| Time |
Nick |
Message |
| 04:00 |
|
MTDiscord joined #luanti-dev |
| 04:56 |
|
Alias joined #luanti-dev |
| 05:51 |
|
SFENCE joined #luanti-dev |
| 08:06 |
|
citrons_ joined #luanti-dev |
| 08:37 |
|
SFENCE joined #luanti-dev |
| 09:23 |
|
YuGiOhJCJ joined #luanti-dev |
| 09:53 |
pgimeno |
isn't #11523 closed by 17036 mentioned above? |
| 09:53 |
ShadowBot |
https://github.com/luanti-org/luanti/issues/11523 -- If player receives a formspec while in the death screen, the death screen disappears, breaking the game |
| 09:57 |
pgimeno |
ah nvm, 17036 is not merged yet |
| 10:19 |
|
Desour joined #luanti-dev |
| 11:15 |
Krock |
will merge #17022 in 15 minutes |
| 11:15 |
ShadowBot |
https://github.com/luanti-org/luanti/issues/17022 -- [NO SQUASH] Break include chains and tidy Part 3 by SmallJoker |
| 14:13 |
Desour |
just out of interest: what does it mean when luanti shows a grey screen (usually observable on a segfault), and what if it is green? are we using grey (and sometimes green) as clear color? |
| 14:15 |
Krock |
depends on the background color drawn after beginScene |
| 14:16 |
Krock |
> this->sky->getSkyColor(); |
| 14:17 |
Krock |
if a mod changes the sky color, it might be possible that you're stuck with a purple screen |
| 14:18 |
Desour |
I see. thx! ^^ |
| 14:18 |
Krock |
at least the main menu default background is #8CBAFA |
| 14:18 |
Krock |
(light blue) |
| 14:19 |
Desour |
yeah, I wonder where I get the green from. maybe something devtest specific |
| 14:20 |
Krock |
if the debugger trips while rendering, then you might as well have corrupted vertex data, though that might look different than plain green |
| 14:21 |
Desour |
it also was always the same green |
| 14:21 |
Krock |
you could use the opengl3 renderer and start it with renderdoc, get a capture and find out when exactly the frame buffer is that way... |
| 14:22 |
Krock |
though I don't think you'd actually see that due to buffering |
| 14:22 |
Desour |
hm, builtin supports setting sky color via a menu_theme setting (TIL), but devtest doesn't set that |
| 14:29 |
Desour |
I've made a breakpoint at GUIEngine::setMenuSkyColor, and observed a color of 0x7bff (`p/x color.color`) but that's not green |
| 14:42 |
sfan5 |
if a shader somehow totally messes up you could get some random color too |
| 14:46 |
Desour |
maybe it just wasn't random because it was in a build with sanitizers |
| 15:29 |
Krock |
Unless there are objections, I'd like to push this trivial fix in 20-30 minutes: https://pastebin.com/raw/MfWUU3CE (removed some duplicated code) |
| 15:30 |
Krock |
can be tested with devtest: //lua S = core.get_translator("testtranslations"); core.kick_player("singleplayer", S("Testing .tr files: untranslated"), false) |
| 16:30 |
sofar |
seems more than reasonable |
| 18:26 |
[MatrxMT]_ |
<Zughy> yo what the fuck, 10 new PRs in 24 hours? |
| 18:39 |
Krock |
the power of AI, partially. |
| 18:41 |
sofar |
sorry |
| 18:41 |
sofar |
it was this or do outside yard work. |
| 22:36 |
MTDiscord |
<greenxenith> I opted for outside yard work |
| 23:14 |
[MatrxMT]_ |
<Zughy> too late, another PR |
| 23:14 |
[MatrxMT]_ |
<Zughy> I don't know, it seems the best way to overwhelm people and make them quit |
| 23:15 |
sofar |
there's no rush for any of this though |
| 23:15 |
sofar |
8 year old feature request. plus there's a different PR with something similar. |
| 23:16 |
rubenwardy |
I mean for me it depends on whether the PRs are of appropriate quality or if they're just generated and untested |
| 23:16 |
rubenwardy |
like generating so quickly doesn't leave much time for polishing up |
| 23:16 |
[MatrxMT]_ |
<Zughy> sure but if you keep publishing PRs faster than I can label them, I can't even imagine how core devs can feel to the idea of testing and reviewing them |
| 23:17 |
sofar |
the only way I can make these is with extensive testing (see screenshots, there's 100+ nodes in the background used for testing each step needed. BTW the node text thing was maybe 6 hours of my time) |
| 23:18 |
[MatrxMT]_ |
<Zughy> you didn't even link the issue :\ |
| 23:18 |
MTDiscord |
<wsor4035> >we have to many issues prs made >no, not like that |
| 23:20 |
rubenwardy |
also sofar, fyi as a previous core dev you have the option of just asking c55 to be readded if you plan on being active now |
| 23:24 |
sofar |
Zughy: I stumbled upon it only after most of this PR was done, not sure whether referencing is a good idea now. Perhaps the other PR needs prio first - I'm admitting that it's not making it easier. |
| 23:25 |
sofar |
rubenwardy: I'm fine being "allowed in" as such. I've stopped doing useless stuff and my natural pull to contribute back is what is driving me right now. ITB is coming up on 10 years very soon. |
| 23:25 |
sofar |
Last time I did this much nore was a spindly student. Now look where they are. |
| 23:26 |
sofar |
;) |
| 23:33 |
|
panwolfram joined #luanti-dev |
| 23:34 |
sofar |
actually the word I was looking for wasn't "allowed in" - "tolerated" is what I wanted to say. |
| 23:58 |
sofar |
FYI I strongly encourage everyone here to provide direction if possible. If some of this energy I have right now is better used in different places then I'm happy to segway. e.g. I built out my network client packet fuzzer for the one out of bounds fix. I could maybe screen every single value (not easy, but, probably worth while doing anyway). |
