Time |
Nick |
Message |
00:07 |
|
AnotherBrick joined #minetest-dev |
00:16 |
|
Taoki joined #minetest-dev |
00:43 |
|
Wuzzy joined #minetest-dev |
00:52 |
|
Wayward_Tab joined #minetest-dev |
01:31 |
|
AnotherBrick joined #minetest-dev |
01:38 |
|
Wuzzy2 joined #minetest-dev |
01:48 |
|
Miner_48er joined #minetest-dev |
02:59 |
|
technomancy joined #minetest-dev |
03:01 |
technomancy |
I have a question about whether my idea for a mod would work with the limitations of the minetest engine. I want to do something similar to http://computercraft.info/ where the game adds computer blocks with a simple OS on them, but it seems like this would be extremely problematic because Lua has no support for pre-emptive multitasking out of the box; one problematic script could take down the whole server. |
03:03 |
technomancy |
does minetest do anything to help in this case? does allowing the user to run sandboxed lua code threaten the stability of the server? |
03:03 |
technomancy |
I mean, it's easy to sandbox what resources a given script is allowed to access, except for time. |
03:40 |
|
Hunterz joined #minetest-dev |
03:42 |
|
FR^4 joined #minetest-dev |
03:50 |
|
FR^4 joined #minetest-dev |
04:00 |
|
FR^4 joined #minetest-dev |
04:18 |
est31 |
technomancy, look here https://github.com/Jeija/minetest-mod-mesecons/blob/master/mesecons_luacontroller/init.lua#L368 |
04:18 |
est31 |
its using that method |
04:18 |
technomancy |
est31: iiiiinteresting. that's promising. |
04:19 |
technomancy |
so does that force a yield, or just kill the coroutine? |
04:20 |
est31 |
http://www.lua.org/pil/23.2.html |
04:23 |
technomancy |
I see. In this case it kills it, but it should be possible to simply yield instead. |
04:23 |
est31 |
? |
04:24 |
technomancy |
I would like the execution to be allowed to continue as long as the rest of the engine is given time to run. |
04:25 |
est31 |
as I said, there is no real concurrency |
04:25 |
technomancy |
yeah, I don't need actual concurrency. |
04:26 |
technomancy |
it seems like a forced yield after a max number of instructions can approximate pre-emptive multitasking even if you don't actually have multiple things happening at once |
04:26 |
technomancy |
but at the very least it's possible to kill out-of-control programs, and that may be good enough |
05:06 |
|
cib0 joined #minetest-dev |
05:26 |
technomancy |
https://stackoverflow.com/questions/8459459/lua-coroutine-error-tempt-to-yield-across-metamethod-c-call-boundary could be problematic |
05:26 |
technomancy |
but I don't know enough about minetest's engine to know whether I'd run into this |
05:27 |
technomancy |
oh... apparently might not be a problem with luajit? |
05:27 |
|
Wayward_One joined #minetest-dev |
05:27 |
est31 |
dunno |
05:28 |
est31 |
I don't even know whether coroutines will work with minetest |
05:28 |
|
RealBadAngel joined #minetest-dev |
05:28 |
technomancy |
what luajit version is used? |
05:30 |
|
Wayward_Tab joined #minetest-dev |
05:32 |
est31 |
minetest just uses the lib |
05:32 |
est31 |
so no limitation :) |
05:32 |
est31 |
or there is but still more than one |
05:34 |
est31 |
if I remember right, there is code that warns you if you use buggy luajit |
05:36 |
technomancy |
oh, you provide whatever luajit version you want? |
05:36 |
technomancy |
cool |
05:48 |
|
Hijiri joined #minetest-dev |
05:49 |
|
Hunterz joined #minetest-dev |
06:50 |
|
jin_xi joined #minetest-dev |
07:00 |
|
Darcidride joined #minetest-dev |
07:02 |
jin_xi |
technomancy: others have already tried and done some work, see this thread: https://forum.minetest.net/viewtopic.php?f=47&t=10185 |
07:02 |
jin_xi |
this even resulted in a pull request for keyboard input, which got shut down |
07:03 |
jin_xi |
https://github.com/minetest/minetest/pull/1737 which killed the enthusiasm it seems |
07:03 |
technomancy |
jin_xi: huh... but somehow luacontroller is able to get keyboard input? |
