| Time |
Nick |
Message |
| 00:31 |
DS-minetest |
is it possible to lua-unittest with json? |
| 00:31 |
DS-minetest |
(I mean using core.parse_json in a unittest) |
| 00:32 |
DS-minetest |
(it's implemented in c++) |
| 00:45 |
rubenwardy |
!tell DS-minetest No, you cannot use any engine functions. However, you can import any Lua libraries and use them to mock engine functions |
| 00:45 |
ShadowBot |
rubenwardy: O.K. |
| 01:42 |
|
nephele_ joined #minetest-dev |
| 02:09 |
|
Foz joined #minetest-dev |
| 04:51 |
|
QwertyDragon_ joined #minetest-dev |
| 05:00 |
|
QwertyDragon joined #minetest-dev |
| 05:22 |
|
realz joined #minetest-dev |
| 06:19 |
|
absurb joined #minetest-dev |
| 06:54 |
|
Kimapr joined #minetest-dev |
| 07:06 |
|
mizux joined #minetest-dev |
| 07:15 |
|
calcul0n joined #minetest-dev |
| 08:00 |
|
calcul0n joined #minetest-dev |
| 08:03 |
|
Kimapr joined #minetest-dev |
| 08:04 |
|
Beton joined #minetest-dev |
| 08:26 |
|
calcul0n_ joined #minetest-dev |
| 08:38 |
|
ShadowNinja joined #minetest-dev |
| 08:40 |
|
hlqkj joined #minetest-dev |
| 10:33 |
|
proller joined #minetest-dev |
| 10:57 |
|
DS-minetest joined #minetest-dev |
| 11:05 |
|
Fixer joined #minetest-dev |
| 11:21 |
|
erlehmann_ joined #minetest-dev |
| 11:34 |
|
calcul0n__ joined #minetest-dev |
| 11:45 |
|
turtleman joined #minetest-dev |
| 11:49 |
DS-minetest |
I don't think I can trust some lua json library to do exactly the same as minetest's core.write_json |
| 12:00 |
|
DS-minetest joined #minetest-dev |
| 12:01 |
|
kb1000 joined #minetest-dev |
| 12:16 |
|
hlqkj joined #minetest-dev |
| 12:18 |
|
erlehmann joined #minetest-dev |
| 12:20 |
|
erlehmann joined #minetest-dev |
| 12:47 |
|
erlehmann joined #minetest-dev |
| 12:50 |
|
erlehmann_ joined #minetest-dev |
| 13:04 |
|
erlehmann joined #minetest-dev |
| 13:24 |
|
kilbith joined #minetest-dev |
| 13:26 |
|
kilbith joined #minetest-dev |
| 13:49 |
|
kb1000 joined #minetest-dev |
| 13:53 |
|
Wuzzy joined #minetest-dev |
| 14:19 |
|
EricD joined #minetest-dev |
| 14:20 |
|
EricD left #minetest-dev |
| 14:27 |
Wuzzy |
My games now get warnings because my ABM takes 3ms. Is this intentional? I remember in previous versions, the warning is triggered at 200ms |
| 14:30 |
sfan5 |
the budget isn't a constant https://github.com/minetest/minetest/blob/master/src/serverenvironment.cpp#L1366 |
| 14:30 |
sfan5 |
3ms is way too low, though |
| 14:33 |
|
Wuzzy2 joined #minetest-dev |
| 14:37 |
behalebabo |
Wuzzy2: I got that warning, somehow my server step setting had been set really low |
| 14:38 |
|
Wuzzy joined #minetest-dev |
| 14:41 |
Wuzzy |
behalebabo: what is this setting called? |
| 14:41 |
Wuzzy |
dedicated_server_step? |
| 14:41 |
Wuzzy |
for me its 1e-05... O_O |
| 14:41 |
behalebabo |
I think that was the one, it had gotten set down at the bottom of my minetest.conf |
| 14:41 |
behalebabo |
yeah |
| 14:41 |
behalebabo |
same |
| 14:42 |
Wuzzy |
i never touched this setting. must be a new default |
| 14:42 |
behalebabo |
I don't know why |
| 14:43 |
Wuzzy |
hmm but it appears in my minetest.conf somehow. how did THAT happen? i wonder if it was caused by minetest |
| 14:43 |
Wuzzy |
ok so the default is 0.09. makes more sense |
| 14:43 |
sfan5 |
more likely explanation: some mod author thought it was a good idea to set that from his mods |
