| Time |
Nick |
Message |
| 00:01 |
paramat |
thanks |
| 00:06 |
|
proller joined #minetest-dev |
| 00:24 |
|
YuGiOhJCJ joined #minetest-dev |
| 00:57 |
|
scr267 joined #minetest-dev |
| 00:57 |
scr267 |
hi, where can i find a channel about lua modding? |
| 00:58 |
rubenwardy |
scr267, #minetest |
| 00:59 |
scr267 |
ok thanks - hey i saw your github account today |
| 01:00 |
scr267 |
nice mods you did |
| 01:00 |
|
scr267 left #minetest-dev |
| 01:29 |
|
Megaf-Haiku joined #minetest-dev |
| 01:29 |
Megaf-Haiku |
Guys, can you please have a look at this? http://paste.debian.net/plain/985183 |
| 01:29 |
Megaf-Haiku |
Thats due to minetest using something that is not standard from something |
| 01:29 |
Megaf-Haiku |
it can be fixed |
| 01:40 |
|
Megaf-Haiku joined #minetest-dev |
| 01:48 |
Megaf-Haiku |
Or maybe I Haiku team can change Haiku |
| 01:51 |
|
Robby joined #minetest-dev |
| 04:35 |
|
paramat joined #minetest-dev |
| 04:46 |
|
proller joined #minetest-dev |
| 06:01 |
|
Robby joined #minetest-dev |
| 06:01 |
|
ssieb joined #minetest-dev |
| 06:09 |
|
nerzhul joined #minetest-dev |
| 06:59 |
|
Hunterz joined #minetest-dev |
| 07:10 |
|
proller joined #minetest-dev |
| 07:44 |
|
proller joined #minetest-dev |
| 08:20 |
|
lisac joined #minetest-dev |
| 09:05 |
|
Raven262 joined #minetest-dev |
| 09:22 |
|
DS-minetest joined #minetest-dev |
| 09:54 |
|
basicer joined #minetest-dev |
| 09:58 |
|
Raven262 joined #minetest-dev |
| 10:41 |
|
Hunterz joined #minetest-dev |
| 11:01 |
|
Krock joined #minetest-dev |
| 11:27 |
|
lisac joined #minetest-dev |
| 11:27 |
|
DS-minetest joined #minetest-dev |
| 11:28 |
|
wwwutz joined #minetest-dev |
| 11:34 |
|
wwwutz left #minetest-dev |
| 11:35 |
|
wwwutz joined #minetest-dev |
| 11:37 |
|
cx384 joined #minetest-dev |
| 12:02 |
|
lisac joined #minetest-dev |
| 12:09 |
Megaf |
Folks, I'm trying to build mt for windows here on linux, but I get this error. Can you please have a look? http://git.megaf.info/Megaf/GitPaste/raw/master/CompilingMTforWindowsOnLinuxError |
| 12:12 |
Krock |
old source code |
| 12:14 |
Krock |
define USE_CPP11_THREADS ? |
| 12:15 |
Megaf |
Krock, that happens with both 0.5 and 0.4 |
| 12:15 |
Krock |
there's no threads.h in the current master |
| 12:16 |
Krock |
it got removed in May (PR 4047) |
| 12:17 |
Megaf |
Ok, so how am I supposed to build 0.4? |
| 12:20 |
Krock |
try using the compiler flag -std=c++11 |
| 12:20 |
Megaf |
ok |
| 12:21 |
Krock |
however, it's strange that the compiler shows its c++11 support in its __cplusplus macro, but then doesn't know about std::thread. the header is included |
| 12:25 |
Megaf |
same error |
| 12:27 |
Megaf |
I'm using the buildbot by the way |
| 12:35 |
Megaf |
Krock, ideas? |
| 12:36 |
Krock |
git checkout -f |
| 12:37 |
Megaf |
Your branch is up-to-date with 'origin/stable-0.4'. |
| 12:37 |
Megaf |
gonna try with master just in case |
| 12:38 |
Krock |
stable-0.4 appears to be different than the 0.4.16 tag in master |
| 12:40 |
|
srifqi joined #minetest-dev |
| 12:40 |
Megaf |
similar error with master Krock |
| 12:41 |
Krock |
similar or the same? |
| 12:43 |
Megaf |
so, this is the error on stable-0.4 http://git.megaf.info/Megaf/GitPaste/src/master/Error_Compiling_Minetest_stable-0.4_for_Windows_On_Linux |
| 12:43 |
Megaf |
This is the error on master http://git.megaf.info/Megaf/GitPaste/src/master/Error_Compiling_Minetest_master_for_Windows_On_Linux |
| 12:44 |
Megaf |
error: ‘thread’ is not a member of ‘std’ |
| 12:44 |
Megaf |
weird thing is that both build fine for Linux |
| 12:45 |
