| Time |
Nick |
Message |
| 04:05 |
|
jyfl987 joined #minetest-dev |
| 09:01 |
|
Calinou joined #minetest-dev |
| 09:40 |
|
PilzAdam joined #minetest-dev |
| 10:37 |
|
celeron55 joined #minetest-dev |
| 12:42 |
|
SpeedProg joined #minetest-dev |
| 13:53 |
|
PilzAdam joined #minetest-dev |
| 14:28 |
|
nyuszika7h joined #minetest-dev |
| 14:55 |
|
doserj joined #minetest-dev |
| 15:44 |
|
Calinou joined #minetest-dev |
| 15:49 |
|
hmmmm joined #minetest-dev |
| 15:54 |
PilzAdam |
I suggest a vote wether we should add this to upstream: https://github.com/celeron55/minetest_game/pull/33 |
| 15:54 |
PilzAdam |
*whether |
| 15:55 |
Calinou |
nice tracking avatar you have there |
| 15:56 |
Calinou |
no cotton, PilzAdam? :sadface: |
| 15:56 |
PilzAdam |
cotton can be easily added |
| 15:56 |
Calinou |
PilzAdam: your wheat textures. |
| 15:56 |
Calinou |
they're from minecraft |
| 15:56 |
PilzAdam |
no, they are from RealBadAngel |
| 15:57 |
Calinou |
they're still from minecraft |
| 16:01 |
thexyz |
rba is so awesome, he can relicense minecraft's textures under wtfpl |
| 16:03 |
Calinou |
rba is jeb in disguise |
| 16:07 |
hmmmm |
i thought certain textures were released under a really unrestrictive license |
| 16:07 |
|
sstrandberg joined #minetest-dev |
| 16:07 |
hmmmm |
besides, mojang is aware of minetest.. if they worry about their texture being used, all it takes is a simple email |
| 16:10 |
Calinou |
not really |
| 16:10 |
Calinou |
it's not because someone asked permission they forever learned about it |
| 16:11 |
Calinou |
even a minecraft splash says "Don't bother with the clones!" |
| 16:12 |
hmmmm |
no i mean they are aware of minetest in particular; that is, they asked someone if they could use a texture used in minetest released under wtfpl |
| 16:12 |
hmmmm |
there was a forum post about it |
| 16:12 |
hmmmm |
but nevermind this, it's not engine development chat and i feel bad for being so off-topic |
| 16:13 |
hmmmm |
i ought to start idling in #minetest-delta as well |
| 16:39 |
|
sstrandberg joined #minetest-dev |
| 17:00 |
|
VanessaE joined #minetest-dev |
| 17:11 |
|
sstrandberg joined #minetest-dev |
| 17:34 |
|
rubenwardy joined #minetest-dev |
| 17:34 |
|
doserj joined #minetest-dev |
| 17:41 |
sfan5 |
wtf? |
| 17:41 |
sfan5 |
how did realbadangel push this https://github.com/celeron55/minetest/commit/8540871ebcfca6584051e1fc042951baa0ed3275 ? |
| 17:42 |
PilzAdam |
https://github.com/celeron55/minetest/commit/8c3ffa3907dae0c16bfe63965ed6380c7e89df84 |
| 17:42 |
sfan5 |
:/ |
| 17:43 |
sfan5 |
there should be extra notes if the commit comes from a pull request |
| 17:44 |
PilzAdam |
thats a GItHub thing |
| 17:44 |
VanessaE |
and it's irritating. it clutters up the commit history |
| 17:44 |
VanessaE |
it should just be one entry: authored by X, merged by Y. |
| 17:58 |
celeron55 |
for single commits, i tend to git remote add foo https://github.com/foo/minetest.git; git fetch foo; git cherry-pick <commit>; <test that it works>; git push |
| 17:58 |
celeron55 |
maybe you should too |
| 18:00 |
celeron55 |
works up to a couple of commits by running cherry-pick sequentially for the commits; after that the overhead of a merge isn't that much but i tend to do a rebase then |
| 18:02 |
celeron55 |
a matter of taste, really; and a matter of whether one wants to see the inter-github-repo merges visually in github's crappy visualizers |
| 18:02 |
celeron55 |
(nobody wants to) |
| 18:17 |
|
sstrandberg joined #minetest-dev |
| 18:20 |
Calinou |
<test that it works>: unknown command |
| 18:20 |
Calinou |
^ too few FOSS developers installed the package that lets you use that command. :P |
| 18:23 |
celeron55 |
s/<test that it works>/make -j2 && bin/minetest/ |
| 18:24 |
darkrose |
forgot to escape the / in the path |
| 18:24 |
celeron55 |
oops, well 8) |
| 18:25 |
Calinou |
