| Time |
Nick |
Message |
| 00:37 |
VanessaE |
well the thing is, I'm not so sure the distance sorting works. |
| 00:37 |
VanessaE |
I've seen stuff in the distance load before stuff closer to the player does |
| 00:54 |
sapier |
vanessae block priorization is done by querying blocks if a block is in there once It's gonna be loaded so if you turn distant blocks in old direction may be loaded prior to near ones in new direction |
| 00:55 |
sapier |
https://github.com/sapier/minetest/pull/new/security_v2 maybe security can be added if it's disabled by default :-( |
| 00:57 |
|
sapier left #minetest-dev |
| 01:16 |
|
VanessaE joined #minetest-dev |
| 02:04 |
|
hmmmm joined #minetest-dev |
| 03:26 |
|
kmmmmm joined #minetest-dev |
| 03:26 |
kmmmmm |
hello |
| 03:27 |
|
kmmmmm left #minetest-dev |
| 04:01 |
|
ffoxin joined #minetest-dev |
| 04:34 |
|
ffoxin_entry joined #minetest-dev |
| 06:00 |
|
ffoxin_work joined #minetest-dev |
| 07:45 |
|
proller joined #minetest-dev |
| 07:58 |
|
ecube joined #minetest-dev |
| 08:20 |
celeron55 |
< sapier> off topic I've added seek support to fileapi some days ago |
| 08:20 |
celeron55 |
wtf |
| 08:21 |
celeron55 |
you should just wrap the regular lua stuff but only allow opening files from a certain location |
| 08:22 |
celeron55 |
limiting what operations someone can do on a file is not security, it's just idiotic |
| 08:24 |
celeron55 |
like: function open_secure_file(name) if not filename_is_secure(name) then return nil end return io.open(safe_directory.."/"..name) end |
| 08:25 |
* celeron55 |
shakes his head once again |
| 08:25 |
proller |
celeron55, https://github.com/celeron55/minetest/pull/492 |
| 08:35 |
celeron55 |
thexyz: re. http://irc.minetest.ru/minetest/2013-02-16#i_2876764 - fixing this should be your first priority because most probably it was caused by the freetype patch |
| 08:36 |
celeron55 |
even reverting is better than not fixing it in a few days |
| 08:37 |
celeron55 |
because it's very bad |
| 08:45 |
thexyz |
any idea how to reproduce this? |
| 08:45 |
thexyz |
http://pastebin.com/NbxEtUs5 |
| 08:48 |
celeron55 |
umm... you need to run the game, not only the menu |
| 08:49 |
thexyz |
I did |
| 08:49 |
thexyz |
oh, well, apparently, wrong log |
| 08:51 |
thexyz |
still, it works |
| 08:52 |
thexyz |
ok, I can also test it in windows |
| 08:52 |
thexyz |
once I install one |
| 08:52 |
celeron55 |
i can't try anything at the moment, but i could give it a try in 12 hours |
| 08:54 |
celeron55 |
hmm, actually, i *can* try it |
| 08:54 |
celeron55 |
because it should be able to fail as a server too |
| 08:55 |
celeron55 |
oh by the way |
| 08:55 |
celeron55 |
po/ro/minetest.po: warning: Charset "CHARSET" is not a portable encoding name. |
| 08:55 |
celeron55 |
po/ko/minetest.po: warning: Charset "CHARSET" is not a portable encoding name. |
| 08:56 |
thexyz |
those weren't changed at all |
| 08:56 |
thexyz |
http://translate.minetest.ru/projects/minetest/core/ |
| 08:56 |
thexyz |
I guess I'll just remove them, right? |
| 08:58 |
celeron55 |
apparently the headers in the new translations are not filled in |
| 08:58 |
celeron55 |
the old ones have UTF-8 as charset and a proper author and date and whatever |
| 08:58 |
thexyz |
yes, because those are not "new translations" but minetest.pot copied |
| 08:59 |
thexyz |
weblate adjust headers as soon as someone touches the translation |
| 08:59 |
celeron55 |
ok; just let them be then |
| 09:02 |
celeron55 |
when i do LC_ALL=fi_FI bin/minetestserver, i get |
| 09:02 |
celeron55 |
bin/minetestserver: warning: could not set default locale |
| 09:02 |
celeron55 |
11:05:21: ERROR[main]: Could not set numeric locale to C |
| 09:02 |
celeron55 |
but it doesn't fail |
| 09:03 |
celeron55 |
shit, i don't even have fi_FI, that explains |
| 09:04 |
|
rubenwardy joined #minetest-dev |
| 09:10 |
thexyz |