Krock |
that's not the key of the question. there's a file called minetest/src/threads.h that doesn't exist in the recent master branch. have you tried to clear the cache yet? |
| 12:45 |
|
Fixer joined #minetest-dev |
| 12:46 |
Megaf |
../../src/mesh_generator_thread.h |
| 12:46 |
Megaf |
../../src/threading/thread.h |
| 12:46 |
Megaf |
../../src/util/thread.h |
| 12:46 |
Megaf |
those are the files on the 0.4 |
| 12:47 |
|
proller joined #minetest-dev |
| 12:47 |
Megaf |
Krock, its a fresh git clone to a directory that didnt exist |
| 12:47 |
Megaf |
./src/util/thread.h |
| 12:47 |
Megaf |
./src/threading/thread.h |
| 12:47 |
Megaf |
./src/mesh_generator_thread.h |
| 12:47 |
Megaf |
That on master |
| 12:48 |
Megaf |
so same files in both branches |
| 12:48 |
Krock |
any src/threads.h on master? |
| 12:49 |
Megaf |
https://github.com/minetest/minetest/blob/master/src/threading/thread.h |
| 12:49 |
Megaf |
all thread.h present are the ones I pasted |
| 12:49 |
Megaf |
no src/thread.h |
| 12:49 |
* Megaf |
will read the code |
| 12:51 |
Megaf |
this is the only part of the code that says src/thread.h https://github.com/minetest/minetest/blob/3cea7a349ac55df93b3eac1bf40365e00e472480/util/travis/clang-format-whitelist.txt#L324 |
| 12:53 |
|
nerzhul joined #minetest-dev |
| 12:54 |
Krock |
okay. then we're on the right path, at least. Now there's TravisCI and the gitlab pipeline, where the latter's Fedora setup works fine with GCC 6.3 |
| 12:54 |
nerzhul |
hi, what is the build problem ? |
| 12:55 |
Krock |
std::mutex and std::thread is not known in his GCC 6.3 compiler |
| 12:55 |
nerzhul |
wtf |
| 12:55 |
Megaf |
This file is indeed been created by cmake /dev/shm/Minetest_stable-0.4-Win64/minetest/src/threads.h |
| 12:55 |
nerzhul |
oh it's not GCC 6.3 it's mingw |
| 12:55 |
nerzhul |
GCC 4.8 supports that |
| 12:56 |
Megaf |
nerzhul, you got the backlog or want the link? |
| 12:56 |
nerzhul |
just use the buildbot script and stop trying non standard minetest provided builds |
| 12:56 |
nerzhul |
:) |
| 12:56 |
Megaf |
nerzhul, precisely what I'm doing |
| 12:56 |
nerzhul |
if travis and gitlab succeed you should just follow what they are doing :) |
| 12:56 |
Fixer |
i'm using msys2/mingw64 on windows with slightly modified buildbot, no problems |
| 12:57 |
Megaf |
nerzhul, the only change on my buildbot is that it doesnt build gettext nor luajit |
| 12:57 |
Megaf |
[13:43:09] <Megaf> so, this is the error on stable-0.4 http://git.megaf.info/Megaf/GitPaste/src/master/Error_Compiling_Minetest_stable-0.4_for_Windows_On_Linux |
| 12:57 |
Megaf |
[13:43:18] <Megaf> This is the error on master http://git.megaf.info/Megaf/GitPaste/src/master/Error_Compiling_Minetest_master_for_Windows_On_Linux |
| 12:57 |
Megaf |
those are the errors ^ |
| 12:57 |
nerzhul |
oh it's on stable 0.4 |
| 12:57 |
nerzhul |
then before C++11 |
| 12:57 |
Megaf |
nerzhul, stable and master |
| 12:57 |
Krock |
Megaf, how about purge mingw and do a fresh install? |
| 12:57 |
nerzhul |
but GCC 6 enabled -std=c++11 by default if i remember |
| 12:58 |
Megaf |
and I have the flag set in CXXFLAGS anyway |
| 12:58 |
Megaf |
coflags="-O2" |
| 12:58 |
Megaf |
cxxflags="$coflags -std=c++11" |
| 12:59 |
Megaf |
Krock, I can try that... |
| 13:01 |
Megaf |
redownloading it |
| 13:02 |
nerzhul |
to use pthread you should have another flag mentionned somewhere in buildbot |
| 13:02 |
nerzhul |
but just use the builbot script |
| 13:03 |
Megaf |
That's what I am doing |
| 13:03 |
Megaf |
can you try to reproduce the error at your end at least? |
