| Time |
Nick |
Message |
| 00:03 |
SoniEx2 |
<SoniEx2> make world saving/loading scripted |
| 00:03 |
SoniEx2 |
<SoniEx2> if it isn't already |
| 00:03 |
SoniEx2 |
<SoniEx2> that way you can test your algorithms and stuff |
| 00:03 |
SoniEx2 |
(that was like 6 hours ago btw) |
| 01:41 |
SoniEx2 |
so is everyone ignoring me or something? |
| 01:46 |
RealBadAngel |
i have answered and you said nothing... |
| 01:47 |
SoniEx2 |
RealBadAngel, I'm asking about scripted world saving/loading algorithms |
| 01:48 |
SoniEx2 |
and everyone is like completely ignoring me |
| 01:48 |
SoniEx2 |
as if... uhh idk... |
| 01:49 |
SoniEx2 |
I said it twice and got no comments at all |
| 01:50 |
RealBadAngel |
well, lookin at what was happenin in the channel i can guess that no one was here |
| 01:51 |
RealBadAngel |
no wonder, its 4am for many of us ;) |
| 01:54 |
SoniEx2 |
I originally posted the thing over 8 hours ago...? |
| 01:55 |
SoniEx2 |
anyway |
| 01:55 |
SoniEx2 |
that's my suggestion |
| 02:04 |
Exio4 |
i don't think that is very nice for the performance of map saving-loading |
| 02:11 |
VanessaE |
SoniEx2: it's not that everyone's ignoring you, it's that virtually no one is here right now that can comment |
| 02:13 |
SoniEx2 |
Exio4, well anything that makes it so you don't have to recompile the whole thing every time you want to test new map encoding would be good |
| 02:14 |
SoniEx2 |
even if it requires LuaJIT |
| 02:14 |
SoniEx2 |
(altho I'm not sure how I feel about features that require LuaJIT and stuff) |
| 02:16 |
|
werwerwer_ joined #minetest-dev |
| 02:27 |
Exio4 |
i still think it isn't a nice idea to have scripts in _that_ place, maybe in some debug mode only |
| 02:30 |
SoniEx2 |
Exio4, well it would be like a million times more flexible and would make it easier to play with world saving/loading |
| 02:32 |
SoniEx2 |
which means it would be easier to play with compression algorithms or use your own save format so you can bridge minetest and your own game etc |
| 02:42 |
SoniEx2 |
(the latter being the main reason why I want this) |
| 02:43 |
SoniEx2 |
anyway, I'm going to bed |
| 02:43 |
SoniEx2 |
good night |
| 02:43 |
|
BrandonReese joined #minetest-dev |
| 03:05 |
|
proller joined #minetest-dev |
| 03:50 |
|
khonkhortisan_ joined #minetest-dev |
| 04:27 |
|
nore joined #minetest-dev |
| 04:30 |
|
jin_xi joined #minetest-dev |
| 05:08 |
|
kaeza joined #minetest-dev |
| 05:12 |
|
troller joined #minetest-dev |
| 05:37 |
|
Hunterz joined #minetest-dev |
| 05:52 |
|
Hunterz1 joined #minetest-dev |
| 05:59 |
|
darkrose joined #minetest-dev |
| 06:17 |
|
LemonLake joined #minetest-dev |
| 06:22 |
|
CraigyDavi_ joined #minetest-dev |
| 06:26 |
|
troller joined #minetest-dev |
| 06:26 |
|
CraigyDavi__ joined #minetest-dev |
| 06:31 |
|
proller joined #minetest-dev |
| 06:37 |
|
proller joined #minetest-dev |
| 06:37 |
|
Calinou joined #minetest-dev |
| 07:20 |
|
VanessaE joined #minetest-dev |
| 07:26 |
VanessaE |
ok wtf? |
| 07:26 |
VanessaE |
suddenly I've got mapblocks disappearing in front of me |
| 07:26 |
VanessaE |
already-loaded blocks |
| 07:26 |
VanessaE |
if I turn around and look at them again, they come back |
| 07:27 |
VanessaE |
seems to happen if I shorten the view range too far |
| 07:27 |
VanessaE |
like 25 or so |
| 07:28 |
VanessaE |
yet the point of disappearance is more like 5 meters in front of me, not 25. |
| 07:28 |
|
troller joined #minetest-dev |
| 07:44 |
|
Calinou joined #minetest-dev |
| 07:45 |
|
iqualfragile joined #minetest-dev |
| 08:17 |
|
daswort joined #minetest-dev |
| 08:37 |
|
Exio4 joined #minetest-dev |
| 09:09 |
|
proller joined #minetest-dev |
| 09:44 |
|
darkrose joined #minetest-dev |
| 10:01 |
|
Jordach joined #minetest-dev |
| 10:02 |
|
jin_xi joined #minetest-dev |
| 10:19 |
|
proller joined #minetest-dev |
| 11:18 |
|
Gethiox joined #minetest-dev |
| 11:22 |
|
Megaf joined #minetest-dev |
| 11:27 |
|
ImQ009 joined #minetest-dev |
| 12:24 |
|
proller joined #minetest-dev |
| 12:25 |
Exio4 |
SoniEx2: million times more flexible also means million times slower, doesn't it? |
| 12:27 |
|
kaeza joined #minetest-dev |
| 12:27 |
celeron55 |