it works fine on russian winxp for me too |
| 09:10 |
celeron55 |
by quick testing, it doesn't seem to fail |
| 09:11 |
celeron55 |
could it be some strange combination of not having UTF-8 locales and not having locales with . as decimal points? |
| 09:11 |
celeron55 |
or... something |
| 09:14 |
thexyz |
lua has its own os.setlocale so the workaround for this problem may be to add os.setlocale('C') somewhere |
| 09:15 |
thexyz |
to the top of builtin.lua maybe |
| 09:15 |
thexyz |
too bad PilzAdam is not here |
| 09:17 |
thexyz |
http://pastebin.com/zcQbw7DP |
| 09:30 |
rubenwardy |
what is wrong with the main loop? |
| 09:30 |
rubenwardy |
" badly structured main loop (which works well, though). " |
| 09:35 |
rubenwardy |
also, is there an up-to-date todo list anywhere? |
| 09:36 |
rubenwardy |
ping: thexyz, celeron55 |
| 09:37 |
thexyz |
sure there is |
| 09:37 |
thexyz |
https://github.com/celeron55/minetest/issues?state=open |
| 09:39 |
rubenwardy |
oh, off course |
| 09:39 |
rubenwardy |
of |
| 09:40 |
|
celeron55 joined #minetest-dev |
| 09:48 |
thexyz |
celeron55: apparently, it fixes the issue https://github.com/celeron55/minetest/issues/494 |
| 09:51 |
|
jin_xi joined #minetest-dev |
| 09:54 |
|
darkrose joined #minetest-dev |
| 09:54 |
|
proller joined #minetest-dev |
| 09:59 |
celeron55 |
thexyz: good |
| 09:59 |
thexyz |
I'm commiting it then, right? |
| 10:00 |
celeron55 |
i guess there is no reason not to |
| 10:00 |
celeron55 |
hmm |
| 10:00 |
celeron55 |
you might want to ask tforsman to try running the C++ unit tests |
| 10:01 |
celeron55 |
they test some Settings stuff that should make sure that decimal points work properly in there too |
| 10:03 |
celeron55 |
actually; the main configuration has so many values with decimal points that tem failing should be surely noticeable |
| 10:03 |
celeron55 |
them* |
| 10:44 |
|
nil_ joined #minetest-dev |
| 10:54 |
rubenwardy |
does this go in a .h or a .cpp? http://dev.minetest.net/Mapgen |
| 11:10 |
celeron55 |
.h, altough you'll probably want to implement stuff like MapgenTest::makeChunk in a .cpp |
| 11:15 |
rubenwardy |
http://pastebin.com/5BFCXJPv |
| 11:15 |
rubenwardy |
woops |
| 11:15 |
rubenwardy |
http://pastebin.com/QypS4ghU |
| 11:16 |
rubenwardy |
nm |
| 11:16 |
rubenwardy |
hmmm |
| 11:16 |
rubenwardy |
11:16:36: ERROR[main]: EmergeManager: mapgen "plus" not registered |
| 11:16 |
rubenwardy |
Segmentation fault (core dumped) |
| 11:18 |
celeron55 |
lol, it segfaults if a mapgen is not found? |
| 11:19 |
celeron55 |
if you add it with registerMapgen("plus", new MapgenFooFactory()), it shouldn't be able to fail that way |
| 11:20 |
thexyz |
celeron55: have you seen https://github.com/celeron55/minetest/pull/492 yet? |
| 11:47 |
|
ffoxin joined #minetest-dev |
| 12:52 |
celeron55 |
thexyz: what are your opinions on it as-is? |
| 12:57 |
celeron55 |
i can see some minor details to be fixed, but as a non-player and a non-server-maintainer i can't really judge |
| 12:59 |
celeron55 |
my main doubts are about security, as it that makes it completely trivial to get many players to connect to anything, not even requiring any "social engineering" |
| 12:59 |
celeron55 |
-it |
| 13:07 |
thexyz |
celeron55: isn't that true about any server list done that way? |
| 13:08 |
proller |
internet is unsecure by design |
| 13:12 |
celeron55 |
maybe we should try to break client security before taking that into use |
| 13:13 |
celeron55 |
just to make sure doing it isn't too small a matter |
| 13:19 |
celeron55 |
there are practically three points that could be exploitable |
| 13:21 |
celeron55 |
1) the minetest protocol, 2) external libraries handling stuff from the server like libpng and libogg, 3) irrlicht's handling of data coming from the server (models; is there anything else?) |
| 13:23 |
celeron55 |