| 13:04 |
Megaf |
using buildbot64 |
| 13:04 |
Megaf |
buildwin64* |
| 13:07 |
sfan5 |
Megaf: buildbot works fine, no errors |
| 13:07 |
sfan5 |
something is broken with your config |
| 13:07 |
nerzhul |
it works on my pc too |
| 13:07 |
sfan5 |
sound like non-c++11 headers with -std=c++11 |
| 13:16 |
Megaf |
will reinstall my whole compiler stack... |
| 13:18 |
Megaf |
[14:07:53] <sfan5> sound like non-c++11 headers with -std=c++11 |
| 13:18 |
sfan5 |
or just use the compiler the travis buildbot uses |
| 13:18 |
Megaf |
how can I check that? |
| 13:18 |
sfan5 |
dunno |
| 13:20 |
Megaf |
travis uses a different toolchain. ubuntu-toolchain-r-test |
| 13:21 |
Megaf |
Not sure if it has anything to do tho |
| 13:22 |
sfan5 |
no no |
| 13:23 |
sfan5 |
look into util/travis/before_install.sh the windows section |
| 13:25 |
Megaf |
it uses a different compiler? |
| 13:25 |
Megaf |
yep |
| 13:25 |
Megaf |
totally different compiler |
| 13:26 |
sfan5 |
ubuntu-toolchain-r-test is used for installing clang btw |
| 13:26 |
Megaf |
and it just extracts its to your /usr |
| 13:26 |
Megaf |
the script changes your system, not good |
| 13:26 |
sfan5 |
you can obviously use the toolchain standalone |
| 13:26 |
Megaf |
Will modify the buildbot to use travis compiler |
| 13:26 |
Megaf |
^ |
| 13:27 |
sfan5 |
lol ok |
| 13:27 |
sfan5 |
i was pointing you to what travis uses because that is known to work |
| 13:28 |
Megaf |
so it's not a problem in my compiler/system afterall |
| 13:29 |
Megaf |
someone just changed a compiler so it could build the buggy code! |
| 13:29 |
Megaf |
xP |
| 13:30 |
|
proller joined #minetest-dev |
| 13:30 |
Megaf |
I'm sure freeminer doesnt have this issue |
| 13:38 |
srifqi |
I was using buildbot64.sh, how to make it not reconfigure so that I don't have to compile ALL of them. |
| 13:43 |
Megaf |
hm, cmake totally ignoring -DCMAKE_CXX_COMPILER=path to compiler |
| 13:51 |
|
Wuzzy joined #minetest-dev |
| 13:54 |
Megaf |
can't you just make a code that works with standard compilers? =/ |
| 13:58 |
Megaf |
sfan5, cmake on hte buildbot is just totally ignoring both export CXX/CC and -DCMAKE_CXX_COMPILER compiler flags |
| 13:58 |
Megaf |
ah |
| 13:58 |
Megaf |
toolchain |
| 13:58 |
sfan5 |
*sigh* |
| 13:58 |
sfan5 |
it's using a toolchain file |
| 13:58 |
Megaf |
yep |
| 13:58 |
Megaf |
Just figured that out |
| 14:01 |
Megaf |
-- Check for working CXX compiler: /tmp/compiler/bin/x86_64-w64-mingw32-g++ |
| 14:01 |
Megaf |
-- Check for working CXX compiler: /tmp/compiler/bin/x86_64-w64-mingw32-g++ -- works |
| 14:01 |
Megaf |
finally... |
| 14:01 |
Megaf |
will it build tho? |
| 14:02 |
Megaf |
Nope |
| 14:02 |
Megaf |
very same error |
| 14:02 |
Megaf |
now you can't blame the compiler |
| 14:03 |
Megaf |
Errors like /dev/shm/Minetest-0.5-Win64/minetest/src/util/../threading/mutex_auto_lock.h:29:40: error: 'mutex' is not a member of 'std' |
| 14:03 |
Megaf |
using MutexAutoLock = std::unique_lock<std::mutex>; |
| 14:03 |
Megaf |
std::map<std::thread::id, std::string> m_thread_names; |
| 14:04 |
sfan5 |
ï½™ï½ï½•ï½’ ｃï½ï½ï½ï½‰ï½Œï½…ï½’ ｉｓ ｂｒï½ï½‹ï½…ｎ |
| 14:05 |
Krock |
^ |
| 14:05 |
Megaf |
sfan5, I'm using THIS COMPILER => http://minetest.kitsunemimi.pw/mingw_w64_x86_64_ubuntu12.04_4.9.1.7z |
| 14:05 |
sfan5 |
>4.9.1 |
| 14:05 |
sfan5 |
that's outdated |
| 14:05 |
Megaf |
FUCK YOU |
| 14:05 |
sfan5 |
i should delete that from my server actually |
| 14:05 |
Megaf |
fuck |
| 14:05 |
sfan5 |
thanks for reminding me |
| 14:06 |
Megaf |
that's the fucking fuck compiler on the fucking scripts in the fuck git |