the main issue with lua for that case is it's dynamic nature |
| 12:27 |
celeron55 |
you only know if it's going to work by running it, and if you risk messing up world data for regular users, that sucks a lot |
| 12:28 |
celeron55 |
currently the C++ implementation which, while it uses a format full of legacy crap, is very solid |
| 12:28 |
celeron55 |
-which |
| 12:29 |
celeron55 |
maybe there should be some way to override it via some simple Lua interface though; if it's simple and not used by default, it won't hurt anyone |
| 12:37 |
|
kaeza joined #minetest-dev |
| 13:10 |
sfan5 |
I think we should add C++ modding support |
| 13:13 |
celeron55 |
who's going to build a version of each C++ mod for linux-x86-gcc, linux-x86-clang, linux-arm-gcc, linux-arm-clang, android-arm-gcc, android-arm-clang, windows-mingw, windows-msvc and whatnot? |
| 13:13 |
celeron55 |
or wherever the ABI compatibility happens to break (with C++, it breaks almost everywhere) |
| 13:14 |
* sfan5 |
shows his hand |
| 13:14 |
sfan5 |
also clang & gcc have the same C++ ABI |
| 13:14 |
Megaf |
I want this! http://www.cavium.com/OCTEON-III_CN7XXX.html |
| 13:15 |
Megaf |
"OCTEON III also features a revolutionary, low latency coherency architecture that enables multiple OCTEON III chips to appear as a single logical high-performance processor with up to 384 cores, providing up to 960GHz compute, up to 800+ Gbps of application performance and up to 2 Terabytes of memory capacity." |
| 13:15 |
Megaf |
That's the CPU minetest needs |
| 13:15 |
* sfan5 |
throws a http://cubox-i.com/ at Megaf |
| 13:15 |
Megaf |
several cores that can be use as a single thing |
| 13:15 |
Megaf |
MIPS Power! |
| 13:16 |
celeron55 |
Megaf: how is this appropriate for #-dev |
| 13:16 |
Megaf |
How is this not? |
| 13:17 |
Megaf |
What if I manage to get a board with that SoC and run minetest in it? |
| 13:17 |
* Megaf |
runs |
| 13:17 |
sfan5 |
<general stuff> <general stuff> <minetest>; |
| 13:17 |
celeron55 |
Megaf: it would be completely meaningless |
| 13:17 |
sfan5 |
that is 66% non-minetest |
| 13:17 |
celeron55 |
Megaf: now, go talk on #minetest instead |
| 13:17 |
celeron55 |
sfan5: i don't believe you building those; absolutely every person on this planet is too lazy to do that |
| 13:18 |
sfan5 |
qemu vm with build tools, every dev gets 10 minutes ssh session, auto-resetting |
| 13:18 |
sfan5 |
or even an automatic build bot for approved mods |
| 13:19 |
|
hmmmm joined #minetest-dev |
| 13:19 |
Megaf |
celeron55: we might need a #Minetest-OffTopic if we want to be so rigid about things we can talk |
| 13:19 |
sfan5 |
wat |
| 13:19 |
celeron55 |
Megaf: this isn't #minetest-offtopic |
| 13:19 |
celeron55 |
Megaf: go away already |
| 13:19 |
sfan5 |
#minetest is #minetest-offtopic |
| 13:19 |
sfan5 |
(just that it is about minetest too) |
| 13:20 |
celeron55 |
Megaf: also, you can make that channel any second you want; i don't own freenode |
| 13:20 |
sfan5 |
anyway, to continue my pointless discussion: what about embedding a c++ compiler into minetest? |
| 13:20 |
sfan5 |
</joke> |
| 13:21 |
Megaf |
someone is not in a good mood, is there any way I can help you? Minetest core is doing great lately, it's been a week without a single crash on my servers |
| 13:21 |
sfan5 |
num. crashes on your server != status of minetest core |
| 13:22 |
celeron55 |
sfan5: if you can make a system where every minetest user can use every mod like currently and any modder can build their mod for all those systems without any effort at all, then fine; of course assuming that you will also figure out how to extract a stable C++ API to the minetest core for them to use |
| 13:22 |
sfan5 |
I don't want to replace lua mods |
| 13:22 |
sfan5 |
but sometimes it would be nice to change stuff in c++ directly |
| 13:22 |
celeron55 |
anyone being able to do that just seems fully impossible, |
| 13:22 |
celeron55 |
-, |
| 13:23 |
sfan5 |
it would also be way faster to have e.g. mob AI in C++ |
| 13:23 |
celeron55 |
sfan5: how are you going to "change stuff in C++" without a huge effort on a C++ API which will convolute the internals way more than they have ever been? |
| 13:23 |
sfan5 |
hm |
| 13:24 |
celeron55 |