"make -j2 && bin/minetest/" won't work if you're doing that from src |
| 18:25 |
Calinou |
your binary is not in src/bin, it's in bin/ |
| 18:25 |
VanessaE |
s:<test that it works>:make -j2 && bin/minetest/: |
| 18:25 |
VanessaE |
:-) |
| 18:25 |
Calinou |
make -j2 && cd ../bin && ./minetest |
| 18:25 |
Calinou |
make sure you run the binary from bin |
| 18:25 |
* PilzAdam |
uses make -j16 |
| 18:25 |
Calinou |
make -j8 here, quadcore hyperthreaded cpu |
| 18:26 |
VanessaE |
-j12 here, out of 2*#cores habit :-) |
| 18:26 |
celeron55 |
Calinou: minetest isn't generally built in src |
| 18:26 |
celeron55 |
either the root or something completely different |
| 18:27 |
celeron55 |
also, real computer geeks buy as many cores as they have .cpp files in their projects |
| 18:27 |
celeron55 |
make -j for the win |
| 18:27 |
VanessaE |
hah |
| 18:28 |
Calinou |
-j is borked |
| 18:28 |
Calinou |
makes compilation much slower for some reason.. |
| 18:28 |
celeron55 |
of course it does |
| 18:28 |
celeron55 |
you for example run completely out of memory |
| 18:28 |
Calinou |
2012: cannot detect how many CPU cores a CPU has |
| 18:28 |
Calinou |
the make developers(R) |
| 18:28 |
celeron55 |
If the -j option is given without an argument, make will not limit the number of jobs that can run simultaneously. |
| 18:29 |
celeron55 |
it isn't supposed to even try |
| 18:29 |
celeron55 |
also there are many cases when make isn't cpu-bound |
| 18:30 |
Calinou |
well that's stupid imo, they should make it use all cpu cores |
| 18:30 |
celeron55 |
in which case limiting arbitrarily by cpu cores wouldn't make any sense |
| 18:30 |
Calinou |
so that newbies with quadcore cpus aren't slowed down by using one core |
| 18:30 |
celeron55 |
well that is an another thing, and it's mostly a distro thing |
| 18:31 |
celeron55 |
or, well, dunno |
| 18:31 |
celeron55 |
whatever |
| 18:31 |
celeron55 |
off to something more interesting -> |
| 18:32 |
Calinou |
"nothing" is more interesting! |
| 18:32 |
thexyz |
newbies usually don't compile from source |
| 18:32 |
Calinou |
what when you tell them to do so |
| 18:32 |
thexyz |
tell them -j4 |
| 18:32 |
Calinou |
slows down people with singlecore cpus |
| 18:32 |
Calinou |
(and dualcore without hyperthreading) |
| 18:32 |
thexyz |
i don't give a shit |
| 18:32 |
PilzAdam |
people with one core are alrady slowed down |
| 18:36 |
hmmmm |
usually it's recommended to use number_of_cores * 2 + 1 for the argument to -j |
| 18:36 |
Calinou |
why +1? |
| 18:37 |
hmmmm |
iunno, i'd say it depends a lot on the type of scheduler your OS uses |
| 18:37 |
darkrose |
and some jobs aren't cpu bound, if their io bound +1 will help speed things up |
| 18:37 |
darkrose |
s/their/they're/ |
| 18:38 |
hmmmm |
darkrose: exactly! :) that's the reasoning behind my threaded mapgen thing |
| 18:38 |
hmmmm |
nobody has any feedback/comments/suggestions on that i see |
| 18:39 |
VanessaE |
you're always talking about it when no one's around :-) |
| 18:39 |
hmmmm |
well that's because i'm working on it when i talk about it |
| 18:39 |
hmmmm |
perhaps i just need to get the idea out in text because it helps me structure my thoughts |
| 18:40 |
VanessaE |
it's a good idea actually |
| 18:42 |
celeron55 |
i have a dualcore with no HT and -j2 vs. -j3 is (just tested with a random project) 0m33.882s vs. 0m34.562s |
| 18:42 |
hmmmm |
i think it can wait though, the whole multithreading thing is strictly an enhancement that i personally can't even benefit from (i have a single core) |
| 18:42 |
celeron55 |
that is, -j2 is better |
| 18:43 |
celeron55 |
(on this thing) |
| 18:43 |
Calinou |
"(just tested with a random project)" is that project named minetest :OP |
| 18:43 |
Calinou |
:P* |
| 18:43 |
celeron55 |
no |
| 18:44 |
hmmmm |
indeed, it's probably better because the third thread keeps thrashing the cores' caches if it bounces between cores |
| 18:45 |