| 14:06 |
Megaf |
that you told me to use |
| 14:06 |
* dzho |
takes a deep breath |
| 14:06 |
sfan5 |
stable branch yea |
| 14:07 |
sfan5 |
https://github.com/minetest/minetest/blob/master/util/travis/before_install.sh#L22 |
| 14:08 |
Megaf |
this thing is way too overcoplicated |
| 14:08 |
Megaf |
I can't handle that |
| 14:08 |
sfan5 |
maybe you are just too stupid 🤔 |
| 14:08 |
Megaf |
I just wanted a piece of code that worked |
| 14:08 |
nerzhul |
please be more polite, and cross compile is not easy |
| 14:09 |
Megaf |
nerzhul, it was easy, I used to make windows builds several months ago |
| 14:09 |
Megaf |
it just worked |
| 14:09 |
dzho |
\o/ |
| 14:09 |
sfan5 |
Megaf: the current master scripts work for compiling both the stable release and -dev |
| 14:09 |
nerzhul |
build has not changed in stable 0.4 |
| 14:09 |
sfan5 |
you are trying to compile 0.5-dev with the compiler used by 0.4.16 |
| 14:09 |
sfan5 |
because of the c++11 changes this does not work |
| 14:09 |
sfan5 |
simple as that |
| 14:10 |
dzho |
all that is good and wonderful to those who debug and reproduce build methods: <3 |
| 14:10 |
Megaf |
And I do have a strong lack of concentration and very poor organizational abilities because my brain, if you are not clear and direct I will not understand |
| 14:11 |
Megaf |
I didnt know I needed a whole different toolchain for 0.4 and 0.5 |
| 14:11 |
sfan5 |
you know now |
| 14:11 |
Megaf |
and didnt know none of the versions work with standard compilers |
| 14:11 |
Megaf |
its finally building this thing now |
| 14:11 |
nerzhul |
all are standard compilers, but versions can change |
| 14:11 |
sfan5 |
they do work with "standard" compilers (whatever those are) if they are not too old |
| 14:15 |
dzho |
Megaf: so what is it, the 14.04 compiler that is working, then? |
| 14:15 |
|
Robby joined #minetest-dev |
| 14:16 |
Megaf |
dzho, this compiler is working for stable-0.4, that's the version I wanted to build. http://minetest.kitsunemimi.pw/mingw_w64_x86_64_ubuntu12.04_4.9.1.7z |
| 14:16 |
Megaf |
version 4.9... that's old as hell |
| 14:17 |
Megaf |
well, I have no idea actually, don't care anymore |
| 14:17 |
Megaf |
x86_64-w64-mingw32-gcc (GCC) 4.9.1 |
| 14:18 |
Megaf |
I had version 6.3.0 |
| 14:19 |
Megaf |
so sfan5 nerzhul dzho !!! I t ' s n o t m y c o m p i l e r t h a t i s b r o k e n o r o u t d a t e d !!! |
| 14:19 |
sfan5 |
yes it is |
| 14:20 |
Megaf |
how come version 6.3.0 is older than 4.9.1? |
| 14:21 |
sfan5 |
what |
| 14:21 |
nerzhul |
maybe it's our head, and strop crying |
| 14:22 |
Megaf |
you are using a compiler for ubuntu 12.04, that reached end of life on april of this year |
| 14:22 |
* Megaf |
inhales deeply and slowly exhales. |
| 14:24 |
Megaf |
I just want you to stop blaming my compiler/system when is your code and your choice of compiler that is outdated... |
| 14:24 |
Megaf |
your compiler/choice https://packages.ubuntu.com/trusty/mingw-w64 the one I have https://packages.debian.org/stretch/mingw-w64 |
| 14:24 |
Megaf |
bye |
| 14:27 |
nerzhul |
... |
| 14:34 |
srifqi |
I even need to use WSL to compile for Windows. |
| 14:49 |
|
Megaf joined #minetest-dev |
| 14:50 |
Megaf |
sfan5, did you just remove http://minetest.kitsunemimi.pw/mingw_w64_x86_64_ubuntu12.04_4.9.1.7z from your server? |
| 14:55 |
Megaf |
I was working on a new restructured version of the buildbot... http://git.megaf.info/Megaf/GitPaste/src/master/NewBuildWinBot64.sh |
| 15:25 |
dzho |