this is just completely infeasible |
| 13:24 |
sfan5 |
I guess you're right |
| 13:24 |
Megaf |
Did anyone discuss about javascript already? I'm just talking about that because JS has been improving a lot, several big names are puting a lot of money and time on JS |
| 13:24 |
celeron55 |
if you do it, then fine; but i am not going to put any effort on that because it is impossible |
| 13:24 |
sfan5 |
Megaf: https://www.destroyallsoftware.com/talks/wat - then come back |
| 13:24 |
Megaf |
ok |
| 13:25 |
celeron55 |
Megaf: lua is better than javascript in various ways in this context, on addition to making such a change being just a ridiculous amount of work |
| 13:25 |
sfan5 |
I think what I want is somehow removing the overhead that remains even with LuaJIT |
| 13:26 |
celeron55 |
you should measure what makes that overhead |
| 13:26 |
celeron55 |
otherwise you're just shouting useless comments around |
| 13:26 |
sfan5 |
lua type checks? unpacking stuff from tables? |
| 13:26 |
celeron55 |
maybe you will know some day |
| 13:27 |
sfan5 |
I think I'm going to do more productive stuff now |
| 13:27 |
celeron55 |
i personally don't know what kind of overheads there are, and i can't even guess |
| 13:28 |
celeron55 |
and this is not a priority for me |
| 13:29 |
sfan5 |
not for me either |
| 13:29 |
Megaf |
sfan5: I'm lmao with that presentation |
| 13:30 |
sfan5 |
you see |
| 13:30 |
sfan5 |
this is why JS is bad |
| 13:30 |
sfan5 |
additionally: https://wiki.theory.org/YourLanguageSucks#JavaScript_sucks_because |
| 13:35 |
Kray |
that article sucks |
| 13:39 |
celeron55 |
i like javascript |
| 13:39 |
celeron55 |
but i like lua too |
| 13:39 |
celeron55 |
and simultaneously hate both too |
| 13:39 |
Kray |
is this article supposed to be funny even though it isn't, or is this just plain bad |
| 13:39 |
celeron55 |
they are the class of languages which have many good and bad sides to them |
| 13:39 |
Kray |
there's so much subjective and/or questionable points that's not really useful |
| 13:40 |
celeron55 |
and no, browser vendors won't do any work for us; they aren't interested |
| 13:40 |
celeron55 |
on the other hand lua is focused on being "the" embeddable language |
| 13:41 |
celeron55 |
we should probably write these things down on a wiki page so we could link it to anyone trying to argue about this |
| 13:42 |
sfan5 |
>Irrlicht log: Irrlicht Engine version 1.8.1 |
| 13:42 |
sfan5 |
this is annoying |
| 13:43 |
sfan5 |
can we shorten it to something like [IRR] ? |
| 13:47 |
celeron55 |
yes |
| 13:49 |
|
BrandonReese joined #minetest-dev |
| 14:24 |
|
vifino joined #minetest-dev |
| 14:28 |
|
LemonLake joined #minetest-dev |
| 14:34 |
RealBadAngel |
sfan5, no problemo |
| 14:35 |
RealBadAngel |
https://github.com/minetest/minetest/blob/master/src/main.cpp#L296 |
| 14:36 |
|
BlockMen joined #minetest-dev |
| 14:41 |
BlockMen |
before it gets completely lost, i have rebase #542 now and updated/imroved it with the second commit here: https://github.com/BlockMen/minetest/commits/item_drop |
| 14:41 |
BlockMen |
can it get pushed? |
| 14:41 |
BlockMen |
RealBadAngel^ |
| 14:50 |
|
smoke_fumus joined #minetest-dev |
| 14:51 |
|
NakedFury joined #minetest-dev |
| 14:56 |
|
PilzAdam joined #minetest-dev |
| 15:06 |
|
kaeza joined #minetest-dev |
| 15:07 |
|
tomasbrod joined #minetest-dev |
| 15:11 |
|
Calinou joined #minetest-dev |
| 15:17 |
|
Piggybear87 joined #minetest-dev |
| 15:20 |
|
kaeza joined #minetest-dev |
| 15:30 |
|
kaeza joined #minetest-dev |
| 15:32 |
|
zat joined #minetest-dev |
| 15:37 |
|
Megaf_ joined #minetest-dev |
| 15:37 |
|
spillz joined #minetest-dev |
| 15:42 |
RealBadAngel |
BlockMen, ofc |
| 15:54 |
|
Garmine joined #minetest-dev |
| 16:05 |
BlockMen |
ok, then i gonna push it within next minutes |
| 16:06 |
RealBadAngel |
btw, https://github.com/minetest/minetest/blob/master/builtin/mainmenu/tab_settings.lua#L96 |
| 16:06 |
RealBadAngel |
what it is doing here? |
| 16:08 |
BlockMen |
was it removed already? if so, i guess it came back with the formspec toolkit |
| 16:15 |
RealBadAngel |
propably youre right |
| 16:16 |
RealBadAngel |
i already had to re-add generate normals setting thx to that commit |
