| Time |
Nick |
Message |
| 00:00 |
paramat |
when drowning in issues and PRs already (and CSM issues) it seems wise to avoid non-essential big changes |
| 00:00 |
celeron55 |
i'm looking towards years from now with these |
| 00:01 |
pgimeno_ |
celeron55: yes it does, it runs without love2d if you use luajit and have SDL... something to look into is whether it needs luajit for normal operation, as love2d only comes with luajit |
| 00:02 |
celeron55 |
well, it won't run out of the box anyway |
| 00:02 |
celeron55 |
but SDL support means there's a good point in which to hook up a different backend |
| 00:02 |
celeron55 |
being MT |
| 00:08 |
QwertyCool |
so if the roadmap doesn't actually indicate what needs to be done for release, what does? |
| 00:08 |
QwertyCool |
how do i know where to direct my contribution efforts to be most effective in the team moving forward? |
| 00:13 |
pgimeno_ |
celeron55: sorry, I must have misremembered, apparently Luigi doesn't include multiline textboxes either |
| 00:14 |
pgimeno_ |
a multiline textbox is difficult, it's pretty much an ascii text editor (and something Irrlicht got terribly wrong) |
| 00:16 |
paramat |
QwertyCool we don't have any single document with plans for 0.5, following github threads, especially issues, will give an idea. some are marked with the 0.5 milestone https://github.com/minetest/minetest/issues?q=is%3Aopen+is%3Aissue+milestone%3A0.5.0 |
| 00:17 |
paramat |
also see https://github.com/minetest/minetest/projects |
| 00:18 |
paramat |
there are some PRs labelled as 'adoption needed', some open , some closed |
| 00:23 |
|
ThomasMonroe joined #minetest-dev |
| 00:25 |
QwertyCool |
what does "release needed" tag mean? |
| 00:26 |
QwertyCool |
won't be merged till release? |
| 00:28 |
|
Fixer joined #minetest-dev |
| 00:30 |
paramat |
'rebase'? |
| 00:30 |
paramat |
we don't have a 'release needed' label |
| 00:34 |
QwertyCool |
sorry yes. |
| 00:34 |
QwertyCool |
misread |
| 01:40 |
|
Andrej1 joined #minetest-dev |
| 03:23 |
|
Andrej1 joined #minetest-dev |
| 05:56 |
|
TC01 joined #minetest-dev |
| 10:49 |
|
YuGiOhJCJ joined #minetest-dev |
| 11:26 |
|
YuGiOhJCJ joined #minetest-dev |
| 11:53 |
|
YuGiOhJCJ joined #minetest-dev |
| 12:19 |
|
Fixer joined #minetest-dev |
| 13:27 |
|
Jordach joined #minetest-dev |
| 13:52 |
|
ThomasMonroe joined #minetest-dev |
| 14:13 |
|
AndroBuilder joined #minetest-dev |
| 14:13 |
AndroBuilder |
http://dev.minetest.net/Intro is this still correct? or not since a quick test does not show the png-tile but shows a greenish block :D |
| 14:22 |
|
ThomasMonroe joined #minetest-dev |
| 14:27 |
|
Sockbat joined #minetest-dev |
| 14:39 |
|
ThomasMonroe joined #minetest-dev |
| 15:01 |
|
dmurph joined #minetest-dev |
| 15:10 |
|
Lone-Star joined #minetest-dev |
| 15:33 |
|
antims joined #minetest-dev |
| 15:35 |
|
YuGiOhJCJ joined #minetest-dev |
| 16:02 |
|
RobbieF joined #minetest-dev |
| 16:02 |
|
RobbieF left #minetest-dev |
| 16:23 |
|
YuGiOhJCJ joined #minetest-dev |
| 17:55 |
|
Gael-de-Sailly joined #minetest-dev |
| 18:02 |
|
ThomasMonroe joined #minetest-dev |
| 18:15 |
|
Darcidride_ joined #minetest-dev |
| 18:39 |
|
ThomasMonroe joined #minetest-dev |
| 18:52 |
|
Krock joined #minetest-dev |
| 18:58 |
|
QwertyCool joined #minetest-dev |
| 18:58 |
QwertyCool |
buenos dias |
| 19:00 |
ThomasMonroe |
buenos dias QwertyCool |
| 19:00 |
|
TMcSquared joined #minetest-dev |
| 19:00 |
ThomasMonroe |
comos estas tu? |
| 19:10 |
Krock |
ThomasMonroe, QwertyCool, please continue this discussion in #minetest if you consider to carry on |
| 19:10 |
ThomasMonroe |
ok Krock |
| 19:24 |
|
paramat joined #minetest-dev |
| 19:35 |
|
ssieb joined #minetest-dev |
| 19:39 |
pgimeno_ |
celeron55: I think this library includes most or all widgets that formspecs support, though I'm afraid it's very tightly tied to the engine: https://www.kennyshields.net/git/loveframes/ , also the support for multiline text isn't very complete (it lacks selection and copy/paste) |