07:03 |
technomancy |
is this about keyboard input outside a form? |
07:03 |
jin_xi |
yes |
07:03 |
jin_xi |
well, outside a text element |
07:04 |
technomancy |
gotcha. yeah, the text element is pretty weak sauce vs what's available in computercraft. =\ |
07:06 |
technomancy |
jin_xi: it's hard to tell what's going on in this conversation without knowing which comments are form core devs. has anyone spoken with authority on this change, or has it just been enough random people shitting on it to make the original author give up? |
07:06 |
technomancy |
*from core devs |
07:07 |
|
blaze joined #minetest-dev |
07:08 |
jin_xi |
the thread starts off badly, but then its two dudes working on a mod. core devs have special green names in the forums |
07:08 |
technomancy |
oh sorry; I meant the github issue |
07:09 |
technomancy |
wow, I didn't realize luacontrollers couldn't have loops and functions =\ |
07:09 |
jin_xi |
hm, sapier used to be a coredev but has not been around |
07:10 |
technomancy |
he is not listed on https://github.com/minetest/minetest/graphs/contributors |
07:11 |
jin_xi |
weird, he did many commits |
07:12 |
technomancy |
"I have pretty much dropped this idea because formspecs are very limited." =( |
07:12 |
jin_xi |
where is that from? |
07:13 |
technomancy |
https://forum.minetest.net/viewtopic.php?id=8399 |
07:13 |
|
nore joined #minetest-dev |
07:14 |
jin_xi |
yeah anything remotely dynamic is very much doomed in minetest |
07:14 |
technomancy |
blargh |
07:14 |
technomancy |
just when I was feeling very accomplished for figuring out how to get sandboxing and pre-emption working |
07:14 |
technomancy |
it looks like one of these guys just switched to computercraft |
07:15 |
technomancy |
but I'd rather just give up than start building on a proprietary platform |
07:15 |
jin_xi |
thats the spirit |
07:16 |
jin_xi |
you can get some dynamic stuff done via stitching together images, it works well enough for a tetris |
07:17 |
technomancy |
for output, yeah |
07:17 |
technomancy |
seems like without this input PR it's dead in the water though |
07:18 |
jin_xi |
i gave it my +1 for what its worth and am in full support for reviving that discussion |
07:18 |
technomancy |
thanks. I'll look into it some more. |
07:19 |
jin_xi |
also sapier is gone and left many a feature done badly |
07:19 |
technomancy |
heh |
07:19 |
jin_xi |
so that argument maybe is not so strong |
07:19 |
technomancy |
it's good to know what's out there. I did some searching but didn't find this discussion. |
07:19 |
nore |
sfan5: can I merge game#511? |
07:19 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/511 -- Use connect_to_raillike for default:rail (2nd try) by Wuzzy2 |
07:21 |
technomancy |
jin_xi: the most encouraging thing is that ninnghazad seems to imply that the debug.sethook approach works to prevent scripts from blocking the rest of the engine |
07:22 |
technomancy |
the most cringeworthy part is that his mod is called "dronetest" |
07:23 |
jin_xi |
lets hope no one makes a wedding mod |
07:24 |
jin_xi |
but from the repo it looks like quite some work has been done |
07:25 |
technomancy |
nothing since december, but yeah. less for me to do. |
07:26 |
technomancy |
I was pretty excited to get my toy little OS working on my own, but I don't mind throwing that away |
07:31 |
|
Calinou joined #minetest-dev |
07:45 |
OldCoder |
sfan5, present? |
07:47 |
|
Zeno` joined #minetest-dev |
07:47 |
OldCoder |
Zeno`, hi |
07:56 |
Zeno` |
hi, how's things? |
08:05 |
|
Yepoleb joined #minetest-dev |
08:09 |
|
kilbith joined #minetest-dev |
08:10 |
|
cib0 joined #minetest-dev |
08:17 |
|
selat joined #minetest-dev |
08:47 |
|
est31 joined #minetest-dev |
09:25 |
|
Megaf_ joined #minetest-dev |
09:45 |
|
cib0 joined #minetest-dev |
09:59 |
|
chchjesus joined #minetest-dev |
10:27 |
|
Calinou joined #minetest-dev |
