| Time |
Nick |
Message |
| 00:43 |
|
reductum joined #minetest-dev |
| 01:17 |
|
fluxflux joined #minetest-dev |
| 01:34 |
|
nephele joined #minetest-dev |
| 03:19 |
|
Kimapr joined #minetest-dev |
| 04:46 |
|
QwertyDragon joined #minetest-dev |
| 04:56 |
|
QwertyDragon joined #minetest-dev |
| 05:18 |
|
QwertyDragon joined #minetest-dev |
| 05:25 |
|
QwertyDragon_ joined #minetest-dev |
| 06:01 |
|
calcul0n joined #minetest-dev |
| 06:08 |
|
reductum joined #minetest-dev |
| 06:38 |
|
calcul0n_ joined #minetest-dev |
| 08:25 |
|
Icedream joined #minetest-dev |
| 08:39 |
|
ShadowNinja joined #minetest-dev |
| 09:12 |
|
absurb joined #minetest-dev |
| 09:19 |
|
kilbith joined #minetest-dev |
| 09:23 |
kilbith |
https://github.com/minetest/minetest/issues/9557#issuecomment-607979943 |
| 09:23 |
kilbith |
sfan5, so why we should deprive from small fixes during the feature freeze? |
| 09:24 |
kilbith |
and why we should ship out a broken feature for 5.2? |
| 09:24 |
kilbith |
add this to the milestone as well for https://github.com/minetest/minetest/pull/9583 |
| 09:27 |
kilbith |
it's exactly like walking in the street and not bending down to pick up $2 deliberately |
| 09:31 |
|
AntumDeluge joined #minetest-dev |
| 09:31 |
kilbith |
your weak-ass PR management is so bad that you have occulted trivial fixes like that: https://github.com/minetest/minetest/pull/9319 |
| 09:32 |
kilbith |
or that: https://github.com/minetest/minetest/pull/9058 |
| 09:49 |
sfan5 |
ah yes great style, leaving a rant and leaving immediately |
| 09:50 |
sfan5 |
~tell kilbith if you had looked at the timestamps, you would have noticed that my comment was posted 3 minutes before the PR was created |
| 09:50 |
ShadowBot |
sfan5: O.K. |
| 09:53 |
sfan5 |
...and the reason why such trivial PRs are often left lying around is that barely any of the devs has available time right now |
| 10:28 |
|
Fixer joined #minetest-dev |
| 10:38 |
|
proller joined #minetest-dev |
| 11:05 |
sfan5 |
merging game#2631 |
| 11:05 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2631 -- Fix respawn not working at all when suitable position was not found by sfan5 |
| 11:31 |
|
Beton joined #minetest-dev |
| 11:47 |
|
Fixer joined #minetest-dev |
| 11:48 |
|
karamel joined #minetest-dev |
| 12:06 |
|
mizux joined #minetest-dev |
| 12:19 |
|
Gael-de-Sailly joined #minetest-dev |
| 12:25 |
|
turtleman joined #minetest-dev |
| 12:57 |
Krock |
sfan5: yes, looks good |
| 12:58 |
Krock |
opening a PR for that collision stuff |
| 13:08 |
|
kilbith joined #minetest-dev |
| 13:09 |
kilbith |
still not added to the milestone... |
| 13:11 |
Krock |
#9591 |
| 13:11 |
ShadowBot |
https://github.com/minetest/minetest/issues/9591 -- Revert collision tweaks #9365 and #9327 by SmallJoker |
| 13:11 |
Krock |
kilbith: where? |
| 13:11 |
kilbith |
sfan5, I haven't seen you paying attention to the formspec PRs at all but marking hypertext as unstable in the docs, that you can do with enthusiasm |
| 13:12 |
kilbith |
great style |
| 13:12 |
sfan5 |
not sure where you are reading enthusiasm into my actions |
| 13:12 |
kilbith |
that thumb up to this PR (marking unstable) and merging |
| 13:13 |
kilbith |
and btw this "no avalaible time" is a BS argument |
| 13:13 |
kilbith |
first off, we are all locked up |
| 13:13 |
sfan5 |
yes, I agreed with that PR |
| 13:13 |
kilbith |
and you have time to create new PRs |