| 16:19 |
|
tomreyn joined #minetest-dev |
| 16:19 |
|
Hunterz joined #minetest-dev |
| 16:23 |
|
tomreyn joined #minetest-dev |
| 16:28 |
|
spillz joined #minetest-dev |
| 16:35 |
|
Calinou joined #minetest-dev |
| 16:39 |
|
zat joined #minetest-dev |
| 16:39 |
|
sapier joined #minetest-dev |
| 16:46 |
sapier |
maybe we should write down somewhere WHY we don't believe in a c/c++ mod interface ... we wouldn't have to discuss it once a month |
| 16:49 |
celeron55 |
that and also the javascript thing |
| 16:49 |
celeron55 |
those get asked all the time |
| 16:50 |
sapier |
well for the javascript thingy I'd just answer "write it" ... since I separated scriptapi It's just a lot of typewriting ;-) |
| 16:51 |
sapier |
I don't believe in anyone completing that task |
| 16:52 |
|
rubenwardy joined #minetest-dev |
| 16:52 |
sfan5 |
if we are honest, a pull request for javascript scriptapi would not get merged anyway |
| 16:53 |
sapier |
well if you'd write it including all builtin code AND switchable on compiletime it may be merged |
| 16:53 |
celeron55 |
there's no chance in this world that it would be good enough |
| 16:53 |
celeron55 |
it would be a mess |
| 16:53 |
sapier |
but I'd guess you'd be finished in about half a year full time |
| 16:53 |
celeron55 |
but if it wasn't and it was beautiful and nicely switchable and all, then yes |
| 16:53 |
celeron55 |
it just won't happen |
| 16:54 |
sapier |
noone spends half a year to a year development time for what benefit? none? |
| 16:55 |
sapier |
you'd have to implement all those fancy runtime optimization javascript things yourself as none of them (I know) is license compatible to lgpl2.1 either |
| 16:57 |
celeron55 |
V8 is BSD licensed |
| 16:57 |
celeron55 |
so that's not a problem |
| 16:58 |
SoniEx2 |
I have a question |
| 16:59 |
SoniEx2 |
why not have multiple ABI DLLs |
| 16:59 |
SoniEx2 |
like 1.0 1.1 etc |
| 16:59 |
celeron55 |
probably the easiest way to add javascript support would be making a lua->V8 wrapper library and building minetest with it instead of regular lua |
| 16:59 |
SoniEx2 |
then you call the DLLs |
| 16:59 |
SoniEx2 |
instead of doing whatever |
| 17:00 |
celeron55 |
SoniEx2: for what is this a solution? |
| 17:00 |
sapier |
a wrapper for something already claimed to be slow? |
| 17:01 |
SoniEx2 |
celeron55, it doesn't stop development? and the DLL would be more of a way to request what to do instead of asking the mod to do stuff? |
| 17:01 |
SoniEx2 |
hmm wait that would be kinda scripting with C++... |
| 17:01 |
celeron55 |
i don't know what you are trying to solve with that |
| 17:01 |
SoniEx2 |
ok uhh |
| 17:02 |
SoniEx2 |
you tell the DLL to setup some pointers |
| 17:02 |
sapier |
but still could someone please spend time on doing the really important things? I spent the whole last week to get about 5-10% performance ... I guess someone knowing irrlicht could get a lot more within that time |
| 17:02 |
SoniEx2 |
and then you don't have to care about Minetest's internal ABI |
| 17:02 |
SoniEx2 |
as the DLL will autoupdate to the internal ABI |
| 17:02 |
SoniEx2 |
while keeping the same ABI |
| 17:02 |
sapier |
SoniEx2: ok and you're gonna implement a dll loader for linux |
| 17:02 |
SoniEx2 |
if that makes sense |
| 17:03 |
SoniEx2 |
sapier, doesn't linux use .so or something? |
| 17:03 |
celeron55 |
to me this makes zero sense; let's just move on 8) |
| 17:03 |
sapier |
once you did this plz integrate wine to minetest and later maintain all the api changes |
| 17:03 |
tomasbrod |
SoniEx2, yes an there could be Makefile in mod dir to compile the mod to .dll or .so |
| 17:03 |
sapier |
you're right celeron |
| 17:03 |
SoniEx2 |
ok uhh |
| 17:04 |
SoniEx2 |
the DLL is basically a dynamic function pointer array |
| 17:04 |
SoniEx2 |
and you set it up |
| 17:04 |
sapier |
any dynamic code requires a abi |
| 17:04 |
SoniEx2 |
the DLL is the ABI |
| 17:04 |
sapier |
and that's something changed about once a week in minetest |
| 17:04 |
SoniEx2 |
the DLL stays the same |
| 17:04 |
sapier |
NO |
| 17:05 |
SoniEx2 |
Minetest's internal ABI changes |
| 17:05 |
sapier |
the dll is just a piece of code |
| 17:05 |
SoniEx2 |
