| Time |
Nick |
Message |
| 00:04 |
Jordach |
hmmmm, there are way way way way way too many caves |
| 00:04 |
Jordach |
1/45 sounds good over 16 blocks |
| 00:05 |
Jordach |
(as 16^2 sections) |
| 01:31 |
ShadowNinja |
Is there a way to do foo:bar(y) --> baz.bar(foo, y) in Lua with metatables? Probably with __index or __call. |
| 03:02 |
ShadowNinja |
Comments? http://ix.io/6xq |
| 03:05 |
ShadowNinja |
kahrl: Is there a way to do hypot(x, y, z) instead of the two calls? It was simple the other way... |
| 03:20 |
|
khonkhortisan joined #minetest-dev |
| 03:42 |
ShadowNinja |
There, got it. |
| 04:01 |
ShadowNinja |
Well this way is about 100x slower. But they come up with different results... |
| 04:05 |
ShadowNinja |
Aaand dual call way wins. |
| 04:14 |
ShadowNinja |
1 billion times in 0.33 seconds. Not bad. |
| 04:23 |
hmmmm |
what do you guys think about taoki's fog |
| 04:23 |
hmmmm |
i think i'm gonna merge it soon |
| 04:41 |
ShadowNinja |
http://ix.io/6xt |
| 04:41 |
ShadowNinja |
Haven't tried it but it looks good. |
| 04:42 |
ShadowNinja |
hmmmm: How does that look? |
| 05:02 |
|
ffoxin joined #minetest-dev |
| 05:03 |
ShadowNinja |
*ahem* Anyone? |
| 05:12 |
hmmmm |
what's with the vector.length?? |
| 05:23 |
ShadowNinja |
hmmmm: Useful for direction vectors. (See normalize) |
| 05:23 |
hmmmm |
no, i mean why is it coded that convoluted way instead of just sqrt(x*x + y*y + z*z) |
| 05:25 |
ShadowNinja |
hmmmm: http://irc.minetest.ru/minetest-dev/2013-07-06#i_3183265 |
| 05:26 |
hmmmm |
oh, there's no way it's going to overflow |
| 05:26 |
|
RealBadAngel joined #minetest-dev |
| 05:26 |
RealBadAngel |
hi |
| 05:29 |
ShadowNinja |
Well the speed is identical. |
| 05:31 |
ShadowNinja |
If anything hypot is a tad faster. |
| 05:31 |
hmmmm |
that's nonsense, it does more arithmetic |
| 05:31 |
RealBadAngel |
Nore has pulled computer code into technic_game and im pretty suprised with it |
| 05:31 |
hmmmm |
there's no reason why it isn't slower |
| 05:32 |
RealBadAngel |
the code uses precompiled libraries .so and .dll |
| 05:32 |
ShadowNinja |
13.29. 13.3. Run ten billion times. |
| 05:33 |
RealBadAngel |
i had no idea thats possible |
| 05:33 |
ShadowNinja |
RealBadAngel: Yep, I use LuaSocket for the IRC mod. |
| 05:34 |
hmmmm |
it could be that luajit notices the pattern and optimizes it into a single FMA instruction followed by fsqrt |
| 05:34 |
RealBadAngel |
holly molly, that makes many weird ideas possible |
| 05:35 |
hmmmm |
shadowninja, those are seconds btw, right? |
| 05:35 |
ShadowNinja |
hmmmm: Yes. |
| 05:35 |
RealBadAngel |
that means i can do Z80 and CP/M without changes to the engine at all |
| 05:36 |
RealBadAngel |
only the screen will be needed |
| 05:39 |
ShadowNinja |
hmmmm: Anything else? |
| 05:40 |
hmmmm |
nah |
| 05:42 |
ShadowNinja |
Ok, preparing pull... |
| 05:47 |
ShadowNinja |
hmmmm: https://github.com/minetest/minetest/pull/810 |
| 05:51 |
kahrl |
good morning |
| 05:51 |
RealBadAngel |
morning kahrl |
| 05:53 |
kahrl |
nice to see the vector helpers committed |
| 05:54 |
kahrl |
a few ideas: should vector.normalize handle null vectors? should there be a vector.dot? |
| 05:54 |
kahrl |
for the first one: http://paste.dy.fi/iNp |
| 05:54 |
kahrl |
erm http://paste.dy.fi/icL |
| 05:54 |
hmmmm |
nobody really uses the dot product |
| 05:54 |
hmmmm |
if it's needed we could add it |
| 05:55 |
kahrl |
yeah |
| 05:55 |
kahrl |
same with cross product |
| 05:55 |
hmmmm |
and yes, normalize should handle 0 vectors |
| 05:58 |
|
nore joined #minetest-dev |
| 06:02 |
nore |