| 13:15 |
kilbith |
I find it quite paradoxical to let it mark stuff as "unstable" while not reviewing small fixes aside... this is pure laziness |
| 13:15 |
kilbith |
you are probably too buzy jerking off in front of your hentai |
| 13:15 |
kilbith |
hmmm? |
| 13:16 |
sfan5 |
I rarely go through my notifications backlog to look at PRs that were posted weeks before, so that should explain why the others are left lying around |
| 13:17 |
kilbith |
Krock, https://github.com/minetest/minetest/issues/9573 and https://github.com/minetest/minetest/pull/9583 |
| 13:17 |
sfan5 |
I would too like to know why everyone else doesn't have much time considering corona is currently a thing, but I can't look into other people's reallife |
| 13:17 |
kilbith |
and https://github.com/minetest/minetest/pull/9319 |
| 13:17 |
Krock |
kilbith: not game breaking, thus no milestone |
| 13:17 |
Krock |
although it's trivial enough to be merged |
| 13:17 |
sfan5 |
regarding you last comment, I cannot think of a reason why you'd think such comments are productive in any way |
| 13:18 |
sfan5 |
in fact you had been banned before for making inappropriate comments at coredevs, it should be in your interest not to repeat that |
| 13:18 |
kilbith |
I don't care sfan5 |
| 13:18 |
kilbith |
it's not of my interest to do that |
| 13:18 |
sfan5 |
I get that you're upset at easy fixes not being merged, but we cannot personally care for hypertext just because you are commercially developing it |
| 13:19 |
kilbith |
it's for the people who actually play the game |
| 13:19 |
sfan5 |
every PR receives equal amount of dev attention, often enough that is none |
| 13:19 |
sfan5 |
if you cared about the general benefit of the Minetest community, you would also know better than to start shit with devs |
| 13:20 |
Krock |
can you please stay civilized in here? |
| 13:22 |
sfan5 |
that said, one or two of those PRs are definitely trivial enough that they don't risk causing any problems and I'd be fine with merging them for 5.2 |
| 13:24 |
|
calcul0n joined #minetest-dev |
| 13:24 |
kilbith |
this is probably my last release with MT, my contract ends in June |
| 13:27 |
sfan5 |
if I may ask, do you plan to keep working on education-related things? |
| 13:28 |
kilbith |
not at all! |
| 13:29 |
kilbith |
I plan to knock the door of Ubisoft |
| 13:33 |
kilbith |
sfan5, regarding that Windows DPI's PR, there's certainly a macro to distinguish Windows 10 from the other versions |
| 13:34 |
Krock |
WINVER |
| 13:34 |
sfan5 |
sure but the displaycaps thing does work in w10, doesn't it? |
| 13:34 |
sfan5 |
... |
| 13:35 |
|
kilbith joined #minetest-dev |
| 13:35 |
kilbith |
I lose the connection everytime my iPhone moves |
| 13:36 |
kilbith |
wouldn't that be safer to use GetDpiForWindow? |
| 13:36 |
sfan5 |
safer? no idea, the result of the two should be exactly the same |
| 13:37 |
sfan5 |
the new w10 api can also support runtime dpi switching, but that can't be easily implemented in MT and would require support from Irrlicht too |
| 13:38 |
Krock |
well, Minetest could register a hook, and forward that to Irrlicht. Not worth the effort, though. A constant DPI value on startup is good enough |
| 13:39 |
Krock |
whereas "forward" would probably be to re-initialize the render |
| 13:41 |
sfan5 |
