| Time |
Nick |
Message |
| 01:53 |
|
Noisytoot joined #minetest-dev |
| 05:00 |
|
MTDiscord joined #minetest-dev |
| 06:34 |
|
YuGiOhJCJ joined #minetest-dev |
| 06:56 |
|
ivanbu joined #minetest-dev |
| 07:31 |
|
calcul0n joined #minetest-dev |
| 08:29 |
|
diceLibrarian joined #minetest-dev |
| 10:29 |
|
Niklp9 joined #minetest-dev |
| 11:41 |
sfan5 |
merging #14106, #14096 in 10m |
| 11:41 |
ShadowBot |
https://github.com/minetest/minetest/issues/14106 -- Update porting.h to fix build errors on MacOS 14 / Xcode 15 by superfloh247 |
| 11:41 |
ShadowBot |
https://github.com/minetest/minetest/issues/14096 -- [no squash] MapBlock structure optimizations by sfan5 |
| 12:22 |
|
appguru joined #minetest-dev |
| 12:40 |
MTDiscord |
<josiah_wi> Thank you for the review, grorp! |
| 12:41 |
MTDiscord |
<josiah_wi> I would like to contribute to fixing the Windows crash if I can, because I have a Windows machine. |
| 12:48 |
|
pmp-p joined #minetest-dev |
| 12:55 |
|
beanzilla joined #minetest-dev |
| 13:06 |
|
MTDiscord1 joined #minetest-dev |
| 15:27 |
|
Niklp9 joined #minetest-dev |
| 18:10 |
|
appguru joined #minetest-dev |
| 18:12 |
MTDiscord |
<josiah_wi> MSVC is issuing warnings about the <codecvt> header being deprecated (https://en.cppreference.com/w/cpp/header/codecvt). |
| 18:12 |
MTDiscord |
<josiah_wi> This is going to go away completely in C++26, without a replacement yet. |
| 18:13 |
|
appguru joined #minetest-dev |
| 18:20 |
MTDiscord |
<josiah_wi> Anyway, sfan5: https://github.com/minetest/irrlicht/pull/261. |
| 18:33 |
MTDiscord |
<paradust> this line is causing the win 32-bit crash: |
| 18:33 |
MTDiscord |
<paradust> https://github.com/minetest/irrlicht/blob/9b52d6fb0db1e491b1d2787ef33c5c74aebb3558/source/Irrlicht/mt_opengl_loader.cpp#L761 |
| 18:33 |
MTDiscord |
<paradust> it's absolutely obliterating the stack |
| 18:36 |
Krock |
@paradust what if you revert 8c521939b9a938 ? |
| 18:47 |
Krock |
or set a breakpoint before the .reserve() call |
| 18:53 |
MTDiscord |
<paradust> if I comment out line 759 to 776, it works |
| 18:54 |
MTDiscord |
<paradust> turns out the second part is also corrupting the stack, not as badly, but enough to cause another weird issue |
| 18:54 |
MTDiscord |
<paradust> i'll try reverting that, one sec |
| 18:56 |
Krock |
are you sure it's corrupting the stack? ContextManager resides in heap, as well as the unordered_set |
| 18:57 |
Krock |
oh. OpenGLProcedures GL; is in stack |
| 18:58 |
MTDiscord |
<paradust> reverting 8c521939b9a938 is not enough, it still crashes |
| 18:59 |
Krock |
same reason? if so, the corruption must've happened earlier |
| 18:59 |
MTDiscord |
<paradust> but if I comment out the remaining part of LoadAllProcedures, it works again |
| 19:00 |
MTDiscord |
<paradust> something might be wrong with with the codes for EXTENSIONS and NUM_EXTENSIONS, not sure. I couldn't find the OpenGL doc where it lists the code |
| 19:01 |
Krock |
grep GL_NUM_EXTENSIONS /usr/include/GL/*.h reveals that it's code 0x821D |
| 19:03 |
MTDiscord |
<paradust> Unfortunately I gotta sleep. Brain shutting off. But I'll be back in 11 hours or so. I'll take a look again if it is still a mystery |
| 19:05 |
Krock |
you could also try to reproduce the issue with the software OpenGL driver: https://fdossena.com/?p=mesa/index.frag |
| 19:05 |
Krock |
drop the opengl32.dll file next to minetest.exe |
| 21:51 |
|
Noisytoot joined #minetest-dev |
| 22:44 |
|
lhofhansl joined #minetest-dev |
| 22:46 |
lhofhansl |
Hi all. Planning to merge #14073 is a few. |
| 22:46 |
ShadowBot |
https://github.com/minetest/minetest/issues/14073 -- Allow cheaper culling checks at a distance by lhofhansl |
| 23:04 |
lhofhansl |
Done. |
| 23:33 |
|
panwolfram joined #minetest-dev |