you know, this is the sort of thing I like about, eg, the ports collection approach (or, to those who might be more familiar, portage or AUR)--its meant to be used with upstream, instead of forking an evanscently-relevant single version and distributing that |
| 15:28 |
Megaf |
dzho, My biggest issue is people blaming the software I use, they assume, it's Debian, so it must be outdated. |
| 15:28 |
Megaf |
They could have just said, you need to use the version x.y.x of the compiler, the code is not compatible with other versions. |
| 15:29 |
Megaf |
Done, problem solved |
| 15:29 |
dzho |
Megaf: yeah, I get that. |
| 15:29 |
Megaf |
scroll back and see how many times they complained me or my compiler |
| 15:30 |
Megaf |
My compiler is fine, I can build plib, luajit, irrlich, opensnecegraph, the linux kernel, and my own software |
| 15:31 |
dzho |
I'm just saying, I guess, that a ports file or ebuild or whatever is almost essentially self-documenting configuration rather than that thing we saw up above where "oh yeah you have to use this exact thing from this exact custom URL that leads to the developer's personal web site" |
| 15:32 |
nore |
#6398 is a trivial maintenance pr |
| 15:32 |
ShadowBot |
https://github.com/minetest/minetest/issues/6398 -- Cleanup & document better conversions between block position and integers by Ekdohibs |
| 15:32 |
dzho |
"my compiler is fine" a) casts things in a not-that-helpful personal light, because it's not *your* compiler. It's one you happen to be using for certain tasks, for which tasks, yes, it is fine. but b) it's not bug-compatible with whatever it is that's being targeted, apparently. |
| 15:34 |
dzho |
this is a tricky thing over time because generally one builds with a certain set of things and has no way of knowing that *future* releases of those things will not work, at least, not until they are tried. |
| 15:34 |
dzho |
at which point most have moved on to newer releases of both the target code base and the build tools. |
| 15:35 |
dzho |
this sort of skew *is* maddening, understandably, but it's also understandable that it happens and how it happens and how folks tend not to care too much about it. |
| 15:38 |
Megaf |
All easily solved we we stick to the latest stable version of ISO C |
| 15:40 |
Megaf |
all latest stable versions of C/C++ compilers should be compatible with all code that follow latest C/C++ version, am I wrong? |
| 15:40 |
Megaf |
and I mean ansi C |
| 15:40 |
Megaf |
This https://en.wikipedia.org/wiki/C11_(C_standard_revision) |
| 15:41 |
Megaf |
and of course the same for C++... |
| 15:45 |
Megaf |
also, nowhere in the code is defined the C version the software supports. __STDC_VERSION__ |
| 15:45 |
Megaf |
again. It's not my computer/emotion/software problem. Code is bad |
| 15:46 |
Megaf |
I suggest chose a C/C++ version and stick to it, properly defining it in the code. |
| 15:46 |
nerzhul |
it's already the case. |
| 15:47 |
nerzhul |
0.4.16 = c++03, 0.5.X = c++11 |
| 15:47 |
nerzhul |
now fix your build system or use official builds. |
| 15:52 |
Megaf |
nerzhul, I dont know if my build system is broken. I came here asking for help, and the only one who tried to help was Krock |
| 15:56 |
nerzhul |
we told you to use buildbot script from 0.5 branch in stable 0 |
| 15:56 |
nerzhul |
0.4 |
| 15:58 |
Megaf |
define buildbot |
| 16:02 |
sfan5 |
Megaf: i did, because it was outdated and is only used by travis for 0.4.16 (which isn't needed anymore) |
| 16:02 |
sfan5 |
just use the 7.1.1 build the 0.5 branch uses |
| 16:04 |
Megaf |
as buildbot I understand this https://github.com/minetest/minetest/tree/master/util/buildbot |