oh and even if you use the new api, unless you opt in to runtime switching (using the exe manifest) you never get notified of changes anyway |
| 13:41 |
sfan5 |
..which is what leads me to believe the APIs return the exact same value |
| 13:42 |
Krock |
#9591 btw |
| 13:42 |
ShadowBot |
https://github.com/minetest/minetest/issues/9591 -- Revert collision tweaks #9365 and #9327 by SmallJoker |
| 13:43 |
Krock |
will merge trivial #9554 in 10 minutes |
| 13:43 |
ShadowBot |
https://github.com/minetest/minetest/issues/9554 -- Fix 'the the' typos in comments by lnjX |
| 13:44 |
Krock |
sfan5: thank you for updating the translations, looking towards a release this weekend |
| 13:44 |
sfan5 |
sure |
| 13:44 |
Krock |
will check the changelog to update it |
| 13:44 |
sfan5 |
figuring out where the steps are documented was roughly the hardest part |
| 13:45 |
Krock |
sorry, I could've helped you out if I had more time in the past few days |
| 13:52 |
Krock |
merging |
| 14:04 |
sfan5 |
merging #9583 in 10m |
| 14:04 |
ShadowBot |
https://github.com/minetest/minetest/issues/9583 -- Fix cursor still visible after closing formspec while on HyperText by pyrollo |
| 14:17 |
sfan5 |
almost on time |
| 14:29 |
rubenwardy |
Lockdown changes nothing for me - I still have the same hours of work. In fact, it may even be harmful because I'm less likely to look at Mt when I'm in a bad mood, which I will be when stressed and at home all day |
| 14:33 |
rubenwardy |
I can spend some time on mt today though |
| 14:34 |
rubenwardy |
And I'll probably make minetest my "job" if I'm furloughed in a few weeks |
| 14:43 |
|
QwertyDragon_ joined #minetest-dev |
| 15:13 |
|
kilbith joined #minetest-dev |
| 15:21 |
|
DS-minetest joined #minetest-dev |
| 15:55 |
|
kilbith joined #minetest-dev |
| 16:11 |
|
AntumD joined #minetest-dev |
| 16:11 |
|
hlqkj joined #minetest-dev |
| 16:19 |
|
hlqkj joined #minetest-dev |
| 16:45 |
|
fluxflux joined #minetest-dev |
| 16:50 |
|
hlqkj_ joined #minetest-dev |
| 17:06 |
|
Jordach joined #minetest-dev |
| 17:08 |
rubenwardy |
ok, reviewed those two bug fix PRs |
| 17:08 |
rubenwardy |
anything else I can do? |
| 17:11 |
rubenwardy |
dev meeting tonight Krock nore sfan5 sofar ? |
| 17:11 |
sfan5 |
sure |
| 17:12 |
Krock |
will be here |
| 17:12 |
rubenwardy |
I'm free anytime, no plans |
| 17:41 |
sfan5 |
https://old.reddit.com/r/Minetest/comments/fuvsdp/weird_texture_glitch/ |
| 17:41 |
sfan5 |
recently I've heard of 3 people who experience this glitch |
| 17:41 |
sfan5 |
thankfully the object shading changes in 5.2 seem to have accidentally fixed it |
| 17:41 |
sfan5 |
because otherwise I'd have no idea why this is suddenly broken |
| 17:53 |
Krock |
also players reporting the wireframe glitch |
| 17:58 |
rubenwardy |
edits Meetings: 500 Internal Server Error |
| 17:58 |
rubenwardy |
hurray |
| 17:58 |
rubenwardy |
and magically worked |
| 17:58 |
|
kilbith joined #minetest-dev |
| 18:00 |
rubenwardy |
!dev Meetings |
| 18:00 |
ShadowBot |
Meetings - Minetest Developer Wiki -- http://dev.minetest.net/Meetings |
| 18:02 |
|
Fixer_ joined #minetest-dev |
| 18:03 |
|
kilbith_ joined #minetest-dev |
| 18:07 |
Krock |
network desync? |
| 18:07 |
|
erlehmann_ joined #minetest-dev |
| 18:10 |
sfan5 |
the only potential problem I know of is that the client has a budget of 100ms for packet processing per step |
| 18:10 |
sfan5 |
but that should be more than enough |
| 18:12 |
rubenwardy |