hmmmm, should that be merged? https://github.com/minetest/minetest/pull/655 |
| 06:02 |
hmmmm |
maybe |
| 06:03 |
nore |
and will the sky color pull request be merged too? |
| 06:03 |
hmmmm |
man wtf is with this |
| 06:03 |
hmmmm |
+ float x; m_time_of_day >= 0.5 ? x = (1 - m_time_of_day) * 2 : x = m_time_of_day * 2; |
| 06:04 |
hmmmm |
can he try a *little* harder in terms of code quality |
| 06:04 |
nore |
it is the same as abs(x-0.5)*2, no? |
| 06:06 |
hmmmm |
spaces mixed with tabs for indentation |
| 06:06 |
kahrl |
nore: 1 minus that, if I see that right |
| 06:06 |
hmmmm |
and then he declares a function inside of the sky ctor |
| 06:06 |
hmmmm |
it's called m_horizon_blend() |
| 06:06 |
hmmmm |
m_horizon_blend is a great method name |
| 06:07 |
nore |
yes, I am seeing that |
| 06:07 |
nore |
what are those m_ at the beginning of function/parameters names |
| 06:07 |
nore |
? |
| 06:07 |
RealBadAngel |
i say yes for #655 its really needed |
| 06:07 |
ShadowNinja |
https://github.com/minetest/minetest/pull/811 |
| 06:07 |
RealBadAngel |
49 limit is pain in the ass |
| 06:08 |
nore |
And https://github.com/minetest/minetest/pull/606, since now force_load is possible, the counter-arguments do not hold anymore |
| 06:08 |
nore |
with voxel manip |
| 06:08 |
nore |
and it avoids reading the map, etc |
| 06:09 |
RealBadAngel |
and comparing to other request that changed the limit, this one allows to configure it |
| 06:09 |
hmmmm |
it's probably a good time to bring this up, i dislike how the horizon in the sky abruptly cuts off |
| 06:10 |
nore |
when I made the request, it was the same, then I thought it was good to be able to configure it |
| 06:11 |
kahrl |
hmmmm: that's what I meant when I said it looks like standing in a giant cardboard box |
| 06:12 |
kahrl |
it not noticeable with the current fog but very with #799 |
| 06:12 |
hmmmm |
yeah |
| 06:13 |
hmmmm |
if you're going to draw special attention to the sky background like that, it better at least look nice |
| 06:13 |
hmmmm |
code style aside, because it's so damn noticable, i don't think i want to merge this |
| 06:14 |
RealBadAngel |
i also noticed that effect |
| 06:14 |
RealBadAngel |
i like the colours but the box is annoying |
| 06:14 |
hmmmm |
nore, no the sky color pull request won't be merged. |
| 06:15 |
|
VanessaE joined #minetest-dev |
| 06:15 |
RealBadAngel |
hi VanessaE |
| 06:15 |
VanessaE |
hey |
| 06:16 |
RealBadAngel |
imho the fog shall be drawn in circle, that would be ok |
| 06:17 |
VanessaE |
*checks log* |
| 06:17 |
kahrl |
RealBadAngel: that would look like a cardboard cylinder :P |
| 06:24 |
VanessaE |
I suggested taoki just do some kind of simple vertical gradient |
| 06:24 |
ShadowNinja |
hmmmm: My pull? |
| 06:24 |
VanessaE |
he wanted to save that for some future tweak |
| 06:24 |
hmmmm |
?? |
| 06:24 |
ShadowNinja |
https://github.com/minetest/minetest/pull/811 |
| 06:25 |
hmmmm |
in a bit |
| 06:25 |
ShadowNinja |
VanessaE: s/he/She/ |
| 06:26 |
RealBadAngel |
kahrl, when you turn around your eyes dont further or closer, your view is a cylinder |
| 06:26 |
RealBadAngel |
*dont see |
| 06:26 |
VanessaE |
if I may point something out: |
| 06:27 |
VanessaE |
the fog coloring is really nice and all, but the fog color also needs to be tuned toward gray as the view distance shrinks. |
| 06:27 |
VanessaE |
THAT would probably help the "I'm in a cardboard box" appearance. |
| 06:28 |
kahrl |
VanessaE: it's really the sky color that does that |
| 06:28 |
hmmmm |
what do you guys think about https://github.com/minetest/minetest/pull/773 |
