| Time |
Nick |
Message |
| 00:01 |
|
kenanbit joined #minetest-dev |
| 00:32 |
RealBadAngel |
thexyz, ive finished polish translation |
| 01:49 |
|
rsiska joined #minetest-dev |
| 02:02 |
hmmmm |
you know it might be neat if you could separate log files, e.g. --verbosestream verbose.txt --errorstream debug.log, et cetera |
| 02:34 |
RealBadAngel |
request: logging, writing gigabytes of data to disk, shall be disabled by default |
| 02:34 |
RealBadAngel |
and enabled only by devs when they need it |
| 02:41 |
RealBadAngel |
im gettin now 50+ fps instead of 25-30 (with 16x/32x textures), vanessaE have 40-45 instead of 25 (with 256x textures) |
| 02:41 |
VanessaE |
with a view range of 80, in my case. |
| 02:49 |
hmmmm |
i guess i could do that really quick. not right now though. |
| 02:50 |
RealBadAngel |
imho conf option should be best option |
| 02:50 |
hmmmm |
obviously |
| 02:51 |
hmmmm |
i'll add a config entry called "log_level" which will be a number representing how much is logged (0 = no logging, 1 = errors, 2 = action, 3 = info, 4 = verbose, 5 = dstream) |
| 02:52 |
RealBadAngel |
flawless |
| 02:53 |
hmmmm |
you know, when i started out playing minetest, the log file was my biggest frustration, and it kept making things from already slow to even slower to the point where i couldn't place blocks |
| 02:53 |
hmmmm |
as you just saw in the main channel it's the same for a lot of people |
| 02:54 |
RealBadAngel |
yeah |
| 02:54 |
RealBadAngel |
thats why i started to test it |
| 02:55 |
RealBadAngel |
if i could see a few fps change, i would just say its not worth it |
| 02:55 |
RealBadAngel |
but when i see 100 to 180% fps change.... |
| 03:03 |
hmmmm |
you know i'm noticing that most of the time, you acquire the two envlocks in emergethread immediately |
| 03:03 |
hmmmm |
maybe it's a different story once you have a lot of clients on your server |
| 03:04 |
hmmmm |
when it is locked, though, it takes mostly either 33ms or 400ms for either |
| 03:04 |
hmmmm |
that's 400ms of sitting there and doing nothing like a shmuck |
| 03:04 |
Exio |
well, i keep getting "virtual irr::u32 TextureSource::getTextureId(const std::string&): Assertion 'result.key == name' failed." any ideas? |
| 03:05 |
* hmmmm |
hides |
| 03:05 |
Exio |
and it worked now! |
| 03:05 |
Exio |
weird |
| 03:10 |
RealBadAngel |
you can back hmmm |
| 03:10 |
RealBadAngel |
lol |
| 04:17 |
hmmmm |
celeron55, i have a question that you probably you have the answer to: are _any_ other lua callbacks done while envlocked aside from on_generate in EmergeThread? |
| 04:17 |
hmmmm |
i am pretty sure add_node is safe |
| 04:26 |
hmmmm |
well doesn't look like it. |
| 04:26 |
hmmmm |
ahh this is all the more reason why i'd like to make a voxelmanipulator for lua and pass that to on_generate right after makeChunk instead of inside a lock |
| 06:17 |
|
Calinou joined #minetest-dev |
| 06:30 |
|
ecube joined #minetest-dev |
| 08:17 |
|
dv8 joined #minetest-dev |
| 09:41 |
|
proller joined #minetest-dev |
| 09:56 |
|
proller joined #minetest-dev |
| 10:32 |
|
PilzAdam joined #minetest-dev |
| 10:38 |
|
rsiska joined #minetest-dev |
| 11:09 |
|
darkrose joined #minetest-dev |
| 13:04 |
|
Taoki joined #minetest-dev |
| 13:11 |
|
progers joined #minetest-dev |
| 14:20 |
|
hmmmm joined #minetest-dev |
| 15:13 |
|
iqualfragile joined #minetest-dev |
| 15:14 |
|
SpeedProg joined #minetest-dev |
| 15:15 |
|
SpeedProg1 joined #minetest-dev |
| 15:17 |
|
SpeedProg joined #minetest-dev |
| 15:17 |
|
proller joined #minetest-dev |
| 16:25 |
|
Calinou joined #minetest-dev |
| 17:31 |
|
doserj joined #minetest-dev |
| 17:42 |
|
doserj joined #minetest-dev |
| 17:46 |
doserj |
comments about https://github.com/celeron55/minetest/pull/459 (doxygen support)? |
| 17:48 |
thexyz |
i don't think that's really needed, we have dev.minetest.net |
| 17:48 |
Calinou |
^ |
| 17:49 |
darkrose |
doxygen! kill it with fire! |
| 17:49 |
Calinou |
https://github.com/celeron55/minetest/pull/455 <= and this? |
| 17:49 |
* Calinou |
upvotes darkrose. |
| 17:52 |
doserj |
http://dev.minetest.net/Engine_documentation <- explicitely asks for doxygen comments :) |
| 17:54 |
doserj |
And I don't think you want to document individual functions on dev.minetest.net |
| 19:27 |
|
proller joined #minetest-dev |
| 19:42 |
|
doserj joined #minetest-dev |
| 20:04 |
|
kahrl joined #minetest-dev |
| 21:27 |
hmmmm |
lel le upvote |
| 21:27 |
hmmmm |
the person who said that was MarkTraceur |
| 21:27 |
hmmmm |
he may want doxygen comments scattered throughout the code, I certainly don't |
| 21:27 |
hmmmm |
keep our rivers^H^H^H^H^Hcode clean |
| 21:28 |
marktraceur |
hmmmm: celeron55 suggested something automatic as opposed to a wiki |
| 21:28 |
hmmmm |
i vote against it |
| 21:29 |
hmmmm |
automatic documentation doesn't really help, we can figure what a function does, what it takes and what it returns from its name and arguments |
| 21:29 |
hmmmm |
the sort of comments we need are the ones that aren't so obvious that actually help people understand how it works or what quirks it has |
| 21:47 |
|
jin_xi joined #minetest-dev |
| 21:51 |
|
VanessaE joined #minetest-dev |
| 21:57 |
proller |
freebsd port updated to 0.4.4 |
| 22:00 |
hmmmm |
have you taken over port maintainer status? |
| 22:02 |
proller |
no i find alive maintainer |
| 22:05 |
proller |
very close to me |
| 22:08 |
doserj |
hmmmm: what exactly are you voting against? Having comments in the source code that describe what functions do, or aving excessive doxygen markup in the source code, or having a "make doc" target using doxygen? |
| 22:22 |
|
SpeedProg joined #minetest-dev |
| 23:11 |
|
exe_ joined #minetest-dev |
| 23:29 |
|
exe_ joined #minetest-dev |
| 23:49 |
hmmmm |
avoiding doxygen |
| 23:57 |
doserj |
what does that mean? |