1 is probably best handled by just looking through everything for obvious bad data handling (or generating some randomized packets if that much effort is used) |
| 13:24 |
celeron55 |
2 by looking up the versions commonly built with minetest and seeing if there are known exploits for them |
| 13:24 |
celeron55 |
3 probably similarly to 1 |
| 13:44 |
* celeron55 |
is reading through Client::ProcessData() |
| 13:45 |
celeron55 |
there are way too many older things done with bare indices in here; would be a good investment to convert them to the later stringstream way |
| 13:47 |
celeron55 |
also, even more than that, there are separetely read and buffered deserializations while later the istream-reading readX() functions were added |
| 13:48 |
celeron55 |
i refuse to even attempt to try to proofread this 8) |
| 13:57 |
celeron55 |
looking into 2 now... |
| 14:02 |
celeron55 |
this isn't particularly convincing |
| 14:07 |
|
pierreghz joined #minetest-dev |
| 14:10 |
|
hmmmm joined #minetest-dev |
| 14:16 |
celeron55 |
not too bad, but there's no magic here; overflow bugs have been found in roughly everything |
| 14:16 |
celeron55 |
irrlicht's model loading code looks like something that could contain many problems without being immediately apparent |
| 14:17 |
celeron55 |
because there's a lot of it and it isn't in such large use as libpng/jpeg/ogg/vorbis/whatever is |
| 14:18 |
RealBadAngel |
hi all |
| 14:18 |
RealBadAngel |
celeron55, will you do something more with main menu layout? |
| 14:19 |
celeron55 |
i'm quessing we'll have to find the problems in the hard way; i won't be surprised at all if irrlicht or the MT protocol prove easily vulnerable |
| 14:19 |
RealBadAngel |
coz i noticed lately one thing, when played in full screen. theres no exit to os from menu... |
| 14:19 |
celeron55 |
RealBadAngel: no i won't |
| 14:21 |
RealBadAngel |
too bad :( btw i have a problem with rotating textures for nodeboxes |
| 14:22 |
celeron55 |
i'm not disallowing others from doing it though |
| 14:22 |
celeron55 |
which should be obvious) |
| 14:22 |
celeron55 |
+( |
| 14:22 |
RealBadAngel |
flipping with use size and *-1 to size doesnt work there |
| 14:23 |
celeron55 |
you can't rotate textures using texture coordinates |
| 14:23 |
celeron55 |
you have to rotate them using vertex indices or vertices itself |
| 14:23 |
RealBadAngel |
but when i need flip and rotation too? |
| 14:23 |
RealBadAngel |
one can have with vertices |
| 14:24 |
celeron55 |
i don't understand what you are saying |
| 14:24 |
RealBadAngel |
lemme show you code for regular nodes |
| 14:25 |
RealBadAngel |
http://pastebin.com/2XguFdB9 |
| 14:25 |
RealBadAngel |
this is how i made rotations |
| 14:26 |
RealBadAngel |
vertex part work with nodeboxes, flipping doesnt |
| 14:27 |
celeron55 |
i don't get what you mean by vertex part and flipping part |
| 14:27 |
RealBadAngel |
rotations are made with vertices |
| 14:28 |
RealBadAngel |
flipping with texture size |
| 14:28 |
RealBadAngel |
tile.texture.pos.Y += tile.texture.size.Y; |
| 14:28 |
RealBadAngel |
tile.texture.size.Y *= -1; |
| 14:28 |
RealBadAngel |
this is flip over X |
| 14:29 |
RealBadAngel |
it does work for texturing regular nodes |
| 14:29 |
celeron55 |
why do you need to do that |
| 14:29 |
RealBadAngel |
because of facedir rotations |
| 14:30 |
RealBadAngel |
some tiles require flip and rotation together |
| 14:30 |
celeron55 |
why would they |
| 14:31 |
RealBadAngel |
because it is so |
| 14:31 |
RealBadAngel |
i havent invented it |
| 14:32 |
celeron55 |
can't you do a flip and a rotation by using vertices only? |
| 14:33 |
RealBadAngel |
teoretically i could |
| 14:33 |
celeron55 |
that might make some backface culling problems, or alternatively your current way does and it wouldn't |
| 14:33 |
RealBadAngel |
but then texture is not correctly shown |
| 14:33 |
RealBadAngel |
yes |
| 14:33 |
RealBadAngel |
i got texture visible inside of cuboid |