| 14:43 |
Wuzzy |
hmmmmm |
| 14:43 |
Wuzzy |
oh i have an idea |
| 14:44 |
Wuzzy |
ok so it's not Crafter |
| 14:44 |
behalebabo |
it's crafter |
| 14:44 |
behalebabo |
https://github.com/oilboi/Crafter/blob/03e4c50bfc27fc439998eefda09e1747d2a1beac/mods/main/settings.lua#L5 |
| 14:44 |
Wuzzy |
argh. i made a typo in grep ? |
| 14:45 |
Wuzzy |
yes, crafter was my first suspect, too ? |
| 14:45 |
behalebabo |
yeah, abm_interval was the other setting |
| 14:46 |
Wuzzy |
oh my, this game quite violently screws up my settings! grrr |
| 14:46 |
Wuzzy |
but ok, its in alpha, i have no right to complain ? |
| 14:46 |
Wuzzy |
Recommended way for a game to set default settings is via minetest.conf in game folder, right? and not with minetest.settings:set(). Right? |
| 14:47 |
sfan5 |
definitely |
| 14:47 |
Wuzzy |
because Crafter did the latter ? |
| 14:49 |
Wuzzy |
I'm pretty sure a dedicated_server_step of 0.00001 is absurd ? |
| 14:49 |
Wuzzy |
what is a reasonable low value for dedicated_server_step? |
| 14:51 |
sfan5 |
if you absolutely want to try to improve responsiveness (which is what I assume to be the intention), go with 0.008 which equals 125Hz |
| 14:52 |
sfan5 |
I doubt anything below that noticably helps (and it might cause the engine to do weird things as you can see) |
| 14:53 |
celeron55 |
server step of 10 microseconds? lol |
| 14:54 |
Wuzzy |
yeah, some alpha stage game actually set that in lua |
| 14:54 |
celeron55 |
there's a reason for most default values, you don't just go about and change them 10000x |
| 14:54 |
Wuzzy |
yep |
| 14:55 |
celeron55 |
values in the range of about 0.01 to 0.2 might make sense |
| 15:01 |
Wuzzy |
ah. the weird abm warning i got was actually caused by another setting that was touched by the alpha game (Crafter): abm_interval |
| 15:18 |
sfan5 |
merging game#2635, game#2636 soon |
| 15:18 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2635 -- Fix creative category sorting in certain cases by sfan5 |
| 15:18 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2636 -- Fix crash on TNT explosions if tnt is disabled by sfan5 |
| 15:35 |
|
calcul0n joined #minetest-dev |
| 15:35 |
|
turtleman joined #minetest-dev |
| 15:47 |
|
DS-minetest joined #minetest-dev |
| 15:53 |
Wuzzy |
where is the default sun/moon texture stored? |
| 16:10 |
|
calcul0n_ joined #minetest-dev |
| 16:21 |
|
Dargod joined #minetest-dev |
| 16:22 |
Wuzzy |
Is it normal when I get a floating point exception when using the [sheet texture modifier? |
| 16:24 |
Krock |
exceptions from modifiers are not wanted |
| 16:25 |
Krock |
there's no default sun/moon texture |
| 16:25 |
Krock |
it's either one or two rectangles |
| 16:26 |
|
calcul0n joined #minetest-dev |
| 16:29 |
Wuzzy |
oh. apparently i used sheet incorrectly |
| 16:29 |
Wuzzy |
"tex.png^[sheet:4,2:0,0" instead of "tex.png^[sheet:4x2:0,0" |
| 16:29 |
Wuzzy |
the first one causes a floating point exception (and is also a syntax error) |
| 16:55 |
sfan5 |
god I hate working with the networking code |
| 17:01 |
sfan5 |
none of this shit makes any sense |
| 17:04 |
sfan5 |
here's a great excerpt: |
| 17:04 |
sfan5 |
bool invalid_address = true; |
| 17:04 |
sfan5 |
if (invalid_address) { |
| 17:05 |
rubenwardy |
looool |
| 17:05 |
rubenwardy |
I've avoided opening any networking files except protocol.h |