Minetest calls the DLL to tell the DLL where the functions are |
| 17:05 |
SoniEx2 |
mods use the DLL to get where the functions are |
| 17:05 |
sapier |
write it soniex2 |
| 17:05 |
SoniEx2 |
the DLL stays the same |
| 17:05 |
sapier |
once you're done we can talk about it |
| 17:05 |
SoniEx2 |
uhh ok let's think about it as Lua |
| 17:05 |
SoniEx2 |
you have a table |
| 17:05 |
sapier |
No |
| 17:05 |
SoniEx2 |
and functions |
| 17:06 |
SoniEx2 |
you wanna rename a function you just put it in the same place on the table |
| 17:06 |
sapier |
I don't need your explanation how dlls work |
| 17:06 |
SoniEx2 |
and the table would be an array |
| 17:06 |
SoniEx2 |
so t[1] = somefunc |
| 17:06 |
SoniEx2 |
then you change somefunc's name to something like copytable |
| 17:06 |
SoniEx2 |
and then t[1] = copytable |
| 17:06 |
sapier |
just write it and after you did succeed we can talk about it again |
| 17:06 |
SoniEx2 |
think about the name as the ABI location |
| 17:07 |
SoniEx2 |
t[1] is the location in the DLL |
| 17:07 |
celeron55 |
this talk is useless |
| 17:07 |
SoniEx2 |
sure you would have to first get the thing from the array and then call it |
| 17:07 |
SoniEx2 |
but still |
| 17:07 |
SoniEx2 |
it shouldn't be too slow |
| 17:07 |
SoniEx2 |
and it should be faster than Lua |
| 17:07 |
SoniEx2 |
and I'm not sure if you understand my logic |
| 17:07 |
SoniEx2 |
:/ |
| 17:08 |
sapier |
write your code we'll try to understand it once it's working |
| 17:08 |
|
tomasbrod left #minetest-dev |
| 17:10 |
SoniEx2 |
uhh meh :/ |
| 17:14 |
kaeza |
I've seen Guile being mentioned a couple of times too |
| 17:16 |
kaeza |
everybody seems to want to add random languages to the engine, instead of doing it the KISS way and simply actually f*cking learning Lua |
| 17:16 |
celeron55 |
yeah it's like adding support for a language would be much easier than learning a language |
| 17:17 |
celeron55 |
you can learn enough lua for modding in a single day :P |
| 17:17 |
sfan5 |
people usually don't have time |
| 17:17 |
celeron55 |
but people are just laaaaaaaaaaaaaaaaaaaaaaaazy |
| 17:17 |
kaeza |
^ |
| 17:18 |
kaeza |
sfan5, no time to learn but time to implement yet another binding? |
| 17:18 |
sfan5 |
yes |
| 17:19 |
sapier |
can someone plz find a way to speed up v3s32 constructor by factor 2? we'd save a significant amount of time on this ;-) |
| 17:24 |
sfan5 |
sapier: rewrite it in assembler |
| 17:25 |
Megaf |
sapier: How can the client disconnect be improved? I noticed that even if a client just disconnect properly, either by clicking on back to menu or back to os, sometimes (quite orften) I will get a time out error for that client on server side. |
| 17:26 |
celeron55 |
the disconnect packet is easily lost |
| 17:26 |
sapier |
that's a tradeoff, right now client is optimized for quick shutdown on local game |
| 17:26 |
sapier |
that increases risk of packet beeing lost even more |
| 17:26 |
celeron55 |
the other option would be to have the client wait for reply from the server, which would mean very slow shutdown if the server is unresponsive |
| 17:27 |
celeron55 |
maybe there should be some background shutdown packet send repeater for this purpose? 8) |
| 17:27 |
celeron55 |
8D |
| 17:27 |
sapier |
as there's no negative consequence except of that message I'd prefere not to add dirty hacks |
| 17:27 |
celeron55 |
for TCP that already happens because the OS does it |
| 17:27 |
sapier |
that's why you have to open a socket in reuse mode |
| 17:28 |
kaeza |
"This program requires internet access to quit. Please connect to the internet and try again." |
| 17:28 |
* kaeza |
runs |
| 17:28 |
kaeza |
sounds somewhat Microsoft-ish |
| 17:29 |
celeron55 |
on a more serious note, maybe the client could stop for 500ms at disconnect of non-local game and then send a disconnect packet again? |
| 17:29 |
sapier |
better just don't reduce timeout as much as we do now for non local games |
| 17:30 |
celeron55 |
altough, that wouldn't probably help except in some cases |
| 17:30 |
celeron55 |
i think this isn't an important issue |
| 17:31 |
sapier |