I'm updating the credits now |
| 18:12 |
rubenwardy |
I'm waiting for the person to submit a bug report |
| 18:16 |
Krock |
err |
| 18:16 |
Krock |
somebody re-organized the changelog |
| 18:24 |
rubenwardy |
GUI changes were in multiple categories, along with scripting changes |
| 18:27 |
rubenwardy |
what change didn't you like? |
| 18:29 |
Krock |
hmm yes |
| 18:30 |
|
vesper11 joined #minetest-dev |
| 18:30 |
Krock |
I just tried to re-use the previous naming scheme, but probably I added a few points to the wrong category |
| 18:32 |
rubenwardy |
#9592 |
| 18:32 |
ShadowBot |
https://github.com/minetest/minetest/issues/9592 -- Sometimes certain packets don't go from server to client |
| 18:33 |
Krock |
interesting arm |
| 18:38 |
rubenwardy |
wip credits #9593 |
| 18:38 |
ShadowBot |
https://github.com/minetest/minetest/issues/9593 -- [WIP] Update credits for 5.2.0 by rubenwardy |
| 18:41 |
|
calcul0n_ joined #minetest-dev |
| 18:42 |
DS-minetest |
(I'd like to get referred to as "DS" ^_^) |
| 18:42 |
rubenwardy |
ok |
| 18:45 |
sfan5 |
hmm there's one thing I forgot about |
| 18:50 |
|
vesper11 joined #minetest-dev |
| 19:00 |
sfan5 |
#9594 |
| 19:00 |
ShadowBot |
https://github.com/minetest/minetest/issues/9594 -- Builtin: Make item entities glow less by sfan5 |
| 19:01 |
sfan5 |
I remember someone saying they supported this, but I couldn't find that |
| 19:02 |
|
kilbith joined #minetest-dev |
| 19:03 |
rubenwardy |
should ShadowNinja be moved to previous core developers, given that they haven't made a contribution since 2018, and before then 2017? |
| 19:03 |
rubenwardy |
do we have guidance on when that move happens? |
| 19:05 |
rubenwardy |
"for comparison: an unlit item" -> this made me cackle |
| 19:06 |
sfan5 |
previous coredevs is only used for people that explicitly left (I think?) |
| 19:06 |
|
hlqkj_ joined #minetest-dev |
| 19:07 |
rubenwardy |
I don't think pilzadam, sapier, or kahrl ever explicitly left |
| 19:08 |
sfan5 |
hm |
| 19:08 |
sfan5 |
can def.light_source ever be nil? adding a check is safer either way |
| 19:09 |
Krock |
light_source = 0, |
| 19:09 |
Krock |
item.lua L736 |
| 19:09 |
rubenwardy |
it can't |
| 19:09 |
rubenwardy |
I also just checked |
| 19:10 |
sfan5 |
what if some mod goes around and sets it to nil? |
| 19:10 |
Krock |
they don't. |
| 19:10 |
rubenwardy |
hm |
| 19:10 |
rubenwardy |
is the default stuff a meta table or a copy? |
| 19:10 |
sfan5 |
well ok |
| 19:10 |
rubenwardy |
either way, it'll default to the default stuff if the mod sets it to nil |
| 19:10 |
rubenwardy |
I don't mind the check necessarily |
| 19:14 |
sfan5 |
https://github.com/minetest/minetest/commit/fe046fab8aa0bde610f8828fb1c3fc014cf7e193 |
| 19:15 |
sfan5 |
at that time, ShadowNinja applied the "no recent contribs -> sort them out" to coredevs too |
| 19:15 |
Krock |
brb |
| 19:15 |
rubenwardy |
I made that commit |
| 19:15 |
rubenwardy |
and I didn't - I updated based on people leaving the team |
| 19:15 |
rubenwardy |
actually no |
| 19:15 |
rubenwardy |
I did apply that to core devs too |
| 19:15 |
sfan5 |
oh, misread |
| 19:17 |
rubenwardy |
if the cut off for moving a core dev is the past release + a few months, then sofar is just in there :D |
| 19:20 |
Krock |
don't add "[Fixes]" if you don't nkow what they did |
| 19:20 |
rubenwardy |
lol |
| 19:20 |
rubenwardy |