| 06:29 |
VanessaE |
kahrl: oh? hrm |
| 06:30 |
hmmmm |
kahrl, the media thread assert thing, how did you guess that it was caused by having too many files sent? |
| 06:30 |
VanessaE |
well either way it needs tuned to that end. I noticed it myself when I tested Taoki's code. discolored fog doesn't really "feel" right when it's up close. |
| 06:30 |
kahrl |
hmmmm: the ObjectProperties field should be called collide_with_objects |
| 06:30 |
VanessaE |
then it should be bluish-grey |
| 06:30 |
VanessaE |
hmmmm: absolutely YES (if the code works right anyway) |
| 06:31 |
kahrl |
hmmmm: well I don't really know why the assert happens, I just looked at under which circumstances it can fail |
| 06:31 |
kahrl |
why the assert failures happen* |
| 06:31 |
hmmmm |
wouldn't the next logical step be to see how the number of sent files could exceed the number requested? |
| 06:31 |
kahrl |
I did but didn't find anything |
| 06:31 |
hmmmm |
hrmm |
| 06:32 |
hmmmm |
well then i guess you should push that |
| 06:32 |
hmmmm |
ignoring the error is better than exploding |
| 06:32 |
kahrl |
I'll do that |
| 06:56 |
|
loggingbot_ joined #minetest-dev |
| 06:56 |
|
Topic for #minetest-dev is now Minetest core development and maintenance. Chit-chat goes to #minetest. Consider this instead of /msg celeron55. http://irc.minetest.ru/minetest-dev/ http://dev.minetest.net/ |
| 07:45 |
VanessaE |
bbl, bedtime |
| 07:47 |
|
darkrose joined #minetest-dev |
| 07:47 |
|
darkrose joined #minetest-dev |
| 08:05 |
|
kaeza joined #minetest-dev |
| 10:39 |
sfan5 |
any objections angainst merging #799 ? |
| 10:41 |
Taoki |
Probably even the tribes in Africa would find something more to add to that before it's rready :P But I feel it's ready and as best as it can get, and I heard several who approve too |
| 10:41 |
Taoki |
Besides,,, anything else that can be tweaked can be tweaked later on if someone really wants to. I did all I could do |
| 10:51 |
Taoki |
sfan5: My suggestion is to merge it. If a few people find so many horrible issues with it, they can try to fix them later. IMHO the effect and improvement is very good and more important than issues which can be tweaked later (given I don't have more to add and tweak to that code myself) |
| 10:52 |
|
Jordach joined #minetest-dev |
| 10:52 |
sfan5 |
Taoki: I'd do that, but a core dev needs to agree with it |
| 10:52 |
Taoki |
RealBadAngel is one that does |
| 10:52 |
Taoki |
RealBadAngel: Can you confirm that, to be sure I didn't understand wrong |
| 10:52 |
|
PilzAdam joined #minetest-dev |
| 10:54 |
Taoki |
sfan5: celeron55 also tested it a few days ago and said "it's fine for me" |
| 10:55 |
Taoki |
1 or 2 days |
| 10:56 |
RealBadAngel |
Taoki, im after your colours, they make world look way better. But i have to agree with others that the box feelin around the player is terrible. I told you saw from begining. I will try to see if it can be fixed. |
| 10:56 |
Taoki |
ok. Think it shouldn't be added in the meantime till that's solved? |
| 10:57 |
Taoki |
It's fine then, will wait for you to take a look at that |
| 10:59 |
RealBadAngel |
effect is excellent, just to spread it over sphere, or something round |
| 10:59 |
|
ecube joined #minetest-dev |
| 10:59 |
RealBadAngel |
corners in every edge of the world.... bleah ;) |
| 11:00 |
Taoki |
I'll post a note on how to modify that, since it's only one number that influences the sphere |
| 11:01 |
Jordach |
we *ARE* in a voxel world though |
| 11:04 |
|
rubenwardy joined #minetest-dev |
| 11:10 |
RealBadAngel |