megaf plz write a feature request for this as I don't consider it to be as important to require immediate action. If this is fixed it's done as addon to something more important |
| 17:32 |
Megaf |
sapier: I agree, it's not urgent. I would just be nice to fix that in a nice way, not hacky way. |
| 17:34 |
Megaf |
sapier: I think right now you could just change the timeout from ERROR to INFO |
| 17:34 |
Megaf |
to actually get less attention than it deserves |
| 17:34 |
sapier |
maybe that's gonna be the final solution too ... there may not be a really clean AND fast way to solve that issue |
| 17:35 |
Megaf |
indeed |
| 17:35 |
Megaf |
I'm gonna open a ticket anyway |
| 17:36 |
sapier |
yes |
| 17:36 |
sapier |
sometimes issues are fixed even years after they have been written ... e.g. vanessae's shutdown menu glitch |
| 17:37 |
Megaf |
That's quite common actually, |
| 17:38 |
sapier |
crashes are usually fixed way more fast ;-) |
| 17:48 |
Megaf |
#1360 |
| 17:50 |
Megaf |
sapier: how to I change this "errorstream<<m_connection->getDesc()" to INFO? |
| 17:50 |
|
blaise joined #minetest-dev |
| 17:50 |
Megaf |
Issue #1360 |
| 17:51 |
sapier |
error --> info |
| 17:52 |
Megaf |
Isnt it LOG(dout_con<<m_connection->getDesc() ? |
| 17:54 |
Megaf |
I will try here, hold on |
| 17:57 |
sapier |
no |
| 17:57 |
sapier |
dout_con is debug log |
| 17:57 |
sapier |
not info |
| 17:57 |
sapier |
and the LOG macro arround is to fix a race condition |
| 18:00 |
Megaf |
oh, ok |
| 18:05 |
|
zat joined #minetest-dev |
| 18:08 |
Megaf |
sapier: I'm trying to test this little change. but I can't reproduce a time out |
| 18:09 |
Megaf |
even killing the client dont cause a time out |
| 18:10 |
Megaf |
15:09:50: ACTION[ServerThread]: test times out. List of players: |
| 18:10 |
Megaf |
:) |
| 18:18 |
Megaf |
sapier: https://github.com/minetest/minetest/pull/1361 |
| 18:21 |
Megaf |
https://github.com/minetest/minetest/issues/1360#issuecomment-45651586 |
| 18:21 |
Megaf |
My first pull request :) |
| 18:22 |
|
werwerwer joined #minetest-dev |
| 18:51 |
|
Piggybear87 joined #minetest-dev |
| 18:55 |
|
nore joined #minetest-dev |
| 18:57 |
Megaf |
sapier: just changing that from errorstream to infostream worked great, 15:56:42: ACTION[ServerThread]: Megaf times out. List of players: moi |
| 18:58 |
Megaf |
That's good beucase now I can have smaller log files and only with relevant info |
| 18:58 |
sapier |
good ;-) |
| 19:02 |
RealBadAngel |
rotfl, by simple typo i made all the dirts nodes wave instead of water |
| 19:02 |
RealBadAngel |
that looks really cool :) |
| 19:03 |
VanessaE |
haha |
| 19:03 |
|
asl joined #minetest-dev |
| 19:06 |
Megaf |
RealBadAngel: hah, I'd like to see that, where you did that typo and what typo you did? |
| 19:14 |
VanessaE |
RealBadAngel: that reminds me, how is that latest shader work coming along? |
| 19:14 |
RealBadAngel |
makin a pieces of code look a bit nicer |
| 19:15 |
VanessaE |
hope to see it soon. |
| 19:16 |
RealBadAngel |
propably tommorow |
| 19:17 |
RealBadAngel |
instead of headers in each shader file i moved generation to the engine |
| 19:17 |
|
khonkhortisan joined #minetest-dev |
| 19:18 |
RealBadAngel |
https://github.com/minetest/minetest/pull/1303 |
| 19:18 |
RealBadAngel |
as ShadowNinja requested: http://pastebin.com/ |
| 19:18 |
RealBadAngel |
oops |
| 19:19 |
RealBadAngel |
http://pastebin.com/pWBchm3P |
| 19:24 |
asl |
it be nice if there a way to play sounds generated from codes (+1 on client side), that way playing mml, midi, sid and lots other using mod would be possible |
| 19:25 |
VanessaE |
SID files in Minetest! |
| 19:25 |
VanessaE |
I could just see linking Minetest against sidplay2 :P |
| 19:25 |
VanessaE |
asl: really though, it's a nice idea. I guess you saw Inocudom's and my discussion on the forum. |
| 19:28 |
asl |
um no |
| 19:28 |
asl |
link me? |
| 19:29 |
VanessaE |
https://forum.minetest.net/viewtopic.php?f=11&t=2041&start=325#p144148 |
| 19:32 |
asl |
the main reason i thought of it is cause of mml, players can input their own songs and let other people listen to it |
| 19:32 |
asl |
i still can't find a good mml player though :P |
| 19:36 |
sapier |
you know all those files have to be transfered to all clients on initial join? |
| 19:37 |