| 19:40 |
pgimeno_ |
demo here: https://www.kennyshields.net/git/loveframes-demo/ (needs love2d 0.9.1; needs one change in the source to run, namely replacing 'require "libraries.loveframes"' with 'loveframes = require "libraries.loveframes"') |
| 19:43 |
QwertyCool |
oh shit, did celeron already code a replacement?! |
| 19:45 |
pgimeno_ |
it's worth mentioning that many people in love2d are using imgui through a wrapper lib (imgui is in c++) |
| 19:45 |
rubenwardy |
argh |
| 19:46 |
rubenwardy |
imgur literally says in it's readme that it's not aimed at end users |
| 19:46 |
rubenwardy |
but for programmers and designs |
| 19:46 |
rubenwardy |
"Dear ImGui is designed to enable fast iteration and empower programmers to create content creation tools and visualization/ debug tools (as opposed to UI for the average end-user)" |
| 19:47 |
sfan5 |
does it look better than irrlichts UI at least? |
| 19:47 |
rubenwardy |
yes |
| 19:47 |
rubenwardy |
but that's not hard |
| 19:47 |
sfan5 |
true |
| 19:47 |
rubenwardy |
Irrlicht's GUI is not meant for end users as well |
| 19:48 |
QwertyCool |
lel |
| 19:48 |
rubenwardy |
At least one of their developers said |
| 19:49 |
pgimeno_ |
I grudge Irrlicht's multiline textboxes |
| 19:50 |
QwertyCool |
is librocket an option? |
| 19:51 |
sfan5 |
what's that? |
| 19:52 |
rubenwardy |
it's webkit basicall |
| 19:52 |
rubenwardy |
HTML |
| 19:52 |
rubenwardy |
librocket.com |
| 19:52 |
rubenwardy |
err |
| 19:52 |
rubenwardy |
https://github.com/libRocket/libRocket |
| 19:53 |
QwertyCool |
or FLTK? |
| 19:53 |
rubenwardy |
librocket appears dead |
| 19:55 |
QwertyCool |
https://github.com/ocornut/imgui and https://www.wxwidgets.org already discussed? |
| 19:56 |
rubenwardy |
"Dear ImGui is designed to enable fast iteration and empower programmers to create content creation tools and visualization/ debug tools (as opposed to UI for the average end-user)" |
| 19:56 |
pgimeno_ |
imgui was discussed a few lines ago, wxwidgets doesn't sound like a good idea |
| 19:57 |
QwertyCool |
oopps miss read that as ocornut not imgui my bad |
| 19:57 |
QwertyCool |
gotcha |
| 19:58 |
celeron55 |
librocket is stupid in my opinion |
| 19:58 |
celeron55 |
it's way too bloated and we already have a flexible script interpreter in MT |
| 19:59 |
QwertyCool |
celeron55: did you already mock something up in imgui? |
| 19:59 |
pgimeno_ |
plus what would we be doing if we used it? convert formspec to html then let librocket render the html? sounds like one step too much |
| 19:59 |
rubenwardy |
argh |
| 19:59 |
rubenwardy |
imgui isn't appropriate |
| 20:00 |
celeron55 |
QwertyCool: umm... no, i've done nothing at all |
| 20:00 |
pgimeno_ |
celeron55 was asking about Lua GUIs, that's why I brought up one implemented in love2d which uses Lua exclusively |
| 20:00 |
QwertyCool |
ThomasMonroe: awaiting your input https://github.com/minetest/minetest/issues/7025 |
| 20:01 |
QwertyCool |
oh my appologies. |
| 20:01 |
QwertyCool |
too many library names, im green |
| 20:03 |
celeron55 |
anyway, my entire idea depends on server-provided CSM to work |
| 20:03 |
celeron55 |
so it's definitely not something for the near future |
| 20:03 |
rubenwardy |
server-provided CSM shouldn't become required |
| 20:04 |
celeron55 |
well then you discard the idea completely |
| 20:04 |
rubenwardy |
so it should be possible for servers to send GUIs like currently |
| 20:04 |
celeron55 |
and IMO CSM can be dropped completely also then |
| 20:04 |
celeron55 |
the only reason CSM exists is for it to progress to that eventually |
| 20:05 |
celeron55 |
if it doesn't, i might even fork MT |
| 20:05 |
celeron55 |
(that might sound ridiculous, but i'm serious) |
| 20:05 |
rubenwardy |
a replacement should have very code CSM support |
| 20:05 |
rubenwardy |
to allow custom elements and latency reduction |
| 20:06 |
celeron55 |
code? |
| 20:06 |
rubenwardy |
good |
| 20:06 |
rubenwardy |
