| Time |
Nick |
Message |
| 00:02 |
|
Fixer_ joined #minetest-dev |
| 00:08 |
|
ssieb joined #minetest-dev |
| 00:32 |
|
Kimapr joined #minetest-dev |
| 01:20 |
MTDiscord |
<exe_virus> Are there any gotchas I should be worried about when making my first PR to minetest? I checked and followed the coding guidelines, but are there other things to be aware of? |
| 01:29 |
|
Kimapr joined #minetest-dev |
| 02:09 |
|
Kimapr joined #minetest-dev |
| 02:41 |
MTDiscord |
<exe_virus> Oh well, that actually took a while to get all tidied up. Lots of small documentation things needing to be changed. But hopefully I hit them all, haha. |
| 02:41 |
|
Kimapr joined #minetest-dev |
| 03:05 |
|
Kimapr joined #minetest-dev |
| 04:00 |
|
MTDiscord joined #minetest-dev |
| 04:20 |
|
Kimapr joined #minetest-dev |
| 07:30 |
|
hlqkj joined #minetest-dev |
| 08:00 |
|
ShadowNinja joined #minetest-dev |
| 08:25 |
|
Kimapr joined #minetest-dev |
| 08:53 |
|
paxom joined #minetest-dev |
| 09:15 |
|
Fixer joined #minetest-dev |
| 09:18 |
|
Kimapr joined #minetest-dev |
| 09:47 |
|
calcul0n joined #minetest-dev |
| 09:50 |
|
Kimapr joined #minetest-dev |
| 10:09 |
MTDiscord |
<Kimapr> you can monkey patch the minetest.sound_play function to have it save the handle somewhere |
| 10:30 |
celeron55 |
has it been discussed whether there should be a more noob friendly process for building irrlichtmt on common linux distros or so? |
| 10:33 |
|
entuland joined #minetest-dev |
| 10:39 |
celeron55 |
currently the process is as so: 1) try to build minetest, make tells you to get https://github.com/minetest/irrlicht, 2) go there and download and build it separately, which is a relatively simple build process, (personally i name the local repo "irrlichtmt" though, as i also have the regular "irrlicht", i would consider at least recommending to name it so locally, 3) after building irrlichtmt |
| 10:39 |
celeron55 |
you're left with not a lot of instruction how to point the MT build into it but if you're good at ctrl+fing you'll find in the MT readme that you need to do something like cmake . -DIRRLICHT_LIBRARY=~/projects/irrlichtmt/lib/Linux/libIrrlichtMt.a -DIRRLICHT_INCLUDE_DIR=~/projects/irrlichtmt/include/, 4) you're done, the MT build works now as before |
| 10:39 |
celeron55 |
+) |
| 10:40 |
celeron55 |
i think this needs some smoothing |
| 10:42 |
celeron55 |
just mentioning -DIRRLICHT_LIBRARY and -DIRRLICHT_INCLUDE_DIR in the make message would go a long way, any opinions? |
| 10:48 |
celeron55 |
wouldn't be impossible to have an ./util/get_me_irrlichtmt.sh that does it all for you either but that seems like unnecessary maintenance burden |
| 10:54 |
MTDiscord |
<VanessaE> that seems fair (mentioning it in the make output) |
| 10:55 |
MTDiscord |
<VanessaE> on my first go 'round, that's where I got tripped-up, took some googling and a couple of tries to get the two paths right. |
| 10:56 |
MTDiscord |
<VanessaE> ("does it want.. a folder? no... it wants a file.. no, not this one -- ah, THAT one." etc) |
| 10:59 |
sfan5 |
sure that'd be useful |
| 11:00 |
sfan5 |
there's also the case of "neither irrlicht nor irrlichtmt installed", it'll just tell the user to go get irrlicht but without any details |
| 11:01 |
sfan5 |
it'd also make sense to add the same help text there is what I'm saying |
| 11:05 |
celeron55 |
also what about having a standard local directory for irrlichtmt where it'll be automatically found |
| 11:06 |
MTDiscord |
<VanessaE> good idea. |
| 11:06 |
celeron55 |
then the build process literally doesn't change at all, as long as you put it there |
| 11:10 |
sfan5 |
found and built or found if it's already built? |
| 11:11 |
celeron55 |
either way |
| 11:11 |
celeron55 |
i guess it could be built too |
| 11:12 |
celeron55 |
but that again is more maintenance |
| 11:13 |
celeron55 |
making it found on the other hand is just a matter of adding the standard path to cmake/Modules/FindIrrlicht.cmake |
| 11:14 |
celeron55 |
i think it isn't important to automate everything, it's just important to allow a user familiar with the command line but not with cmake to get the build done |
| 11:16 |
celeron55 |
doesn't matter if it's done using automation or using easy to find instructions |
| 11:18 |
celeron55 |
the obvious standard path i guess would be lib/irrlicht |
| 11:18 |
celeron55 |
which would then be added to .gitignore |
| 11:19 |