I add fixes when I scroll down the list, and see lots of fixing commits |
| 19:20 |
rubenwardy |
which don't fit a particular categort |
| 19:20 |
Krock |
well then :) |
| 19:21 |
rubenwardy |
lol |
| 19:23 |
Krock |
so merging #9594 and #9591 in a few minutes? |
| 19:23 |
ShadowBot |
https://github.com/minetest/minetest/issues/9594 -- Builtin: Make item entities glow less by sfan5 |
| 19:23 |
ShadowBot |
https://github.com/minetest/minetest/issues/9591 -- Revert collision tweaks #9365 and #9327 by SmallJoker |
| 19:24 |
sfan5 |
I wish I could ask paramat for his opinion on 9594 but he's not here |
| 19:24 |
Krock |
sfan5: like "Formspec", and "Script API", "builtin" is also a certain code area |
| 19:24 |
Krock |
probably not exact, but the most accurate you could get |
| 19:24 |
rubenwardy |
the credits should be user facing really |
| 19:24 |
rubenwardy |
along with the changelog |
| 19:25 |
sfan5 |
oh I found it |
| 19:27 |
Krock |
great :) |
| 19:27 |
Krock |
so merging these two PRs now |
| 19:27 |
Krock |
done |
| 19:27 |
Krock |
forum topic - is anything prepared yet? |
| 19:28 |
sfan5 |
no but I usually do that when it's time |
| 19:28 |
Krock |
and this time around also when the forums work again |
| 19:29 |
sfan5 |
would be helpful though if someone could list the 5 most important features :) |
| 19:29 |
Krock |
game#2566 is still open |
| 19:29 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2566 -- Fix beds by An0n3m0us |
| 19:29 |
Krock |
move to 5.3.0 |
| 19:29 |
sfan5 |
agreed |
| 19:29 |
Krock |
moved. |
| 19:29 |
Krock |
- Hypertext Formspec element |
| 19:29 |
Krock |
- Object shading |
| 19:30 |
Krock |
- Documentation improvements |
| 19:30 |
Krock |
- New sky API functions |
| 19:30 |
sfan5 |
i don't think users care about the documentation much |
| 19:30 |
Krock |
- New bugs to fix later |
| 19:31 |
Krock |
- Pathfinder improvements |
| 19:31 |
Krock |
- Altered ObjectRef livecycle |
| 19:31 |
Krock |
- 9-slice buttons |
| 19:32 |
Krock |
... + styling |
| 19:32 |
Krock |
I think that's about the most important stuff |
| 19:33 |
Krock |
the other changes are really just bugfixes, and stuff that's not too important for a short read |
| 19:35 |
Krock |
rubenwardy: so you want #9582 in 5.2.0? I tested it too, and it works. but I think leaving this discussion after the release would be helpful |
| 19:35 |
ShadowBot |
https://github.com/minetest/minetest/issues/9582 -- Fix hovercolor not working in global tag by pyrollo |
| 19:35 |
Krock |
deding-wise |
| 19:36 |
Krock |
*design-wise |
| 19:36 |
rubenwardy |
so, fixing it to the documented behaviour is trivial |
| 19:36 |
rubenwardy |
but it may cause problems later |
| 19:36 |
rubenwardy |
I think the documented behaviour is fairly valid |
| 19:38 |
Krock |
* `hovercolor`: Color of <action> tags when mouse is over. |
| 19:39 |
Krock |
that's why there's "<tag .." |
| 19:39 |
Krock |
fair, if there are more elements that make use of "hovercolor", it might make more sense |
| 19:40 |
rubenwardy |
well, I don't mind either way |
| 19:41 |
Krock |
oh my.. so many PRs to review after the release |
| 19:42 |
rubenwardy |
idk |
| 19:42 |
rubenwardy |
*ikr |
| 19:42 |
rubenwardy |
priority being with fixing Android support |
| 20:17 |
|
erlehmann joined #minetest-dev |
| 20:34 |
|
calcul0n joined #minetest-dev |
| 20:38 |
|
kilbith joined #minetest-dev |
| 20:42 |
|
Taoki joined #minetest-dev |