| 17:05 |
rubenwardy |
and that one time I thought about adding a Host field to the hello packet |
| 17:06 |
|
Fixer_ joined #minetest-dev |
| 17:06 |
rubenwardy |
which I'd still like to do, maybe - it allows virtual hosting of servers by having the first packet contain the address given in the connection form |
| 17:06 |
sfan5 |
most files are okay just don't open connection{threads,}.{cpp,h} |
| 17:06 |
rubenwardy |
I see |
| 17:07 |
rubenwardy |
when, I was writing my own netcode, I finished fragmentation and then started reliability channels and thought: "why am I doing this, I want to make a game". And so switched to enet... |
| 17:18 |
Krock |
sfan5: a lot optimization potential |
| 18:06 |
sfan5 |
merging #9611, |
| 18:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/9611 -- Miscellaneous networking improvements by sfan5 |
| 18:06 |
sfan5 |
#9614 and #9598 in 5m |
| 18:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/9614 -- Work around LuaJIT issues on aarch64 by sfan5 |
| 18:06 |
ShadowBot |
https://github.com/minetest/minetest/issues/9598 -- Overall improvements to log messages by sfan5 |
| 18:27 |
|
calcul0n_ joined #minetest-dev |
| 18:36 |
sfan5 |
merging game#2640 |
| 18:36 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2640 -- player_api: Remove local variable. by orbea |
| 18:39 |
DS-minetest |
what is the intended way to run luachecker for builtin? just doing `luachecker builtin` does not ignore the test files |
| 18:40 |
sfan5 |
did you check how travis runs it? |
| 18:40 |
DS-minetest |
no, will do |
| 18:40 |
rubenwardy |
the test files are luachecked |
| 18:40 |
rubenwardy |
and should be |
| 18:41 |
rubenwardy |
if you're having problems due to mocking, you can handle the test files differently like files["**/tests/*"] = { globals = { "thing" } } |
| 18:43 |
DS-minetest |
hm, but I also get warnings if I run luachecker in master |
| 18:46 |
rubenwardy |
are you running from the root folder |
| 18:46 |
rubenwardy |
it's luacheck builtin from the root folder |
| 18:48 |
DS-minetest |
I'm calling from the minetest folder |
| 18:52 |
DS-minetest |
luacheck builtin --exclude-files **/tests/* works fine |
| 18:55 |
rubenwardy |
you probably shouldn't do that |
| 18:55 |
rubenwardy |
what are the errors? |
| 18:55 |
rubenwardy |
I used luacheck on my projects with tests with no issues |
| 18:56 |
DS-minetest |
it's warnings, eg.: builtin/common/tests/vector_spec.lua:178:3: accessing undefined field is_true of global assert |
| 18:57 |
rubenwardy |
weird |
| 18:58 |
DS-minetest |
luacheck installed with apt (lua-check), version: Luacheck: 0.21.1 |
| 18:59 |
|
Hijiri_ left #minetest-dev |
| 19:18 |
|
hlqkj joined #minetest-dev |
| 19:39 |
sfan5 |
rubenwardy: does this look okay? http://sprunge.us/yvgubg?diff |
| 19:40 |
rubenwardy |
looks good, although I'm not sure about the VALGRIND travis job |
| 19:40 |
rubenwardy |
I guess that does the unit coverage? |
| 19:40 |
rubenwardy |
well, unit testings |
| 19:41 |
sfan5 |
the freetype=0 and valgrind=1 ones still run w/o |
| 19:41 |
sfan5 |
and the builds pass https://travis-ci.org/github/minetest/minetest/builds/672661514 |
| 19:43 |
rubenwardy |
lgtm |
| 19:43 |
sfan5 |
thanks, pushing it to master |
| 19:46 |
sfan5 |
what a productive day |
| 19:47 |
rubenwardy |
impressive |
| 19:53 |
|