celeron55 |
and which could be a git submodule but again not important for my goal |
| 11:31 |
MTDiscord |
<exe_virus> I already made the PR, I implemented a new lua function, sound_stop_all() as this is not the first time I or someone else has desired this functionality: https://github.com/minetest/minetest/pull/11241 |
| 11:41 |
sfan5 |
stopping all sounds for all players on the server doesn't sound that useful |
| 11:55 |
|
proller joined #minetest-dev |
| 12:22 |
MTDiscord |
<appguru> ^ |
| 12:22 |
MTDiscord |
<appguru> it's quite easy to implement yourself |
| 12:52 |
MTDiscord |
<exe_virus> 1. No it's not often that such a thing is needed in minetest game or standard blocky games, but in a game with music that plays on loop, and different players having different sounds playing at once, it's good to have a global stop. 2. It's easy to implement yourself? Without adding a dependency to every mod in a game/server, how would you keep track of many different handles across them all? You could overwrite the sound play |
| 12:52 |
MTDiscord |
function to keep track of handles, but again this is not portable. 3. Simple example of why you might want to stop all sounds. Let's say you want to make a server announcement or periodic hunger games style cannon, you would want to stop all sounds before making such an announcement. Obviously ephemeral sounds are ignored in this case. |
| 13:13 |
|
proller joined #minetest-dev |
| 13:18 |
|
tech_exorcist joined #minetest-dev |
| 13:40 |
|
systwi joined #minetest-dev |
| 14:06 |
MTDiscord |
<appguru> re: 2. by overriding all sound play funcs to insert the handle into a table of all handles |
| 14:06 |
MTDiscord |
<appguru> I don't see what's not portable about this? |
| 14:12 |
MTDiscord |
<josiah_wi> sfan5 when you have time irrlicht/#35 should be really fast to review (adds verbose flag) and will help me debug the Zlib issue in my other PR. |
| 14:12 |
ShadowBot |
https://github.com/minetest/minetest/issues/35 -- Can't shutdown server by Ctrl+C in Ubuntu |
| 14:16 |
sfan5 |
you don't need to wait for me and can just push the change to your PR (and revert it later) |
| 14:16 |
sfan5 |
just fyi |
| 14:17 |
MTDiscord |
<josiah_wi> Did I not push the change? |
| 14:19 |
sfan5 |
i'm saying you can push the same commit to irrlicht#31 to debug it right now |
| 14:20 |
MTDiscord |
<josiah_wi> Oh I understand. If I have time tonight to work on it that's what I'll do. Also will shadowbot be getting IrrlichtMt compatibility? It's a little annoying to have it open the wrong link everytime I link an Irrlicht PR. |
| 14:20 |
sfan5 |
ask ShadowNinja about that |
| 14:20 |
sfan5 |
should be easy to add if he has a moment |
| 14:21 |
|
appguru joined #minetest-dev |
| 14:31 |
|
absurb joined #minetest-dev |
| 15:24 |
|
Kimapr joined #minetest-dev |
| 15:28 |
pgimeno |
I'll work on providing a fallback when I have some minutes |
| 15:36 |
pgimeno |
irr#7 |
| 15:36 |
pgimeno |
https://github.com/minetest/irrlicht/issues/7 -- Add EGL DRM backend |
| 15:36 |
pgimeno |
irrmt#7 |
| 15:45 |
|
Foz joined #minetest-dev |
| 18:31 |
|
silver_est joined #minetest-dev |
| 18:49 |
|
hlqkj joined #minetest-dev |
| 19:53 |
|
Kimapr joined #minetest-dev |
| 20:18 |
|
troller joined #minetest-dev |
| 20:20 |
|
proller__ joined #minetest-dev |
| 20:56 |
|
search_social joined #minetest-dev |
| 21:13 |
|
appguru joined #minetest-dev |
| 21:26 |
sfan5 |
https://forum.minetest.net/viewtopic.php?f=6&t=26669 https://forum.minetest.net/viewtopic.php?p=394479#p394479 |
| 21:26 |
sfan5 |
should've backported that mapgen change... |
| 21:27 |
sfan5 |
if anyone has some time please find out which mod/game messes with mapgen settings so we can remove it from contentdb |
| 21:56 |
MTDiscord |
<Jonathon> re mapgen settings, someone claimed hades redo caused it, however i never was able to verify that |
| 21:59 |
MTDiscord |
<exe_virus> Mostly if other people overwrite the sound_play function themselves as well. It starts getting hairy when you set that precedent in built_in where this PR makes use of stopping all sounds. We could make such a sound_stop_all() function purely out of lua and just override the default sound_play() function to support it. And I see no issues with such a plan, as long as that is what the core devs want. |
| 22:49 |
|
absurb_ joined #minetest-dev |
| 22:53 |
|
entuland joined #minetest-dev |