10:29 |
|
rom1504 joined #minetest-dev |
10:31 |
|
Anchakor_ joined #minetest-dev |
10:32 |
|
jin_xi joined #minetest-dev |
10:33 |
|
Ritchie joined #minetest-dev |
10:34 |
|
Yepoleb joined #minetest-dev |
10:36 |
|
FR^4 joined #minetest-dev |
10:36 |
|
Robby joined #minetest-dev |
10:50 |
|
proller joined #minetest-dev |
10:58 |
|
blaze joined #minetest-dev |
11:17 |
|
cib0 joined #minetest-dev |
11:21 |
|
Zeno` joined #minetest-dev |
11:33 |
|
Etzos joined #minetest-dev |
11:35 |
|
ElectronLibre joined #minetest-dev |
11:39 |
|
psedlak joined #minetest-dev |
11:40 |
|
proller joined #minetest-dev |
11:55 |
|
cib0 joined #minetest-dev |
12:58 |
|
selat joined #minetest-dev |
13:28 |
|
AnotherBrick joined #minetest-dev |
14:11 |
|
Darcidride joined #minetest-dev |
14:27 |
|
proller joined #minetest-dev |
14:46 |
|
err404 joined #minetest-dev |
15:07 |
|
cib0 joined #minetest-dev |
15:23 |
|
Amaz joined #minetest-dev |
15:41 |
|
Krock joined #minetest-dev |
15:43 |
|
AnotherBrick joined #minetest-dev |
15:49 |
|
ElectronLibre left #minetest-dev |
15:54 |
|
proller joined #minetest-dev |
16:07 |
|
proller joined #minetest-dev |
16:25 |
|
jin_xi joined #minetest-dev |
16:31 |
|
RealBadAngel joined #minetest-dev |
16:31 |
RealBadAngel |
hi |
16:31 |
RealBadAngel |
http://i.imgur.com/QhKRhLP.png |
16:32 |
RealBadAngel |
finally i managed to show textures as meshes :) |
16:32 |
|
Hunterz joined #minetest-dev |
16:33 |
RealBadAngel |
easy to rotate them in 3d space and have them visible on screen plane |
16:33 |
RealBadAngel |
that will make rotating items in inventories possible even |
16:35 |
kilbith |
impressive |
16:42 |
RealBadAngel |
that would solve many issues, like hud elements not workin properly with interlacing or other 3d rendering modes |
16:42 |
Krock |
next step: tiny 3d model of the minimap, rotateable by couse drag |
16:42 |
Krock |
*mouse drag |
16:42 |
kilbith |
RBA, can we have a round map instead ? |
16:42 |
Krock |
^ |
16:42 |
RealBadAngel |
thats possible |
16:42 |
RealBadAngel |
kilbith, ofc, thats the main idea |
16:42 |
RealBadAngel |
i will just apply round mask to minimap now |
16:43 |
RealBadAngel |
and the map will rotate when player moves |
16:43 |
RealBadAngel |
another advantage it can be shaded with shaders now |
16:44 |
RealBadAngel |
and i do have stored already height map of the scanned area |
16:44 |
RealBadAngel |
so minimap will be 3d |
16:44 |
RealBadAngel |
or at least will look like 3d :) |
16:44 |
|
proller joined #minetest-dev |
16:50 |
|
proller joined #minetest-dev |
16:54 |
|
Hijiri joined #minetest-dev |
17:00 |
|
TheWild joined #minetest-dev |
17:01 |
|
Wuzzy joined #minetest-dev |
17:04 |
|
ElectronLibre joined #minetest-dev |
17:04 |
|
Wayward_Tab joined #minetest-dev |
17:09 |
|
cib joined #minetest-dev |
17:12 |
|
ElectronLibre joined #minetest-dev |
17:15 |
|
MinetestForFun_ joined #minetest-dev |
17:26 |
Calinou |
the arrow should be less ugly :x |
17:26 |
Calinou |
well, it can be removed if minimap is round and rotates |
17:26 |
Calinou |
instead we'll show a slight grey “+†|
17:43 |
|
proller joined #minetest-dev |
17:46 |
|
proller joined #minetest-dev |
17:56 |
|
Hijiri joined #minetest-dev |
17:58 |
|
younishd joined #minetest-dev |
17:59 |
|
est31 joined #minetest-dev |
18:00 |
kilbith |
the grey "+" doesnt indicate the facing direction |
18:00 |
est31 |
RealBadAngel, any feedback from gsmanners about license? |
18:01 |
RealBadAngel |
no, but i wont need him |
18:01 |
est31 |
? |
18:01 |
RealBadAngel |
nothing of his code left by now ;) |
18:02 |
RealBadAngel |
Calinou, map is already round and rotating :) |
18:02 |
Calinou |
cool |
18:02 |
Calinou |
Rei will be angry :3 |
18:02 |
est31 |
http://programmers.stackexchange.com/questions/260347/ship-of-theseus-applied-to-gpl-can-i-relicense-my-program-if-i-replace-all-of |
18:02 |
RealBadAngel |