| 14:34 |
RealBadAngel |
and invisible from outside |
| 14:35 |
celeron55 |
i still don't understand why it might work like that - it would seem natural for it to work right by just rotating vertices |
| 14:35 |
RealBadAngel |
it doesnt |
| 14:35 |
celeron55 |
maybe something flips them somewhere hidden from you |
| 14:36 |
celeron55 |
or maybe you should actually rotate 180 degrees more rather than flipping, or something 8) |
| 14:36 |
RealBadAngel |
when i do flip with vertices result is ok |
| 14:36 |
RealBadAngel |
but shown on the wrong side of face |
| 14:36 |
RealBadAngel |
from inside the box |
| 14:36 |
celeron55 |
you already said that, why are you repeating it |
| 14:37 |
RealBadAngel |
to point i tried everythin |
| 14:37 |
RealBadAngel |
notice im tryin to avoid ^transform |
| 14:38 |
celeron55 |
anyway, i don't have any solutions to you |
| 14:39 |
RealBadAngel |
i will ask another way, is texturing nodeboxes any way different than regular nodes? |
| 14:42 |
RealBadAngel |
there was an issue with coords of boxes, that they have to be sorted |
| 14:42 |
RealBadAngel |
maybe that is somehow related? |
| 14:44 |
RealBadAngel |
nvm, i will find out |
| 14:46 |
RealBadAngel |
btw, somebody would mind if i make texturing every single box not just whole nodebox list? |
| 14:48 |
RealBadAngel |
thats actually easy to make |
| 14:50 |
jin_xi |
different testures for each box in a nodebox model? yes, i think that would be nice |
| 15:02 |
VanessaE |
RealBadAngel: +1000 |
| 15:06 |
RealBadAngel |
consider it done then |
| 15:08 |
RealBadAngel |
VanessaE, have you tried my facedir tree? |
| 15:08 |
VanessaE |
not yet. |
| 15:09 |
VanessaE |
doing it now. |
| 15:09 |
RealBadAngel |
pick the lua mod too |
| 15:10 |
VanessaE |
master branch of your repo? |
| 15:10 |
RealBadAngel |
you can use it on nodeboxes too, just textures are badly rotated on them |
| 15:10 |
RealBadAngel |
yes |
| 15:11 |
RealBadAngel |
i need an idea how to make it work for nodeboxes so good as for regular nodes |
| 15:12 |
RealBadAngel |
for small nodeboxes its hard to pick proper face to rotate it |
| 15:12 |
RealBadAngel |
thx to also small selection boxes |
| 15:13 |
VanessaE |
override the default selection box when using the screwdriver/crowbar then |
| 15:13 |
RealBadAngel |
selection box is defined with node |
| 15:15 |
RealBadAngel |
but i can transform it, maybe i can also alter it at all? will have to try |
| 15:16 |
VanessaE |
RealBadAngel: see pm |
| 15:34 |
|
Calinou joined #minetest-dev |
| 15:38 |
VanessaE |
is there a particular reason the server no longer prints a mod's missing dependencies to the terminal on failure? |
| 15:39 |
Calinou |
failed asserts |
| 15:39 |
VanessaE |
10:34:32: ERROR[main]: The following mods have unsatisfied dependencies: technic |
| 15:39 |
VanessaE |
ok, it wasn't hard to figure out what it wanted. |
| 15:40 |
VanessaE |
but why is the engine/server/whatever no longer bothering to tell me WHAT was missing? |
| 15:40 |
Calinou |
I get failed asserts when having a missing dep when starting a dedicated server :P |
| 15:40 |
Calinou |
exiting gracefully would be better |
| 15:46 |
celeron55 |
VanessaE: that's probably done by doserj's mod selection UI changes |
| 15:46 |
celeron55 |
VanessaE: ask him |
| 15:46 |
celeron55 |
or whoever that was |
| 15:46 |
VanessaE |
probably. |
| 15:56 |
VanessaE |
I'll make a point of mentioning this when I see him next. |
| 15:58 |
|
ffoxin joined #minetest-dev |
| 16:03 |
|
PilzAdam joined #minetest-dev |
| 16:06 |
|
ffoxin joined #minetest-dev |
| 16:26 |
|
ffoxin joined #minetest-dev |
| 16:43 |
PilzAdam |
celeron55, http://minetest.net/download.php <- I use mingw too; maybe add that my builds have gettext and ttf support |
| 17:11 |
celeron55 |
mingw added; i don't really want to flood that page with information so i won't add anything else |
| 17:12 |
