| Time |
Nick |
Message |
| 00:54 |
|
Lone-Star joined #minetest-dev |
| 00:54 |
|
Lone-Star left #minetest-dev |
| 01:21 |
|
erlehmann joined #minetest-dev |
| 01:24 |
|
nephele_ joined #minetest-dev |
| 01:32 |
|
jomat joined #minetest-dev |
| 01:45 |
erlehmann |
i just want to say to all of you that minetest is amazing. rubenwardy when will your book be releazed? |
| 01:45 |
erlehmann |
released? |
| 01:54 |
|
erlehmann joined #minetest-dev |
| 02:17 |
|
nephele joined #minetest-dev |
| 02:18 |
|
erlehmann joined #minetest-dev |
| 02:45 |
|
ssieb joined #minetest-dev |
| 02:46 |
|
vesper11 joined #minetest-dev |
| 04:01 |
|
erlehmann joined #minetest-dev |
| 06:02 |
|
p_gimeno joined #minetest-dev |
| 06:46 |
|
Ruslan1 joined #minetest-dev |
| 07:43 |
|
pyrollo joined #minetest-dev |
| 07:48 |
|
erlehmann joined #minetest-dev |
| 08:39 |
|
ShadowNinja joined #minetest-dev |
| 08:47 |
|
ANAND joined #minetest-dev |
| 10:46 |
rubenwardy |
erlehmann: no lool |
| 10:46 |
rubenwardy |
So, I'm wondering about the future of GUIs in Minetest |
| 10:47 |
rubenwardy |
What we could do is continue with our current development and fully embed the irrlicht GUI code into Minetest. We can then do refractors to modernise it and clean up the styling and such support by replacing skins with something more useful |
| 10:48 |
rubenwardy |
Then use a replacement serialisation format |
| 10:48 |
rubenwardy |
Or, we could could rewrite it or pick up an existing thing starting from scratch |
| 10:49 |
rubenwardy |
I want to increase the reliability of the GUI library by adding unit tests for the rendering, this would be easier if everything went through a class called Render which could be mocked |
| 10:49 |
rubenwardy |
The render class could then contain things like drawInteractableStandardBorder() |
| 10:50 |
rubenwardy |
This is very similar to skins, but should be perelement to allow some styling |
| 10:50 |
rubenwardy |
And also strictly enforced to avoid direct driver calls |
| 10:56 |
rubenwardy |
Most of our issues are actually with out own code and how we use irrlicht GUI |
| 10:57 |
rubenwardy |
Irrlicht GUI will still need to be embedded and maintained by us, but I don't think it's unsalvageable |
| 10:57 |
rubenwardy |
And has the advantage of us being familiar with it, and slightly easier to ensure compatibility |
| 11:03 |
rubenwardy |
The best method would reduce implementation effort, use an existing solution, and give a good result |
| 11:09 |
rubenwardy |
We could use a full time employee for all our problems |
| 11:12 |
|
ensonic joined #minetest-dev |
| 11:13 |
|
ANAND joined #minetest-dev |
| 11:17 |
|
sofar joined #minetest-dev |
| 11:24 |
|
proller joined #minetest-dev |
| 11:26 |
|
sofar joined #minetest-dev |
| 11:47 |
|
sofar joined #minetest-dev |
| 11:54 |
|
proller joined #minetest-dev |
| 11:55 |
|
erlehmann joined #minetest-dev |
| 11:57 |
|
Fixer joined #minetest-dev |
| 12:02 |
|
ANAND joined #minetest-dev |
| 12:06 |
|
sofar joined #minetest-dev |
| 12:17 |
|
sofar joined #minetest-dev |
| 13:06 |
|
Kimapr joined #minetest-dev |
| 13:21 |
|
CrazyDave joined #minetest-dev |
| 13:36 |
|
HDMI_STECKDOSE joined #minetest-dev |
| 14:25 |
|
tomraceror joined #minetest-dev |
| 15:09 |
|
luk3yx joined #minetest-dev |
| 16:22 |
|
Ruslan1 joined #minetest-dev |
| 16:27 |
|
fluxflux joined #minetest-dev |
| 16:29 |
|
BrunoMine joined #minetest-dev |
| 17:30 |
|
BrunoMine joined #minetest-dev |
| 17:46 |
|
ensonic joined #minetest-dev |
| 17:53 |
|
Krock joined #minetest-dev |
| 18:10 |
|
BrunoMine joined #minetest-dev |
| 18:19 |
|
BrunoMine joined #minetest-dev |
| 18:35 |
|
ensonic joined #minetest-dev |
| 18:43 |
|
Taoki joined #minetest-dev |
| 18:45 |
|
BrunoMine joined #minetest-dev |
| 19:11 |
|
tomraceror joined #minetest-dev |
| 19:46 |
Krock |
rubenwardy: how does the default color of text work? |
| 19:46 |
Krock |
as in: styling |
| 19:46 |
rubenwardy |
It doesn't |
| 19:47 |
Krock |
currently it doesn't but I'm trying to fix it |
| 19:47 |
rubenwardy |
It has never worked for label, but it works for labels |
| 19:47 |