| 16:04 |
Megaf |
What do you mean by buildbot? |
| 16:13 |
Megaf |
well, trying with 7.1.1 |
| 16:36 |
Megaf |
sfan5, 7.1.1 from Ubuntu indeed works well here |
| 16:42 |
Megaf |
sfan5, why don't we use clang instead? |
| 16:42 |
* Megaf |
hides |
| 16:46 |
sfan5 |
i haven't seen any (mature) clang -> windows cross-compiling solutions |
| 16:46 |
sfan5 |
Megaf: the buildbot itself does not force a specific compiler, but when travis builds for windows it will download a specific compiler and tell the buildbot to use thatr |
| 16:48 |
Megaf |
I don't understand how clang works on Debian. I know there's an aditional package I have to download to make it work |
| 16:48 |
Megaf |
without it cmake just fails when it tests if the compiler works |
| 16:55 |
|
Natechip joined #minetest-dev |
| 16:55 |
|
Natechip joined #minetest-dev |
| 16:56 |
|
kaeza joined #minetest-dev |
| 17:18 |
Megaf |
Krock, would you be a victim of my build? (That is, can you test it on a windows machine please?) |
| 17:18 |
Megaf |
it ships the minimal subgame |
| 17:18 |
Megaf |
actually minetest_game is coming to |
| 17:23 |
|
paramat joined #minetest-dev |
| 17:25 |
Krock |
Megaf, no sorry, you'll have to search another victim |
| 17:25 |
Krock |
my Win XP setup is 1) not representative if it doesn't work 2) not mounted with a mainboard anymore |
| 17:40 |
|
Raven262 joined #minetest-dev |
| 17:51 |
|
DS-minetest joined #minetest-dev |
| 18:00 |
|
lisac joined #minetest-dev |
| 18:01 |
|
Raven262 joined #minetest-dev |
| 18:06 |
srifqi |
paramat, can you please check #6399 ? |
| 18:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/6399 -- Change the settings interface for noise_params, noise_params_group, flags, and v3f by srifqi |
| 18:06 |
paramat |
yes i'll review later, thanks |
| 18:32 |
|
srifqi joined #minetest-dev |
| 18:33 |
srifqi |
paramat, I can't use that function. Both of them throws an error because they are nil. |
| 18:55 |
paramat |
ok they must be for getting during playing |
| 19:22 |
|
rubenwardy left #minetest-dev |
| 19:22 |
|
rubenwardy joined #minetest-dev |
| 19:29 |
|
lisac joined #minetest-dev |
| 19:30 |
|
Player_2 joined #minetest-dev |
| 19:34 |
rubenwardy |
https://github.com/rubenwardy/minetest/commit/615d00928a9ada87fc948f3d4b0ea57ca5620239 |
| 19:34 |
rubenwardy |
sfan5, Krock ^ |
| 19:34 |
Krock |
+1 |
| 19:34 |
rubenwardy |
nice, will merge in 10 |
| 19:34 |
sfan5 |
find by me, tho maybe add a note that distros/people building from src should use the 0.4.16 tag until 0.4.17 is actually released |
| 19:34 |
sfan5 |
fine* |
| 19:35 |
rubenwardy |
will do |
| 19:37 |
rubenwardy |
one of the reasons I prefer a dev-0.4 is that we can have a PR to discuss which fixes should be merged |
| 19:37 |
rubenwardy |
I guess that could be done with an issue |
| 19:41 |
rubenwardy |
ok, merging... |
| 19:43 |
rubenwardy |
done |
| 19:44 |
Megaf |
Folks, so I'm working with the Haiku OS people, to port Minetest to Haiku |
| 19:44 |
rubenwardy |
!g haiku os |
| 19:44 |
ShadowBot |
rubenwardy: Home | Haiku Project - Haiku is an open-source operating system that specifically targets personal computing. Inspired by BeOS, Haiku is fast and easy to learn but very powerful. </l/?kh=-1&uddg=https%3A%2F%2Fwww.haiku%2Dos.org%2F>, Haiku (operating system) - Wikipedia - Haiku is a free and open-source operating system compatible with the now discontinued BeOS. Its development (6 more messages) |
| 19:45 |
rubenwardy |
www.haikuos.org |
| 19:45 |
rubenwardy |
www.haiku-os.org |
| 19:46 |
|