but it's a bit much to make modders write a csm mod just to show a simple form |
| 20:06 |
rubenwardy |
guess they could just use formspecs |
| 20:06 |
rubenwardy |
or something very similar |
| 20:06 |
celeron55 |
you can always simplify things |
| 20:06 |
celeron55 |
for common cases |
| 20:08 |
celeron55 |
i'm siding on removing CSM though, and with that, drop this, because i'm lazy and people hate all change |
| 20:09 |
pgimeno_ |
celeron55: why did you want the GUI in Lua anyway? just so the GUI code could be sent by the server? |
| 20:09 |
celeron55 |
and are used to the current security model of MT |
| 20:10 |
celeron55 |
so that the gui can run on the client and be responsive and allow anything a modder wants to do |
| 20:10 |
celeron55 |
up to falling back to drawing 2d graphics and not even using the provided gui |
| 20:10 |
celeron55 |
or providing their own |
| 20:10 |
pgimeno_ |
ah |
| 20:11 |
celeron55 |
anyway, some are not willing to ditch the current security model which will stop all attempts at this |
| 20:11 |
celeron55 |
so improved formspecs it is |
| 20:12 |
pgimeno_ |
the current security model = no server provided csms? |
| 20:12 |
celeron55 |
yes |
| 20:15 |
rubenwardy |
I'm very much a proponent to client-side scripting |
| 20:16 |
rubenwardy |
but we can't be negligent and compromise our users |
| 20:16 |
pgimeno_ |
maybe a compromise could be to let the server verify the presence of CSMs... guess that has already been discussed? |
| 20:16 |
celeron55 |
the current CSM API is too wide for sure and allows accessing stuff that wasn't necessarily designed to be secure |
| 20:18 |
pgimeno_ |
I mean the server could do something like: don't allow the user to log in if this or that CSM isn't installed |
| 20:18 |
celeron55 |
i'd probably just have a second one for server-sent stuff or something |
| 20:18 |
celeron55 |
not sure how to actually handle this mess |
| 20:19 |
rubenwardy |
disable filesystem access completely, and load the scripts from a virtual filesystem |
| 20:19 |
rubenwardy |
checking ofc for bytecode |
| 20:19 |
rubenwardy |
and don't just luajit |
| 20:20 |
rubenwardy |
virtual filesystem is complexity to mess up |
| 20:20 |
rubenwardy |
but hopefully not so much as an actual filesystem |
| 20:20 |
celeron55 |
why are you talking about filesystems |
| 20:21 |
rubenwardy |
second one -> environment? |
| 20:21 |
celeron55 |
it literally just needs a list of filename - data pairs |
| 20:21 |
celeron55 |
if even that |
| 20:24 |
celeron55 |
at this point i'd probably need a different name for the system i want to make and implement it separate from current CSM to work on its side |
| 20:25 |
rubenwardy |
remove csm then |
| 20:26 |
rubenwardy |
there's no need for csm other than server-provided scripting |
| 20:26 |
celeron55 |
there's just the problem that i probably can't find time to make what i want to make, so i'm not sure where i'm going with this |
| 20:27 |
celeron55 |
(in a 5-year timespan i might) |
| 20:28 |
celeron55 |
(assuming i manage to make someone else do things that i do now) |
| 20:29 |
celeron55 |
(this is one of the times when i'd like to just shut down everything and turn them back on once i've had time to think and code) |
| 20:32 |
celeron55 |
(anyone happen to fancy a completely outdated 5GB forum? lol) |
| 20:33 |
rubenwardy |
storage space isn't so bad with a HDD (but slow) |
| 20:33 |
rubenwardy |
phpbb though? |
| 20:33 |
* rubenwardy |
shivers in disgust |
| 20:33 |
QwertyCool |
celeron, if tou give me instruction i can help you work on the code youd like to writ |
| 20:34 |
celeron55 |
rubenwardy: one word: backups |
| 20:35 |
celeron55 |
the thing is annoying on many levels, believe me |
| 20:46 |
|
Foz joined #minetest-dev |
| 20:51 |
|
ThomasMonroe joined #minetest-dev |
| 21:25 |
|
Beton joined #minetest-dev |
| 21:49 |
|
Gael-de-Sailly joined #minetest-dev |
| 21:53 |
|
Darcidride_ joined #minetest-dev |
| 21:56 |
|