celeron55 |
maybe |
| 18:45 |
hmmmm |
oh, i needed to ask you about text on signs |
| 18:45 |
Calinou |
next step: compile with -j and swappiness 100 |
| 18:45 |
hmmmm |
have you attempted that at all? |
| 18:45 |
Calinou |
no :) |
| 18:45 |
hmmmm |
how about images on signs or 'pictures? |
| 18:46 |
Calinou |
images on signs would be awesome; but how are users supposed to upload them? |
| 18:46 |
hmmmm |
iunno, how does minecraft do it? |
| 18:46 |
Calinou |
no images on signs on minecraft |
| 18:46 |
VanessaE |
thexyz did that, pilzadam followed up with his own rewrite of that mod. |
| 18:46 |
Calinou |
colored text yes -- images no |
| 18:46 |
celeron55 |
hmmmm: i haven't attempted at all, and don't really have any opinions |
| 18:46 |
Calinou |
(and you need plugins to type colored text on signs) |
| 18:46 |
VanessaE |
(technically, text on signs = images on signs) |
| 18:47 |
hmmmm |
well i was watching a minecraft video the other day, and i saw the orthogonal projection of a node on a sign |
| 18:47 |
Calinou |
I'm talking about pictures |
| 18:47 |
Calinou |
hmmmm: that's an item frame |
| 18:47 |
hmmmm |
and i thought "holy crap that's cool" |
| 18:47 |
Calinou |
not the same thing |
| 18:47 |
hmmmm |
item frame? |
| 18:47 |
hmmmm |
what's that |
| 18:47 |
Calinou |
right click with any item in hand -- item is put in item frame |
| 18:47 |
Calinou |
maps/clocks/compasses are shown as usual |
| 18:47 |
hmmmm |
ah |
| 18:47 |
Calinou |
rotate the item inside with right click |
| 18:48 |
hmmmm |
well maybe we don't need that in particular |
| 18:49 |
hmmmm |
but it seems pretty easy to do - you just get the sign texture, get the INVENTORY_<namehere>_RTT texture, transparent blit the the latter onto the former, and there you go |
| 18:50 |
hmmmm |
that would be really helpful for lots of chests and i'm sure the modding people would find it useful too |
| 18:50 |
|
OdnetninI joined #minetest-dev |
| 18:50 |
hmmmm |
and in the same manner, images can be put on the signs, just that the selection of images to do that with would be limited to what's on the server |
| 18:50 |
celeron55 |
minetest already has a quite flexible and (C++-)extendable "texture modification" thing; adding different kinds of things to it is easy |
| 18:51 |
hmmmm |
but as for the "how are users going to upload pictures" thing - that could come later as it has a lot of potential for abuse and needs a much more elegant solution |
| 18:51 |
celeron55 |
it's still largely undocumented |
| 18:52 |
Calinou |
talking about textures: https://github.com/celeron55/minetest/issues/179 < thoughts? |
| 18:52 |
hmmmm |
i'll check it out. that's the second thing on my list of stuff to do |
| 18:52 |
celeron55 |
tile.{h,cpp} |
| 18:52 |
hmmmm |
i am doing ServerMap cleaning |
| 18:52 |
hmmmm |
isn't the pure white texture already there with clouds? |
| 18:52 |
Calinou |
it'd be awesome if the texmod thingy had the ability to recolor textures, btw :P |
| 18:52 |
Calinou |
hmmmm: well cloud.png does that, but what about pure black and transparent? |
| 18:53 |
PilzAdam |
Calinou, just allow "#000000" in tile definition |
| 18:53 |
celeron55 |
you can basically infinitely append "^[modifier:parameters"-like strings to image names, and it'll take care of doing the modifications to the image |
| 18:53 |
Calinou |
PilzAdam: oh, you can recolor textures? didn't know |
| 18:53 |
celeron55 |
and cache the result |
| 18:53 |
Calinou |
what about rotations/flipping though? |
| 18:53 |
celeron55 |
the visible text sign mods use that to blit images of letters on a base image |
| 18:54 |
celeron55 |
(afaik) |
| 18:54 |
hmmmm |
so it seems pretty simple |
| 18:54 |
hmmmm |
shouldn't take too long since all the hardware is there already |
| 18:54 |
hmmmm |
but i shouldn't call it "easy" just yet because i thought the mapgen replacement was going to be easy until i realized how much unrelated code needed to be reworked for it to fit |