est, we are not philosophers :P |
18:03 |
Calinou |
StackExchange is GPL hate party |
18:03 |
* Calinou |
tries to avoid SE as much as possible |
18:03 |
Calinou |
its login also requires third-party cookies |
18:03 |
RealBadAngel |
you will compare the code |
18:03 |
est31 |
Stackexchange puts all content into free license |
18:03 |
Calinou |
it's not like we'll be sued over a minimap anyway |
18:03 |
est31 |
thats better than other forum sites |
18:03 |
RealBadAngel |
literally nothing is as gsmanner did it |
18:03 |
est31 |
not even minetest forum has it anymore |
18:04 |
RealBadAngel |
im using totally different techniques even |
18:04 |
est31 |
what about just simply *asking* him? |
18:04 |
est31 |
perhaps he sais "yes its ok?" |
18:04 |
RealBadAngel |
i did, no answer |
18:08 |
est31 |
when you list him in the copyright boilerplate inside the header then he must have agreed to the according license |
18:09 |
est31 |
either way, if he doesnt agree, then we still can change the license of minetest to GPL. who needs that "lesser" anyway. |
18:10 |
RealBadAngel |
if there wont be any of his code why should i put him into copyright note? i can just mention that he inspired me |
18:11 |
est31 |
yea I think inspiration is fair use |
18:12 |
est31 |
but I guess whether its GPL or LGPL is decided when one actually wants to use the "lesser" part of the license |
18:12 |
est31 |
if* |
18:13 |
est31 |
so, from my POV, its ok to merge the code |
18:13 |
est31 |
because I don't really care whether its LGPL or GPL |
18:13 |
est31 |
as long as its free |
18:14 |
Calinou |
if it stays GPL, we might include it provided we provide a --no-gpl build option |
18:15 |
Hijiri |
LGPL matters when dynamically linking to minetest, right |
18:16 |
Hijiri |
but how would that work |
18:16 |
est31 |
dunno |
18:16 |
est31 |
perhaps mods can be interpreted that way?? |
18:16 |
Hijiri |
as in the mods are dynamically linking to minetest? |
18:17 |
Hijiri |
actually I'm not sure at all how interpreted programs work with GPL and friends |
18:17 |
Calinou |
ask licensingfsf.org |
18:17 |
est31 |
there is an entry in the GPL faq about stuff like this |
18:18 |
est31 |
they say "if communication is intimate enough, its like static linking" |
18:18 |
Hijiri |
I thought it was "if communication is intimate enough, its like using a library" |
18:18 |
est31 |
I mean its similar to the IPC issue which even GPL doesnt impose any restrictions (or freedoms if you are fsf :p) upon. |
18:19 |
Hijiri |
in the context of people trying to run a separate program and use message channels of some sort to avoid it being technically a library |
18:20 |
est31 |
yea |
18:20 |
est31 |
but applies to simpler examples too |
18:20 |
Hijiri |
GPL FAQ says However, when the interpreter is extended to provide “bindings†to other facilities (often, but not necessarily, libraries), the interpreted program is effectively linked to the facilities it uses through these bindings. |
18:21 |
Hijiri |
minetest provides bindings to minetest API things |
18:21 |
Hijiri |
so I assume that means mod scripts count as linked to an LGPL library |
18:21 |
Hijiri |
or GPL if it's GPL |
18:21 |
est31 |
yea I think that has been the reason for the switch |
18:22 |
est31 |
perhaps LGPL isnt that bad after all... |
18:22 |
Hijiri |
which switch? |
18:22 |
|
selat joined #minetest-dev |
18:22 |
est31 |
minetest has originally been GPL if I'm right |
18:22 |
Hijiri |
oh |
18:23 |
Calinou |
yes, before June 2012 |
18:23 |
Calinou |
marktraceur refused license change to clouds code was rewritten |
18:23 |
Calinou |
est31, ideally we'd be using permissive license from the start |
18:23 |
Calinou |
this is what celeron wanted |
18:23 |
Calinou |
(no one should use GPLv2 on new projects in 2015… either use permissive of GPLv3) |
18:23 |
Calinou |
or* |
18:26 |
est31 |
so why didnt he do it then? :) |
18:26 |
Calinou |