| 20:42 |
|
Wuzzy joined #minetest-dev |
| 20:44 |
rubenwardy |
I hate bundler/gem so much |
| 20:45 |
rubenwardy |
it never works |
| 20:46 |
rubenwardy |
programming language package managers that install system-wide by default need to die already |
| 20:46 |
sfan5 |
I never actually test my changes to the website because I don't want to touch ruby |
| 20:47 |
Wuzzy |
were any new settings added to -dev? |
| 20:47 |
Wuzzy |
i ask because of settingtypes.txt |
| 20:48 |
rubenwardy |
python/pip at least has virtual envs |
| 20:48 |
Wuzzy |
ah, indeed |
| 20:48 |
rubenwardy |
I can check that |
| 20:48 |
Wuzzy |
sorry, there are indeed more settings |
| 20:48 |
sfan5 |
I ran the script yesterday and it did not make any changes to settingtypes.txt |
| 20:48 |
Wuzzy |
which script? |
| 20:49 |
Wuzzy |
I found a new setting: font_path_bold |
| 20:49 |
sfan5 |
generate_from_settingtypes.lua |
| 20:49 |
Wuzzy |
did not exist in 5.1.1 |
| 20:49 |
Wuzzy |
o_O |
| 20:49 |
rubenwardy |
it's incredibly stupid how it's done, I hate it |
| 20:49 |
Wuzzy |
looks like the script managed to miss font_path_bold then :/ |
| 20:49 |
rubenwardy |
also, settingstypes.txt is a source file |
| 20:49 |
sfan5 |
uh well |
| 20:50 |
Wuzzy |
and indeed: no german translation for the new font_* settings, neither in PO nor weblate |
| 20:50 |
sfan5 |
it generates minetest.conf.example, settings_translation_file.cpp FROM settingtypes.txt |
| 20:50 |
rubenwardy |
if there's a new setting that isn't in there then it won't be in minetest.conf.examples |
| 20:50 |
Wuzzy |
i mean source string |
| 20:50 |
rubenwardy |
yeah |
| 20:50 |
sfan5 |
so whoever added the setting didn't add it to settingtypes.txt |
| 20:50 |
Wuzzy |
huh |
| 20:50 |
Wuzzy |
it shows up in all settings O_O |
| 20:51 |
sfan5 |
>./builtin/settingtypes.txt:font_path_bold (Bold font path) filepath fonts/Arimo-Bold.ttf |
| 20:51 |
Wuzzy |
font_path_bold (Bold font path) filepath fonts/Arimo-Bold.ttf |
| 20:51 |
Wuzzy |
so its in settingtypes.txt |
| 20:51 |
Wuzzy |
notice how this doesn't have a description prepended by # |
| 20:51 |
Wuzzy |
maybe the script chokes on settings that don't have a long description |
| 20:51 |
Wuzzy |
? |
| 20:52 |
sfan5 |
./po/de/minetest.po:msgid "Bold font path" |
| 20:52 |
Wuzzy |
wat. |
| 20:52 |
sfan5 |
..? |
| 20:52 |
Wuzzy |
ok |
| 20:53 |
sfan5 |
https://hosted.weblate.org/translate/minetest/minetest/en/?q=%22Bold+font+path%22&checksum=&offset=1#translations |
| 20:53 |
sfan5 |
it very clearly exists |
| 20:53 |
Wuzzy |
ok ok i'm recompiling haha. sorry |
| 20:53 |
Wuzzy |
what did you do to make weblate update? I'm just curius how it works |
| 20:53 |
Wuzzy |
(also for the devwiki :P) |
| 20:53 |
sfan5 |
https://dev.minetest.net/Translating#How_to_merge_translations_from_Hosted_Weblate |
| 20:55 |
Wuzzy |
"A good schedule is once per month." << that doesn't actually happen, right? reality is "once per release" |
| 20:55 |
Wuzzy |
unless i'm mistaken |
| 20:55 |
sfan5 |
correct |
| 20:55 |
sfan5 |
and I think that's fine as long as it's done early enough (on feature freeze) |
| 20:57 |
Wuzzy |
hmmmm maybe there were more updates that i missed |
| 20:57 |
Wuzzy |
because the bold strings are clearly in weblate which are all new |
| 20:58 |
Wuzzy |
i know those did not appear recently |
| 20:58 |
Wuzzy |
maybe someone pushed an update and i missed it ? |