Jordach, sure, but the feelin of fog box is terrible |
| 11:10 |
RealBadAngel |
its just way too many corners |
| 11:25 |
|
Calinou joined #minetest-dev |
| 11:53 |
|
Zeg9 joined #minetest-dev |
| 11:56 |
|
rubenwardy_ joined #minetest-dev |
| 12:53 |
|
kaeza joined #minetest-dev |
| 13:16 |
|
john_minetest joined #minetest-dev |
| 13:49 |
|
diemartin joined #minetest-dev |
| 14:31 |
|
NakedFury joined #minetest-dev |
| 14:55 |
|
sapier joined #minetest-dev |
| 14:56 |
sapier |
VanessaE I fixed up/down keys in listboxes but I had to change enter button handling too. Plz tell me if I accidentaly broke something related to enter |
| 15:07 |
kahrl |
sapier, is this a bug? builtin/mainmenu.lua:580 |
| 15:07 |
kahrl |
shouldn't that be menu.last_world <= #menu.worldlist |
| 15:14 |
kahrl |
oh, yeah indeed I can't delete the last world in the list |
| 15:16 |
|
Jordach joined #minetest-dev |
| 15:42 |
sapier |
yes you're right kahrl |
| 15:56 |
|
john_cephalopoda joined #minetest-dev |
| 16:09 |
ShadowNinja |
https://github.com/minetest/minetest/pull/812 |
| 16:12 |
|
Jordach joined #minetest-dev |
| 16:12 |
|
Jordach_ joined #minetest-dev |
| 16:35 |
|
john_cephalopoda joined #minetest-dev |
| 16:35 |
|
john_minetest left #minetest-dev |
| 16:40 |
|
serengeor joined #minetest-dev |
| 16:57 |
|
sapier1 joined #minetest-dev |
| 17:05 |
|
kaeza joined #minetest-dev |
| 17:07 |
|
ShadowNinja joined #minetest-dev |
| 17:14 |
|
hmmmm joined #minetest-dev |
| 17:15 |
ShadowNinja |
hmmmm: https://github.com/minetest/minetest/pull/812 |
| 17:18 |
hmmmm |
fine, but why would you check if both x and y are 0, and not just x after assigning it? |
| 17:21 |
PilzAdam |
<hmmmm> what do you guys think about https://github.com/minetest/minetest/pull/773 <- I think it should be merged |
| 17:22 |
hmmmm |
sfan5, could you *not* say "hurr any objections to #pull" 5 minutes after we have a detailed discussion about why it shouldn't be merged as-is? |
| 17:22 |
sfan5 |
hmmmm: ...I didn't see the discussion then |
| 17:22 |
sfan5 |
the discussion was before I joined |
| 17:22 |
hmmmm |
are you going to fix the hexadecimal number commit? |
| 17:22 |
sfan5 |
fix == ? |
| 17:23 |
hmmmm |
yes, you forgot to change several color codes, and then there's the problem where the bottom bar is sometimes green |
| 17:23 |
hmmmm |
remember, we discussed this? |
| 17:23 |
sfan5 |
yes |
| 17:23 |
sfan5 |
the bottombar-is-green thing seems to have to do something with irrlich |
| 17:23 |
sfan5 |
irrlicht* |
| 17:24 |
PilzAdam |
its sometimes blue for me |
| 17:24 |
hmmmm |
so you're saying it's an irrlicht bug and you can't do anything to influence the color being black |
| 17:24 |
hmmmm |
can you go into more detail why? |
| 17:24 |
hmmmm |
it seems to have been consistently black before your commit happened |
| 17:24 |
sfan5 |
it was a guess |
| 17:25 |
sfan5 |
I don't know why std::hex shouldn't work correctly |
| 17:25 |
|
iqualfragile joined #minetest-dev |
| 17:26 |
PilzAdam |
reverting 307c57dcc494161d019d7ee394e7cdb1912ca11f fixes it |
| 17:26 |
hmmmm |
obviously. but i thought we'd give him a chance to fix things up |
| 17:28 |
ShadowNinja |
hmmmm: Changed. |
| 17:28 |
hmmmm |
so is builtin/modstore.lua:192 the offending line? |
| 17:28 |
hmmmm |
0x32CD32 seems like it'd be pretty green to me |
| 17:29 |
PilzAdam |
but it was LIM before |
| 17:29 |
sfan5 |
hmmmm: it was LIM for Lime before |
| 17:29 |
hmmmm |
right |
| 17:29 |
hmmmm |
but is that element really that green box with the games? |
| 17:29 |
ShadowNinja |