VanessaE |
sapier: which is why the discussion in the forum turned to MIDI and the like, those files are usually much smaller than an equivalent OGG/MP3. |
| 19:37 |
VanessaE |
(and which is why my "stereo" object in homedecor does not have any sound) |
| 19:38 |
sapier |
midi? you know about midi being beep sound ;-) |
| 19:38 |
VanessaE |
sapier: and you clearly never heard a really good MIDI composition then :P |
| 19:38 |
asl |
mml is just text, it can be set in the meta |
| 19:39 |
VanessaE |
look up one called "06twelve" (spelled exactly so) and play it with some really GOOD equipment or with Timidity++ and EAWpats or freepats as your patch set. you'll wonder if you're still hearing a MIDI piece. |
| 19:39 |
VanessaE |
(its somewhat comedic, mostly instrumental version of "12 days of christmas") |
| 19:40 |
sapier |
maybe ... no matter if good or not I don't see a technical reason against midi |
| 19:42 |
VanessaE |
nope, shouldn't be hard |
| 19:43 |
VanessaE |
link against timidity++, send the music like any other piece of media, make sure the client knows how to tell timidity to play it |
| 19:43 |
VanessaE |
libmikmod for mod/xm/s3m etc. files |
| 19:44 |
celeron55 |
having that kind of stuff in MT would certainly be quite unique, but i wonder if it's what users generally want though |
| 19:44 |
VanessaE |
resid/libsidplay0/sidplay2 for .sid/(.mus,.str) files (since asl mentioned SID music) |
| 19:44 |
asl |
too much link library |
| 19:44 |
VanessaE |
what's the player used for Nintendo NSF files? |
| 19:44 |
asl |
i am sure the sound can be generated from lua |
| 19:44 |
VanessaE |
celeron55: it's hard to say for sure what users would or would not want. you just never know :P |
| 19:44 |
asl |
just need a way to pipe it to the speaker |
| 19:44 |
celeron55 |
when there is client-side lua, it can generate waveform sound whatever way it wants |
| 19:45 |
celeron55 |
but... only then |
| 19:45 |
celeron55 |
so get coding guys |
| 19:45 |
VanessaE |
lua generating realtime digital audio? |
| 19:45 |
VanessaE |
you're insane. |
| 19:46 |
celeron55 |
insane? not even close |
| 19:46 |
VanessaE |
celeron55: been there, done that with assembly, got the headaches to prove it. you don't want to go there, trust me. |
| 19:46 |
VanessaE |
just link to some appropriate library for that purpose. |
| 19:47 |
celeron55 |
luajit is not far from native in crunching raw numbers like that |
| 19:48 |
celeron55 |
but of course the feasibility depends on what kind of end result is good enough |
| 19:49 |
Megaf |
sapier: Wouldnt be a good idea to marge #1361 ? |
| 19:49 |
Megaf |
merge marge |
| 19:51 |
sapier |
it's to hot to merge here ;-) |
| 19:51 |
celeron55 |
VanessaE: actually, it doesn't matter what users want |
| 19:52 |
celeron55 |
VanessaE: i should have said "but i wonder if it's actually useful in the end" |
| 19:52 |
|
ImQ009 joined #minetest-dev |
| 19:52 |
celeron55 |
because, if you provide something, then your userbase will just shift towards those who like that particular thing you provided |
| 19:52 |
Megaf |
sapier: ok, let's let it cool down then. |
| 19:53 |
celeron55 |
there are users for anything that is useful, but not for things that aren't 8) |
| 19:55 |
VanessaE |
celeron55: yes, but please be fair: your definition of "useful" doesn't always mesh well with the rest of the users :) |
| 19:56 |
VanessaE |
of course in-game music isn't particularly useful to me either, since I always have my sound turned off, but I know it does add a nice aesthetic for some folks. |
| 19:56 |
VanessaE |
in this case the point is smaller media downloads than would be possible with ogg/mp3 and the like. |
| 19:56 |
VanessaE |
that is always useful. |
| 19:58 |
VanessaE |
that leave the question of what kind of condensed format is appropriate for Minetest beyond ogg and so forth. that's a question I can neither answer from a technical point of view (because I just don't know) nor from a personal one (because I am heavily biased) |
| 20:00 |
celeron55 |
i like music myself, as long as it's consistent with whatever the game puts me to do otherwise |
| 20:00 |
celeron55 |
and see |
| 20:01 |
celeron55 |