behalebabo joined #minetest-dev |
| 19:46 |
Megaf |
And they did their part in inplementing stuff into the system so it could run Minetest, now is our turn |
| 19:46 |
Megaf |
more details there -> http://git.megaf.info/Megaf/GitPaste/src/master/Porting_Minetest_To_Haiku |
| 19:52 |
rubenwardy |
Sounds interesting |
| 19:54 |
Megaf |
Haiku is a whole "new" Operating System |
| 19:54 |
Megaf |
is not Linux, UNIX, Windows or anything |
| 19:54 |
Megaf |
it's built from the ground up |
| 19:54 |
Krock |
we need to support TempleOS first |
| 19:55 |
Megaf |
and it's almost POSIX |
| 19:55 |
Megaf |
TempleOS is no POSIX |
| 19:55 |
Megaf |
well, nor is Windows... |
| 19:55 |
Megaf |
Windows NT/2000 were almost POSIX |
| 19:55 |
Megaf |
we had a whole layer we could install |
| 19:55 |
rubenwardy |
If you make a PR for it, I'm sure it will be accepted. Well, as long as it doesn't add tons of #ifdef everywhere |
| 19:55 |
Megaf |
called Microsoft Services for UNIX I think |
| 19:56 |
Megaf |
rubenwardy, I could, but I have no idea what to do |
| 19:56 |
Megaf |
I'm no dev |
| 19:58 |
celeron55_ |
we don't use haiku, thus we can't test, thus we won't make a PR or push a commit |
| 19:58 |
celeron55_ |
someone who uses it should make a PR |
| 19:58 |
rubenwardy |
^ |
| 19:58 |
celeron55_ |
that's how it always works |
| 20:22 |
Megaf |
celeron55_, That's sad, but makes sense and it's fair enough |
| 20:23 |
Megaf |
Release is -O2 or -O3? |
| 20:24 |
celeron55_ |
i don't think it's that sad; if the userbase of an entire operating system, no less, is too small to have one interested developer, then it's not something that needs or even should happen |
| 20:25 |
celeron55_ |
anyway, if the OS continues to gain interested, then some day there will be a dev |
| 20:25 |
celeron55_ |
interest* |
| 20:27 |
sfan5 |
does haiku even have 3d accel? |
| 20:28 |
Megaf |
radeon drivers, bluettoth, wireless and all |
| 20:28 |
Megaf |
cpu scalling |
| 20:28 |
Megaf |
it already has irrlicht, luajit and sqlite in its repos |
| 20:33 |
celeron55_ |
what's the benefit as a user? |
| 20:33 |
celeron55_ |
compared to other OSes |
| 20:35 |
rubenwardy |
well, currently you can't be distracted by playing Minetest |
| 20:35 |
Megaf |
celeron55_, It's designed for Desktop and Multimedia production |
| 20:35 |
Megaf |
from the ground up |
| 20:36 |
Megaf |
from it's kernel to the userland and GUI |
| 20:36 |
Megaf |
It it's a tiny OS, yet fully featured, extreme low latency |
| 20:36 |
celeron55_ |
does someone use it for that purpose? |
| 20:36 |
Megaf |
and astonishing SMP and multitasking capabilities |
| 20:37 |
celeron55_ |
does it have a better audio subsystem than linux? |
| 20:37 |
rubenwardy |
if it doesn't support photoshop or after affects, then no |
| 20:37 |
Megaf |
to begin with it doesnt depends on X server and this stuff, all is in the kernel |
| 20:37 |
Megaf |
I had crashes on it, kernel panics, and recovered from them by typing resume |
| 20:38 |
sfan5 |
what i've heard until now is "design is different from linux" and buzzwords |
| 20:38 |
Megaf |
I don't think anyone uses it for multimedia now, not in production, because the OS still in Alpha |
| 20:38 |
Megaf |
but BeOS, the father of Haiku, was used for sure |
| 20:38 |
Megaf |
too bad it was too expensive and PowerPC only at the time |
| 20:39 |
Megaf |
Haiku in my opinion is the best of Windows and Linux |
| 20:42 |
|
Megaf_ joined #minetest-dev |
| 20:48 |
Megaf |
Haiku has a pkgmaneger too |
| 20:55 |
Megaf |