| 20:58 |
Wuzzy |
so that would mean i was wrong |
| 20:59 |
sfan5 |
https://github.com/minetest/minetest/commit/987b2c5f372ea30e90f3d83dd9e76c2286a14a34 < this one added "Bold font path" |
| 21:01 |
Wuzzy |
no i did not mean the commit itself, i meant the moment when weblate was triggered |
| 21:01 |
Wuzzy |
but i see now that "updating weblate" requires quite a few (boring) steps. i see now why its not done so often ? |
| 21:01 |
sfan5 |
yes, I linked the translation update commit |
| 21:01 |
sfan5 |
you can be sure weblate was updated on the same day |
| 21:02 |
Wuzzy |
oops i assumed you meant the commit that added the feature (didn't click) |
| 21:04 |
|
Poikilos_ joined #minetest-dev |
| 21:07 |
sfan5 |
hm |
| 21:07 |
sfan5 |
rubenwardy: you didn't have a chance yet to test the Raspberry Pi build I linked did you? |
| 21:07 |
rubenwardy |
I didn't yet |
| 21:07 |
sfan5 |
thought so |
| 21:07 |
rubenwardy |
I can pull it out now |
| 21:08 |
rubenwardy |
I didn't have a keyboard when you sent a link, I do now |
| 21:19 |
|
Poikilos_ joined #minetest-dev |
| 22:22 |
rubenwardy |
sfan5: build seems to work well, albeit slowly |
| 22:22 |
rubenwardy |
still waiting for media to load |
| 22:23 |
sfan5 |
i'm glad that it works at all |
| 22:23 |
rubenwardy |
i think the slowly thing is more the RPi than the build |
| 22:23 |
rubenwardy |
ah, loaded |
| 22:23 |
rubenwardy |
a glorious 1 FPS |
| 22:24 |
sfan5 |
hm that's not better than I had with my rpi1 |
| 22:24 |
rubenwardy |
I have an RPi 3 B+ |
| 22:24 |
rubenwardy |
it's emulated graphics |
| 22:24 |
rubenwardy |
RPi 4s have integrated graphics |
| 22:25 |
sfan5 |
what does the window title show as renderer? |
| 22:25 |
rubenwardy |
OpenGL ES-CM 1.1 Mesa 19.3.2 |
| 22:25 |
sfan5 |
hm that's not indicative if it's software rendering or not |
| 22:26 |
sfan5 |
ogles2 should work too, btw |
| 22:26 |
sfan5 |
output from eglinfo should suffice to confirm the actual driver |
| 22:28 |
rubenwardy |
ogles2 is much slower - draw time 600ms |
| 22:29 |
rubenwardy |
it's possible my CPU is thermal throttling |
| 22:32 |
|
kilbith joined #minetest-dev |
| 22:34 |
rubenwardy |
there are a bunch of errors, not sure if they're bad |
| 22:34 |
rubenwardy |
1 sdc |
| 22:34 |
kilbith |
rubenwardy: I'm perplex about that "number of changes" criteria, e.g. a small fix of 3 LOC can take long to find out and that's something you cannot evaluate from an algorithm |
| 22:37 |
rubenwardy |
I'll re-run it with a flat weighting |
| 22:37 |
rubenwardy |
it's supposed to be correct on average rather than correct for everything |
| 22:37 |
rubenwardy |
but a flat weighting might have a similar affect and be more fair |
| 22:38 |
sfan5 |
number of commits does not necessarily indicate hours of work either, it's just a heuristic after all |
| 22:38 |
rubenwardy |
yeah |
| 22:38 |
kilbith |
or just drop the contributor list |
| 22:38 |
kilbith |
the "ego" list |
| 22:38 |
rubenwardy |
a weighting of a point for any contribution gives the same names |
| 22:39 |
rubenwardy |
perhaps the "active contributors" list should be _all_ contributors since the last release |
| 22:43 |
sfan5 |
rubenwardy: could you run `eglinfo` on your pi and pastebin the output? |
| 22:43 |
|
kilbith joined #minetest-dev |
| 22:43 |
rubenwardy |
I don't have the installed, I'll google it |
| 22:43 |
sfan5 |