probably didn't think about it |
18:27 |
celeron55 |
because i was a big dummy in 2010 |
18:27 |
celeron55 |
or actually 2011 |
18:28 |
celeron55 |
that's when i accepted the first GPL licensed contributions |
18:28 |
Calinou |
yeah, thought so |
18:29 |
Calinou |
Buildat is Apache 2 for comparison |
18:30 |
|
selat joined #minetest-dev |
18:45 |
|
Wayward_Tab joined #minetest-dev |
19:09 |
|
cib0 joined #minetest-dev |
19:19 |
|
MinetestForFun joined #minetest-dev |
19:27 |
|
TeTpaAka joined #minetest-dev |
19:28 |
|
Miner_48er joined #minetest-dev |
19:30 |
|
crazyR_ joined #minetest-dev |
19:30 |
|
crazyR_ joined #minetest-dev |
19:32 |
|
Hijiri joined #minetest-dev |
19:45 |
|
Amaz joined #minetest-dev |
19:53 |
|
TenPlus1 joined #minetest-dev |
19:54 |
TenPlus1 |
Hi folks... |
19:56 |
TenPlus1 |
Update on std::bad_alloc error... it seems newly generated areas also create map errors which cause this problem... I found out today as a player dug down and at a certain spot the server crashed again |
20:23 |
|
Krock2 joined #minetest-dev |
20:32 |
TeTpaAka |
Why does PlayerSAO::setHP check only for enable_damage when the new hp is smaller than the old hp? Shouldn't enable_damage = false forbid any hp change? |
20:33 |
RealBadAngel |
just realized one important thing that comes with my render to screen plane |
20:33 |
RealBadAngel |
we wont need inventory cubes anymore thx to that |
20:34 |
VanessaE |
oh G*d, please make that a reality |
20:34 |
VanessaE |
:) |
20:34 |
RealBadAngel |
thats damn easy in fact |
20:35 |
RealBadAngel |
you just add a mesh there |
20:35 |
RealBadAngel |
set the mesh rotation, position and you got it |
20:35 |
RealBadAngel |
it can have rotation animator too |
20:45 |
|
crazyR_ joined #minetest-dev |
20:50 |
VanessaE |
RealBadAngel: one thing that's needed is a way to tell the engine to just render the object like it normally does, and specifically to be able to set the location of the virtual camera so that e.g. a small model or nodebox can be scaled up to fit the inventory slot (instead of using an inventory image) |
20:51 |
|
est31 joined #minetest-dev |
20:52 |
RealBadAngel |
you dont need to set the camera for such plane |
20:53 |
RealBadAngel |
you just define meshes position compared to the middle of the screen |
20:53 |
VanessaE |
I mean e.g a mesh node or something else that often ends up wrongly-sized when it's rendered to texture |
20:55 |
RealBadAngel |
you should then just define inventory visual scale perhaps |
20:55 |
VanessaE |
yeah, something like that. |
20:55 |
VanessaE |
but for some objects, setting the camera position may be useful too |
20:56 |
RealBadAngel |
such plane has fixed camera |
20:57 |
RealBadAngel |
only way to move it is to move your monitor lol |
20:57 |
VanessaE |
heh |
20:57 |
RealBadAngel |
its meant i think for hud and shit like that |
20:58 |
RealBadAngel |
but name of it and description doesnt suggest that in any way :) |
20:58 |
RealBadAngel |
https://github.com/zaki/irrlicht/blob/master/tests/material.cpp#L25 |
20:58 |
RealBadAngel |
^^ sample usage |
20:59 |
RealBadAngel |
http://irrlicht.sourceforge.net/docu/classirr_1_1scene_1_1_i_scene_manager.html#a2eb986c6975ebc1aa1f68c878ac8dcda |
21:07 |
RealBadAngel |
VanessaE, http://i.imgur.com/Vd1FIRR.png |
21:08 |
RealBadAngel |
the bigger one is visible on that plane, it rotating when player moves, compare on both orientation on red/green details on the ground |
21:11 |
|
VanessaE joined #minetest-dev |
21:20 |
|
err404 joined #minetest-dev |
21:20 |
|
proller joined #minetest-dev |
21:36 |
|
ElectronLibre left #minetest-dev |
22:24 |
RealBadAngel |
http://i.imgur.com/ioCKtrP.jpg |
23:20 |
|
est31 joined #minetest-dev |
23:20 |
est31 |
RBA: just seen that image: nice! |
23:31 |
est31 |
But its an option to have round vs rectangular maps right? |
23:53 |
RealBadAngel |
i will make it that way propably |