Guys, so this is what happens when I try to build minetest server on Haiku http://paste.debian.net/plain/985308 |
| 20:56 |
rubenwardy |
that's a function in endian.h as part of the linux includes |
| 20:56 |
Megaf |
thats a BSD include thing, that was missing in Haiku and now was added because minetest |
| 20:57 |
Megaf |
now we just have to do the adaptation I pasted here earlier for it to work |
| 20:57 |
sfan5 |
you need to add -D_BSD_SOURCE to your flags like the guy said |
| 21:00 |
Megaf |
nope, not that simple |
| 21:01 |
sfan5 |
well then he was wrong |
| 21:01 |
Megaf |
that flag doest exist |
| 21:02 |
sfan5 |
says who? |
| 21:04 |
rubenwardy |
I think he means that you should add a cmake flag which does #define BSD_SOURCE |
| 21:05 |
sfan5 |
_BSD_SOURCE |
| 21:05 |
rubenwardy |
ah yes |
| 21:05 |
rubenwardy |
#ifdef __HAIKU__ / #define _BSD_SOURCE / #endif / #include <endian.h> |
| 21:05 |
Megaf |
my question is, where to add that |
| 21:06 |
rubenwardy |
above any endian.h includes |
| 21:06 |
sfan5 |
"hello devs i am using a niche platform please do all work for me" |
| 21:06 |
sfan5 |
-DCMAKE_CXX_FLAGS="-D_BSD_SOURCE" should also work |
| 21:07 |
Megaf |
sfan5, Linux is not exactly popular, is it? |
| 21:07 |
Megaf |
nor is Minetest |
| 21:07 |
sfan5 |
fuck off |
| 21:07 |
Megaf |
Please mind your language |
| 21:07 |
Megaf |
lol |
| 21:08 |
rubenwardy |
Linux is very popular |
| 21:08 |
Megaf |
I will go back to Linux to add the define things soon |
| 21:08 |
Megaf |
not on desktops |
| 21:08 |
rubenwardy |
2% is a large number |
| 21:08 |
rubenwardy |
even 0.1% is |
| 21:08 |
rubenwardy |
the 100 people that use Haiku isn't |
| 21:09 |
Megaf |
and dont get me wrong, Im a huge Linux fan, its my main OS since 2003 and co existed with Windows way before that |
| 21:09 |
Megaf |
not sure if theres 100 people using Haiku xP |
| 21:09 |
rubenwardy |
it's also worth noticing that 98% of our users run on the Linux kernel |
| 21:10 |
Megaf |
99% of those are on Android |
| 21:10 |
Megaf |
0.5% on ChromeOS |
| 21:11 |
sfan5 |
mt works on chrome os? |
| 21:12 |
sofar |
theoretically |
| 21:12 |
sofar |
I don't think anyone has a working port |
| 21:13 |
sofar |
I think it's probably a chromebook with linux running on it, more likely |
| 21:13 |
sfan5 |
there's the android "bridge" tho, so it might work on ARM ones |
| 21:14 |
Megaf |
well, chromeos do run android apps |
| 21:19 |
Megaf |
So it turns out there is a dev interesed, but still not working https://github.com/jessicah/minetest/commit/13cd79e75af3a11c5d4b9293d68c6ed50d94ddcc |
| 21:19 |
Megaf |
[21:06] <sfan5> -DCMAKE_CXX_FLAGS="-D_BSD_SOURCE" should also work |
| 21:20 |
Megaf |
that didnt work, nor in Minetest nor with her patch |
| 21:20 |
sfan5 |
did you update your system to actually include the changes made by the haiku dev? |
| 21:21 |
Megaf |
I think I have to reboot |
| 21:23 |
|
Megaf joined #minetest-dev |
| 21:27 |
Megaf |
compiling now, after a reboot and a second update |
| 21:29 |
|
DI3HARD139 joined #minetest-dev |
| 21:30 |
|
Megaf_ joined #minetest-dev |
| 21:59 |
Megaf |
[ 47%] Building CXX object src/CMakeFiles/minetestserver.dir/util/string.cpp.o |
| 21:59 |
Megaf |
/boot/home/minetest/src/util/string.cpp:35:20: fatal error: iconv.h: No such file or directory |
| 21:59 |
Megaf |
compilation terminated. |
| 22:24 |
jcalve |
Looks like you're missing iconv.h mate |
| 22:51 |
|
paramat joined #minetest-dev |
| 23:15 |
|
Player_2 joined #minetest-dev |
| 23:42 |
Megaf_ |
jcalve, precisely, it does look like that |
| 23:44 |
Megaf_ |
I will try to fix that tomorrow |