| 18:56 |
Calinou |
hmmmm: any screenshots of the new mapgen? |
| 18:57 |
hmmmm |
not yet |
| 18:57 |
hmmmm |
it doesn't look pretty just yet :) |
| 18:58 |
OdnetninI |
hmmmm, are you working in a new mapgen? |
| 18:58 |
hmmmm |
you don't really want a screenshot, it's just a crapload of stone everywhere because i need to figure out what combination of values works the best when defining biomes in lua |
| 18:59 |
hmmmm |
odentninI, yeah, along with a bunch of other stuff at the same time, it's pretty much done though |
| 19:00 |
OdnetninI |
Can you add mines, and quarry? |
| 19:00 |
hmmmm |
...mines? you mean like caves? caves are still in there from the previous mapgen |
| 19:00 |
hmmmm |
dunno what you mean by quarry |
| 19:01 |
Calinou |
quarry = huge vertical holes |
| 19:01 |
OdnetninI |
mines, how a real mine, not caves |
| 19:01 |
VanessaE |
bbl |
| 19:01 |
hmmmm |
ehh... that sort of stuff is best left to the Lua people |
| 19:01 |
PilzAdam |
so you want Minecraft? |
| 19:01 |
Calinou |
http://media-mcw.cursecdn.com/thumb/a/ae/2011-12-02_12.18.51.png/300px-2011-12-02_12.18.51.png < what MC players do sometimes |
| 19:01 |
Calinou |
oops |
| 19:01 |
Calinou |
http://media-mcw.cursecdn.com/a/ae/2011-12-02_12.18.51.png |
| 19:02 |
hmmmm |
yikes |
| 19:05 |
Calinou |
(unrelated to dev, but funny, and funny things go to -dev) what the fu is this: http://minetest.net/forum/viewtopic.php?pid=51862#p51862 |
| 19:06 |
hmmmm |
WTF. sounds like he's begging to be ignored by people |
| 19:07 |
Calinou |
(more related to dev) http://minetest.net/forum/viewtopic.php?pid=51803#p51803 < thoughts? |
| 19:07 |
Calinou |
imo i'd put the random-cubes-when-loading back, as a feature |
| 19:07 |
Calinou |
i really liked it |
| 19:08 |
PilzAdam |
but upscale it to the whole screen |
| 19:08 |
Calinou |
that'd be ugly |
| 19:08 |
Calinou |
keep it to lower-left corner of screen |
| 19:09 |
Calinou |
and maybe scale it up a bit so that it is fixed-scale compared to the screen resolution |
| 19:09 |
PilzAdam |
its acuatlly funny to make a bug to a feature |
| 19:09 |
hmmmm |
looks messy, i think it might a good idea but you'd need to make some kind of black alpha blended box around the text |
| 19:09 |
hmmmm |
so the text stands out |
| 19:09 |
hmmmm |
besides, how long do people actually look at that screen for? |
| 19:09 |
hmmmm |
i mean add it if you'd like... but... |
| 19:10 |
Calinou |
about the background, it should stay black imo |
| 19:10 |
PilzAdam |
we are talking about the node images in the bottom left corner and not about the background image |
| 19:10 |
Calinou |
don't copy a minecraft feature that isn't really good :p |
| 19:10 |
PilzAdam |
+1 for black |
| 19:11 |
hmmmm |
it's up to personal preference |
| 19:11 |
hmmmm |
for these kinds of things, instead of forcing your preference on everybody else, it ought to be conservative and stay what it was before |
| 19:11 |
|
jin_xi joined #minetest-dev |
| 19:11 |
Calinou |
if people don't agree, make it an option, but keep a sensible default |
| 19:12 |
Calinou |
white text is easier to read on black background than on light-brown background |
| 19:12 |
Calinou |
so black background would be a sensible default here |
| 19:36 |
OdnetninI |
hmmmm, what do you change to mapgen? |
| 19:36 |
hmmmm |
everything |
| 19:36 |
OdnetninI |
how what? |
| 19:36 |
hmmmm |
it's all new except for the cave code |
| 19:37 |
OdnetninI |
can i test it? |
| 19:37 |
hmmmm |
it's not ready yet |
| 19:37 |
hmmmm |
when i need testing i'll ask for testing |
| 19:37 |
hmmmm |
don't test me, i'm quite testy already |
| 19:37 |
hmmmm |
in fact this entire project is just a test |
| 19:38 |
hmmmm |
...of mines |
| 19:38 |
OdnetninI |
mmm |
| 21:54 |
|
OdnetninI left #minetest-dev |
| 23:46 |
|
sstrandberg joined #minetest-dev |
