that, but I need it locally. 22:11 sapier float to int casting? 22:13 ShadowNinja No, the comment mentions (sz * spread * ofactor). The code does (sx * ofactor / spread). 22:14 ShadowNinja Changing it to multiplication fixes it. 22:22 BlockMen what is this supposed to do? why is quit=true returned when quit_mode is cancel? 22:22 BlockMen https://github.com/minetest/minetest/blob/master/src/guiFormSpecMenu.cpp#L2177 22:24 sapier L2177 sets field named fname to text shown 22:28 BlockMen the question was more where is it for what propose used? 22:38 ShadowNinja This works: http://ix.io/bEL But I don't really know what I'm doing. hmmmm might be able to help. 22:38 hmmmm wtf 22:38 hmmmm shadowninja, stop 22:39 hmmmm don't mess with noise if it already works 22:39 hmmmm the comment is clearly wrong, all you're doing there is allocating a much, much, much larger array than is necessary 22:47 hmmmm that comment came from the original implementation of the optimized noise, when the spread factor was not the divisor, i changed that so people wouldn't have to pre-evaluate the result of 1/250 to put in the config file, for example 23:28 ShadowNinja hmmmm: But it doesn't work. It throws std::bad_aloc. 23:28 ShadowNinja +l