Hijiri joined #minetest-dev |
| 20:00 |
|
EricD joined #minetest-dev |
| 20:02 |
|
proller joined #minetest-dev |
| 20:03 |
|
edalonge joined #minetest-dev |
| 20:05 |
|
edalonge joined #minetest-dev |
| 20:06 |
|
edalonge joined #minetest-dev |
| 20:07 |
|
Fixer joined #minetest-dev |
| 20:29 |
sfan5 |
merging #9604, #9343 in 15m |
| 20:29 |
ShadowBot |
https://github.com/minetest/minetest/issues/9604 -- Allow to select en as language in settings menu by Desour |
| 20:29 |
ShadowBot |
https://github.com/minetest/minetest/issues/9343 -- Collision various fixes by TheTermos |
| 20:30 |
|
kilbith joined #minetest-dev |
| 20:30 |
rubenwardy |
https://www.reddit.com/r/Minetest/comments/fxcnet/520_impressions/ |
| 20:32 |
sfan5 |
do they mean the gui by "looks horrible"? |
| 20:32 |
rubenwardy |
main menu? |
| 20:32 |
sfan5 |
I thin in general, the irrlicht gui looks like it was designed in 2010 |
| 20:33 |
sfan5 |
think* |
| 20:34 |
rubenwardy |
someone needs to come up with a different design |
| 20:35 |
rubenwardy |
I can get my head around UX, and I find that interesting, but I'm not a graphics designer |
| 20:36 |
pmp-p |
ask the guys complaining to give money to hire ux people :D |
| 20:37 |
sfan5 |
applying a pixel-art looking gui skin would fix 99% of complaints |
| 20:37 |
sfan5 |
like the progress bar looks you know |
| 20:39 |
kilbith |
that I made |
| 20:39 |
pmp-p |
yeah someone did that on ultra old ugly looking panda3d gui ( as old as irrlicht default ) and he got that https://www.youtube.com/watch?v=Y3r4JrTHZAA |
| 20:40 |
kilbith |
the craftguide GUI looks like it was designed in 2020 with the 2005 tools |
| 20:51 |
rubenwardy |
sfan5: by looks horrible, they're referring to the pixellated text |
| 20:53 |
sfan5 |
oh |
| 20:53 |
sfan5 |
apple displays are high dpi ("retina") and users are used to very smooth fonts |
| 20:54 |
sfan5 |
it appears neither us nor irrlicht implement any support for dpi scaling on mac |
| 20:54 |
sfan5 |
though wouldn't pixely fonts be considered fitting for minetest? |
| 20:54 |
rubenwardy |
it's nicely pixelly thought |
| 20:54 |
rubenwardy |
it still does anti-aliasing |
| 20:54 |
rubenwardy |
whereas, with a pixel art font you'd turn off anti-aliasing |
| 20:54 |
kilbith |
I'm not sure fonts are scaled like 2D graphics |
| 20:55 |
kilbith |
MacOS uses freetype right? |
| 20:55 |
sfan5 |
macos has CoreText |
| 20:55 |
sfan5 |
but minetest only uses freetype |
| 20:57 |
pmp-p |
iirc freetype subpixel rendering is not available platform without ClearType license |
| 20:58 |
sfan5 |
didn't those patents expire a few years ago? also I don't think lack of cleartype means you get no subpixel rendering *at all* |
| 20:58 |
kilbith |
yes, since freetype 2.7 or 2.8 |
| 21:01 |
pmp-p |
from the doc i'd say it depends on distro build options |
| 21:01 |
pmp-p |
was reading there https://www.freetype.org/freetype2/docs/reference/ft2-lcd_rendering.html |
| 21:05 |
|
absurb joined #minetest-dev |
| 21:12 |
|
calcul0n joined #minetest-dev |
| 21:34 |
|
erlehmann joined #minetest-dev |
| 22:05 |
|
calcul0n_ joined #minetest-dev |
| 22:09 |
|
GreenXenith joined #minetest-dev |
| 22:37 |
|
kilbith_ joined #minetest-dev |
| 22:52 |
|
turtleman joined #minetest-dev |
| 23:12 |
|
kilbith_ joined #minetest-dev |
| 23:29 |
|
GreenXenith joined #minetest-dev |