Calinou |
is there any use for links to 0.4.3/0.4.1/0.3.1? |
| 17:13 |
celeron55 |
dunno; but i consider it good practice because sometimes people do need old versions |
| 17:13 |
celeron55 |
nothing is more annoying than not being able to get one when you need it :P |
| 17:14 |
PilzAdam |
Calinou, AFAIK there are still some 0.3.1 server |
| 17:14 |
VanessaE |
...which is really surprising |
| 17:14 |
VanessaE |
(redcrab maintains one, madchicken also I think) |
| 17:14 |
celeron55 |
0.3.1 is a relatively solid game |
| 17:14 |
celeron55 |
not a good community sandbox, but a game |
| 17:15 |
Calinou |
redcrab runs a 0.4_20120106 server too |
| 17:16 |
celeron55 |
it's quite different in philosophy than 0.4; i'd really like it if somebody made an appropriately solid and updated (in the right spirit) fork out of it |
| 17:16 |
Calinou |
noone goes to that server today :P |
| 17:16 |
Calinou |
a modless 0.4.4 is pretty much an improved 0.3.1 |
| 17:16 |
Calinou |
noone forces you to install mods |
| 17:16 |
celeron55 |
not really |
| 17:17 |
VanessaE |
well there's the caveat that you'd need a mobs mod and something to re-add jungle trees. |
| 17:17 |
VanessaE |
if you wanted a 0.4.4 that was at least equivalent to 0.3.x |
| 17:17 |
Calinou |
I have a simplemobs mod which puts oerkki/rats and DMs back |
| 17:17 |
Calinou |
still lags like hell most of the time :) |
| 17:17 |
Calinou |
(I forked pilzadam's simplemobs to remove the dirt/stone/sand monsters, nothing else) |
| 17:17 |
Calinou |
and celeron55 made a branch which adds jungles and dungeons back (why wasn't it merged?) |
| 17:18 |
VanessaE |
because moretrees is superior to default jungles ;) |
| 17:18 |
celeron55 |
because there was so much going on with the mapgen at that time |
| 17:18 |
PilzAdam |
hasnt hmmmm added mg_flags to readd jungles an dungenons? |
| 17:18 |
celeron55 |
i might do that some day now (if nobody else does) |
| 17:18 |
celeron55 |
he did? |
| 17:19 |
PilzAdam |
https://github.com/celeron55/minetest/commit/97260d09a8e658b6ded7d9100b0aa1040fcfe50f |
| 17:19 |
PilzAdam |
seems they have no use currently |
| 17:19 |
Calinou |
what about jungles? |
| 17:20 |
Calinou |
and when you enable the flag, do they actually generate? |
| 17:20 |
PilzAdam |
I think forests are jungles |
| 17:21 |
Calinou |
do they have jungle trees? /P |
| 17:21 |
PilzAdam |
<PilzAdam> seems they have no use currently |
| 17:36 |
|
hmmmm joined #minetest-dev |
| 19:50 |
|
ffoxin_entry joined #minetest-dev |
| 20:23 |
|
BackupCoder joined #minetest-dev |
| 20:32 |
F00 |
... |
| 20:33 |
F00 |
With the new physics, going down a ladder with fast move causes 6 damage, doing the same in water does 3 |
| 20:36 |
F00 |
Knee-jerk fix might be "disable fall damage when swimming or climbing", but then there's still the issue of flying ~10 blocks into the air at the top of the ladder/water |
| 21:07 |
|
jin_xi joined #minetest-dev |
| 21:33 |
hmmmm |
oh yeah i didn't do that yet |
| 21:33 |
hmmmm |
they have no current use |
| 21:34 |
hmmmm |
but i know a lot of people liked dungeons and jungles, and they were kind of removed on a whim, so the option for them should at least be there |
| 22:07 |
VanessaE |
so I guess I should drop them from moretrees? |
| 22:08 |
VanessaE |
:( |
| 22:18 |
hmmmm |
sometime later when i actually implement it.. |
| 22:18 |
VanessaE |
ok |
| 22:18 |
hmmmm |
there's a ridiculous amount of stuff i need to do |
| 22:19 |
VanessaE |
maybe by then I'll have tuned the jungle trees to be "just right", and you can use my models :-) |
| 22:19 |
hmmmm |
one problem at a time. right now it's emergethread. |
| 22:19 |
VanessaE |
I understand. There's no rush |
| 22:19 |
hmmmm |
just got home from school, gotta eat, then take a nap, then i'll be able to do something |
| 22:27 |
|
proller joined #minetest-dev |
| 23:41 |
|
proller joined #minetest-dev |