rubenwardy |
If that makes.sense lol |
| 19:47 |
Krock |
before: https://i.imgur.com/3WVpfjV.png |
| 19:49 |
Krock |
Text Yellow is an invisible button |
| 19:49 |
Krock |
so where does it take the label color? |
| 19:50 |
Krock |
f (style.isNotDefault(StyleSpec::TEXTCOLOR)) { |
| 19:52 |
rubenwardy |
e->setOverrideColor(style.getColor(StyleSpec::TEXTCOLOR)); |
| 19:52 |
rubenwardy |
https://github.com/minetest/minetest/pull/8383/files#diff-074850a88796e13dad12dd7ae1a8d9f7R709 |
| 19:53 |
rubenwardy |
which you've already found |
| 19:53 |
Krock |
so somewhere in the CGUITTFont there must be a place where it's read |
| 19:53 |
|
ensonic joined #minetest-dev |
| 19:55 |
Krock |
ok. it's not |
| 19:56 |
Krock |
dang .. does it apply the override color somewhere else then?! |
| 20:06 |
Krock |
found it |
| 20:08 |
rubenwardy |
color never worked on label[] |
| 20:08 |
rubenwardy |
but it did on buttons and fields |
| 20:08 |
rubenwardy |
I suspect buttons and fields may have previously been using irrlicht builtin text, rather than our label element which supports colorize |
| 20:09 |
Krock |
it's a matter of what to pass to font->draw() |
| 20:10 |
Krock |
the raw string is needed in order to define a default color |
| 20:10 |
Krock |
since this did not work in StaticText it now also doesn't work in GUIButton |
| 20:11 |
rubenwardy |
probably seems easiest to make it work in statictext |
| 20:11 |
Krock |
haha "easiest" |
| 20:11 |
rubenwardy |
I think the /formspec dialog needs to have an image of what it should look like |
| 20:11 |
rubenwardy |
well, easiest but not reverting |
| 20:12 |
rubenwardy |
what I mean is: the best fix |
| 20:12 |
rubenwardy |
even better: formspec rendering should be unit tested :( |
| 20:29 |
|
BrunoMine joined #minetest-dev |
| 20:59 |
Krock |
thanks, I hate EnrichedText |
| 21:06 |
Krock |
well, yes and no. override colors and EnrichedString just don't fit at all together |
| 21:06 |
Krock |
will try tomorrow |
| 21:09 |
|
BrunoMine joined #minetest-dev |
| 21:27 |
|
Foz joined #minetest-dev |
| 21:47 |
|
paramat joined #minetest-dev |
| 21:59 |
paramat |
(reads logs) about GUI future: "... continue with our current development and fully embed the irrlicht GUI code into Minetest ... has the advantage of us being familiar with it, and slightly easier to ensure compatibility ... reduce implementation effort, use an existing solution ..." i am not an expert on GUI, but this seems better than a complete replacement |
| 22:01 |
|
BrunoMine joined #minetest-dev |
| 22:01 |
paramat |
c55 often links to an article about the danger of rewrites, i somewhat agree with it |
| 22:04 |
paramat |
nice to see someone writing that minetest is "amazing" ^_^ that is a nice thing about someone who has been away for years |
| 22:11 |
|
proller joined #minetest-dev |
| 22:22 |
rubenwardy |
I'm getting very strange behaviour in formspecs |
| 22:22 |
rubenwardy |
I'm not sure what's going on |
| 22:33 |
|
BrunoMine joined #minetest-dev |
| 22:37 |
rubenwardy |
it looks like when a box[] is being parsed is affecting how it is then drawn |
| 22:37 |
rubenwardy |
I wonder if there's broken state |
| 22:42 |
rubenwardy |
fixed it |
| 22:42 |
rubenwardy |
I thought formspec_version[2] was the latest |
| 22:42 |
rubenwardy |
but it's 3 |
| 22:43 |
rubenwardy |
also, apparently real ordering isn't documented in lua_Api |
| 22:43 |
rubenwardy |
also, real ordering also messes up legacy formspecs by non-deterministically shuffling things within a priority |
| 22:48 |
|
Foz joined #minetest-dev |
| 22:48 |
rubenwardy |
any objections to a "Regression" tag? |
| 22:48 |
rubenwardy |
maybe this doesn't count |
| 22:50 |
rubenwardy |
actually, I might just fix it |
| 22:57 |
|
Fixer_ joined #minetest-dev |
| 22:58 |
paramat |
no objection |
| 23:16 |
|
BrunoMine joined #minetest-dev |
| 23:24 |
|
BrunoMine joined #minetest-dev |
| 23:36 |
|
Fixer joined #minetest-dev |
| 23:48 |
rubenwardy |
https://forum.minetest.net/viewtopic.php?f=6&t=23769&view=unread&sid=5c97b87ffba70c6571987c9d47ab5e2e#unread |
| 23:51 |
|
BrunoMine joined #minetest-dev |