should be |
| 22:43 |
sfan5 |
mesa-utils-extra |
| 22:44 |
|
absurb joined #minetest-dev |
| 22:52 |
|
benrob0329 joined #minetest-dev |
| 22:53 |
rubenwardy |
!title https://gist.github.com/rubenwardy/c3830f421a8dbfa398f077a3e12cda65 |
| 22:53 |
ShadowBot |
rubenwardy: RPi 3 B+ info · GitHub |
| 22:54 |
sfan5 |
the jpeg thing is my fault, no idea what's up with the second warning |
| 22:54 |
sfan5 |
the eglinfo output still doesn't give me what I was hoping for |
| 22:54 |
sfan5 |
but thanks for testing so far |
| 22:56 |
sfan5 |
hm, did you enable support for the GL driver in raspi-config? |
| 22:56 |
rubenwardy |
I didn't touch anything |
| 22:56 |
rubenwardy |
it's a fresh install, as-of an hour ago |
| 22:57 |
rubenwardy |
because somehow my previous was bricked |
| 22:57 |
sfan5 |
heh, classic |
| 22:57 |
rubenwardy |
I think it wasn't booting because my external hard drive wasn't plugged in? |
| 22:57 |
rubenwardy |
there was nothing of value on the rpi, and I couldn't be bothered to get the power supply for my external hard drive, so just wiped the sdcard |
| 23:01 |
sfan5 |
try enabling the GL driver please |
| 23:13 |
rubenwardy |
sfan5: I have three options G1 legacy, g2 gl fake mks, g3 gl full kms |
| 23:14 |
rubenwardy |
number 3? |
| 23:14 |
sfan5 |
not sure if KMS matters but pick 3 yeah |
| 23:18 |
rubenwardy |
so much better, it's now 20FPS |
| 23:19 |
rubenwardy |
I'll try gles2 |
| 23:20 |
sfan5 |
nice |
| 23:21 |
rubenwardy |
ogles2 gives me 10FPS (75-80ms). I'm measuring it by opening the same world, letting it load, then point to the same position |
| 23:21 |
rubenwardy |
not entirely scientific |
| 23:22 |
sfan5 |
does the window title show something else? |
| 23:22 |
|
Taoki joined #minetest-dev |
| 23:23 |
rubenwardy |
it does |
| 23:23 |
rubenwardy |
OpenGL ES 2.0 Mesa 19.3.2 |
| 23:23 |
rubenwardy |
I'll copy the glinfo in a sec |
| 23:23 |
sfan5 |
well that's not very different from "OpenGL ES-CM 1.1 Mesa 19.3.2" |
| 23:24 |
sfan5 |
so I wonder how Minetest would detect that it's using software rendering |
| 23:24 |
rubenwardy |
it is still "OpenGL ES-CM 1.1 Mesa 19.3.2" when using ogles1 |
| 23:25 |
sfan5 |
hm, maybe have to read GL_RENDERER |
| 23:26 |
sfan5 |
if you feel like testing more, I'd also like to know whether you consider it playable for general use |
| 23:26 |
sfan5 |
and perhaps which graphics settings could be tweaked to make it more playable |
| 23:26 |
rubenwardy |
https://gist.github.com/rubenwardy/0d9ead3fe07ad2681eaf9102c6e4de43 |
| 23:27 |
sfan5 |
the reason I am asking all this is that I think we should provide official builds for the Raspberry Pi (downloadable .debs or perhaps a repository) |
| 23:27 |
rubenwardy |
yeah, sounds good |
| 23:27 |
rubenwardy |
it would also be nice to have build flags to remove features that may be intensive |
| 23:27 |
sfan5 |
because distros will build with OpenGL which tends(?) to be slower and we also don't get the opportunity to set optimized defaults for weak GPUs |
| 23:27 |
rubenwardy |
I'm not sure what |
| 23:28 |
sfan5 |
I remember Krock saying that stars take lots of render time |
| 23:28 |
rubenwardy |
even during day time? |
| 23:28 |
rubenwardy |
maybe I should profile, but the problem with that is cpu vs gpu time |
| 23:28 |
sfan5 |
I hope they aren't rendered during day time :D |
| 23:36 |
|
Miner_48er joined #minetest-dev |