Lone-Star joined #minetest-dev |
| 22:02 |
QwertyCool |
does fontsize change callback function on linux? |
| 22:03 |
QwertyCool |
do you need to restart client to see font size change? |
| 22:07 |
red-001 |
is there any well tested cross platform way to sandbox a process? |
| 22:08 |
sofar |
depends, what resources do you want to sandbox away? |
| 22:08 |
sofar |
stack+heap? |
| 22:08 |
sofar |
file system? |
| 22:10 |
|
CBugDCoder joined #minetest-dev |
| 22:15 |
pgimeno_ |
https://github.com/CodethinkLabs/sandboxlib#user-content-relationship-to-other-projects |
| 22:15 |
pgimeno_ |
^ red-001 |
| 22:18 |
|
indiana joined #minetest-dev |
| 22:18 |
sfan5 |
it really depends on what you want to sandbox? |
| 22:19 |
sfan5 |
a process with possibly malicous machine code? a process running a scripting lang interpreter? a processing running a JIT engine? |
| 22:20 |
red-001 |
talking in the context on sandboxing the whole CSM/whatever it will be called subsystem |
| 22:21 |
sfan5 |
how would you effectively sandbox that without sandboxing the whole of minetest? |
| 22:21 |
celeron55 |
i guess it'd communicate by RPC by the rest of MT |
| 22:22 |
celeron55 |
or, more like just feed commands in and out through stdin/out 8) |
| 22:22 |
red-001 |
I imagine that would be fairly slow |
| 22:22 |
sfan5 |
sure but what's the alternative? |
| 22:22 |
celeron55 |
well that's what you do |
| 22:22 |
celeron55 |
no other options |
| 22:22 |
sfan5 |
shared mapped memory would work too, if you're careful |
| 22:22 |
celeron55 |
that's how chrome works too afaik |
| 22:22 |
sfan5 |
but no sandboxing without performance impact |
| 22:24 |
|
Darcidride_ joined #minetest-dev |
| 23:28 |
|
paramat joined #minetest-dev |
| 23:28 |
paramat |
game#2041 is good, might merge later with game#2027 |
| 23:28 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2041 -- Adjust kelp placement and add selection box by Ezhh |
| 23:28 |
ShadowBot |
https://github.com/minetest/minetest_game/issues/2027 -- Add marram grass for coastal sand dunes by paramat |
| 23:29 |
Shara |
THanks paramat |
| 23:29 |
Shara |
Any plantlike_rooted nodes should be done that way really |
| 23:31 |
Shara |
paramat: one last thing maybe - thoughts on adding waving = 1 to them as well? |
| 23:32 |
Shara |
It apparently works fine, and big plants like that in water would not be still |
| 23:32 |
sofar |
do they wave both ways? |
| 23:32 |
Shara |
Both ways? |
| 23:32 |
sofar |
e.g. plantlike and leaflike waving? |
| 23:33 |
Shara |
Now I will need to stare at a leaf to try and work out the difference |
| 23:33 |
Shara |
It's plantlike waving from what I can tell, because the base doesn't move |
| 23:34 |
Shara |
I wasn't sure if it was supported for this drawtype, but it looks good |
| 23:40 |
paramat |
i can try it |
| 23:41 |
Shara |
I dropped a gif in the PR thread if you want to check. Seems worth it to me |
| 23:44 |
paramat |
yes it's good |
| 23:45 |
paramat |
please add |
| 23:45 |
Shara |
Gladly. Quite pleased with how that turned out :) |
| 23:47 |
Shara |
Done |
| 23:50 |
QwertyCool |
https://github.com/minetest/minetest/pull/7007 can i get a final word on this? |
| 23:50 |
QwertyCool |
remove the frill error text and resubmit or leave as is |
| 23:51 |
QwertyCool |
progess feels like quicksand when there's so many chefs in the kitchen t.t |
| 23:52 |
paramat |
looking |
| 23:52 |
paramat |
yes i agree with rubenwardy's comment |
| 23:53 |
paramat |
will +1 when done |
| 23:54 |
QwertyCool |
thanks. will do. |
| 23:54 |
paramat |
heh this isn't slow for MT |
| 23:58 |
paramat |
since trivial only needs 1 approval (mine), so will be merged soon |
| 23:58 |
red-001 |
#7007 |
| 23:58 |
ShadowBot |
https://github.com/minetest/minetest/issues/7007 -- improving error message in src/map.cpp - closes #6977 by QwertyCoolMT |
| 23:58 |
QwertyCool |
paramat: what would be the most helpful thing for me to do in your opinion as someone who's green to c++ and the project? |