if it ends up sounding too oldschool due to being midi or some tracker stuff, it won't really match many things |
| 20:01 |
celeron55 |
or, potentially might not be able to |
| 20:01 |
celeron55 |
but on the other hand our artistic production standards are rather low anyway so it's just impossible to know |
| 20:02 |
celeron55 |
streaming or transferring something like ogg is too much bandwidth usage for small servers, that's for sure |
| 20:02 |
celeron55 |
and for many users on the receiving end |
| 20:04 |
celeron55 |
if someone implements something like that nicely and demonstrates some kind of music with it that works well in a subgame, i'll probably accept it |
| 20:04 |
celeron55 |
because it just might work best in this weird technical context |
| 20:04 |
VanessaE |
actually RBA did do something similar |
| 20:04 |
Megaf |
Once I thought of using the VanessaE's homedecor mod toiled sound (minetest)bug to play music on my server, I would play low bitrate ogg files using the toilet |
| 20:05 |
VanessaE |
remember those tunes he wanted to add? Skaven I think was the author? |
| 20:05 |
celeron55 |
i will not judge any tunes in the context of "minetest" |
| 20:05 |
celeron55 |
i will judge them in the context of a subgame for which they are intended by someone |
| 20:06 |
celeron55 |
and so should others |
| 20:07 |
|
BlockMen left #minetest-dev |
| 20:07 |
celeron55 |
i think rba's thing was just playing oggs in the menu, which were rendered from some old songs |
| 20:08 |
VanessaE |
damn it, I can't remember the name of the song |
| 20:08 |
VanessaE |
but they were rendered from IT/XM or MOD or some such |
| 20:08 |
VanessaE |
one of those module tracker formats |
| 20:09 |
VanessaE |
I keep thinking something about mosquitos |
| 20:09 |
* asl |
want mml included too |
| 20:11 |
VanessaE |
anyway the point is, that's some really decent music that still holds up well. |
| 20:22 |
VanessaE |
change of subject: if I have multiple servers (which I do :P )... could I hard-link ipbans.txt between them and expect it to work right? |
| 20:22 |
VanessaE |
e.g. ban from one server = banned from all of them |
| 20:29 |
Megaf |
by |
| 20:29 |
Megaf |
bye |
| 20:33 |
celeron55 |
VanessaE: you can't |
| 20:33 |
VanessaE |
didn't think so. bummer |
| 20:34 |
celeron55 |
maybe it could be modified so that you could; dunno if there are some conflicting priorities though |
| 20:38 |
celeron55 |
if someone makes a reliable-looking patch for that, i guess it could be merged |
| 20:38 |
|
Anchakor1 joined #minetest-dev |
| 20:39 |
|
Garmine42 joined #minetest-dev |
| 20:42 |
|
mrtux joined #minetest-dev |
| 20:47 |
|
proller joined #minetest-dev |
| 20:52 |
|
EvergreenTree joined #minetest-dev |
| 20:57 |
|
luizrpgluiz_ joined #minetest-dev |
| 20:58 |
|
luizrpgluiz_ left #minetest-dev |
| 21:07 |
|
werwerwer joined #minetest-dev |
| 21:10 |
|
cg72 joined #minetest-dev |
| 21:14 |
|
luizrpgluiz joined #minetest-dev |
| 21:15 |
luizrpgluiz |
hi,because the site is down? |
| 21:16 |
SoniEx2 |
someone make LPeg built-in |
| 21:19 |
luizrpgluiz |
mean that someone is trying to attack the site? |
| 21:22 |
celeron55 |
uh oh, the whole host is down, and it's not even a very small one |
| 21:23 |
celeron55 |
looks like it came up now |
| 21:27 |
luizrpgluiz |
:D |
| 21:28 |
|
luizrpgluiz left #minetest-dev |
| 22:04 |
Sokomine |
hi. got a problem with a relatively recent built: when stuck inside of something (i.e. after spawning a building or a tree growing on you) the screen becomes pitch dark - with no way to control anything at all |
| 22:04 |
Sokomine |
this is pretty problematic. has that been reported and solved already? |
| 22:05 |
Sokomine |
one game spawned me underground, and i had the same problem |
| 22:05 |
Sokomine |
not even chat works then |
| 22:05 |
sapier |
hmmm I remember having that issue too |
| 22:06 |
sapier |
thought it'S been something with my android build |
| 22:23 |
|
khonkhortisan joined #minetest-dev |
| 22:51 |
|
kaeza joined #minetest-dev |
| 23:01 |
|
sapier left #minetest-dev |
| 23:12 |
|
khonkhortisan joined #minetest-dev |
| 23:12 |
|
deltib joined #minetest-dev |
| 23:53 |
|
NakedFury joined #minetest-dev |