There is no newline after my name in the credits... |
| 17:29 |
PilzAdam |
sfan5, maybe add some unittests for parseColor()? |
| 17:29 |
hmmmm |
erm, you shouldn't need unit tests for something so simple.... |
| 17:29 |
hmmmm |
this seems like it might be a deeper bug |
| 17:30 |
hmmmm |
shadowninja, try again |
| 17:30 |
ShadowNinja |
hmmmm: I did. |
| 17:30 |
|
RealBadAngel joined #minetest-dev |
| 17:30 |
hmmmm |
so you're saying it's consistently like that? |
| 17:30 |
hmmmm |
as in, you can reproduce *that* bug? |
| 17:31 |
ShadowNinja |
Yes. |
| 17:31 |
hmmmm |
that's good to know, tell sapier about it |
| 17:32 |
thexyz |
sfan5: https://github.com/minetest/minetest/commit/307c57dcc494161d019d7ee394e7cdb1912ca11f#L3R2525 you should use susbtr(0, 2), and so on |
| 17:32 |
thexyz |
because, well, color is 2 characters |
| 17:32 |
sfan5 |
derp |
| 17:33 |
hmmmm |
anyway, I just tried changing that color code to red and it didn't seem to change anything, so that's not what the problem is |
| 17:33 |
ShadowNinja |
sapier1: http://imgur.com/0IXy2BN |
| 17:34 |
PilzAdam |
^ thats what I see too |
| 17:35 |
hmmmm |
oh shoot |
| 17:35 |
hmmmm |
i just noticed they're missing commas |
| 17:35 |
hmmmm |
are those the only things that get concatenated? |
| 17:36 |
PilzAdam |
yes |
| 17:37 |
PilzAdam |
adding commas fixes it |
| 17:38 |
sfan5 |
hmmmm: I can fix the contributor list in my commit |
| 17:38 |
hmmmm |
also don't forget to fix the *rest* of the color codes that you missed |
| 17:38 |
sfan5 |
I did |
| 17:39 |
hmmmm |
great |
| 17:39 |
sfan5 |
in the contributor list and something with mods |
| 17:39 |
sfan5 |
https://gist.github.com/sfan5/5944262 |
| 17:40 |
sfan5 |
^ the diff |
| 17:40 |
hmmmm |
great |
| 17:41 |
hmmmm |
i just wish you'd look at your own code more the next time instead of blaming it on irrlicht and then blaming it on std::hex :/ |
| 17:41 |
sfan5 |
I didn't blame it on std::hex... |
| 17:41 |
hmmmm |
[01:25 PM] <sfan5> I don't know why std::hex shouldn't work correctly |
| 17:42 |
sfan5 |
The bar at the bottom is still sometimes blue or grey for me |
| 17:42 |
|
diemartin joined #minetest-dev |
| 17:43 |
hmmmm |
hmm |
| 17:43 |
hmmmm |
you should try printing out the substrings to make sure they're actually what you want to be parsed |
| 17:44 |
hmmmm |
also, your code can crash minetest if someone has a color code less than 6 characters |
| 17:44 |
hmmmm |
it'd throw out_of_range |
| 17:45 |
sfan5 |
that part is sapiers(?) code |
| 17:45 |
ShadowNinja |
IMO the credits should use a table and table.concat(minetest.contributors, ","). |
| 17:45 |
hmmmm |
well it should be done differently |
| 17:45 |
hmmmm |
u32 color; |
| 17:46 |
hmmmm |
whoops, color is already used for the string name |
| 17:46 |
hmmmm |
u32 color_value; |
| 17:46 |
hmmmm |
iss >> std::hex >> color_value; |
| 17:46 |
sfan5 |
and then bitshift it? |
| 17:46 |
hmmmm |
r = color_value & 0xFF; g = (color_value >> 8) & 0xFF; b = (color_value >> 16) & 0xFF; |
| 17:46 |
hmmmm |
other way around actually |
| 17:48 |
kahrl |
actually you should be able to just call video::SColor(color_value) |
| 17:49 |
hmmmm |
and then set the alpha after |
| 17:49 |
kahrl |
right |
| 17:50 |
hmmmm |
hmmm, as for the listbox concatenation problem, i suppose the reason why sapier didn't see it was because it's easy to miss |
| 17:51 |
hmmmm |
and i don't know why i didn't see it after the first time, my eyes must be that bad |
| 17:51 |
ShadowNinja |
hmmmm: Do you agree that a table would be better? |
| 17:52 |
hmmmm |
not really.... this is such a mundane piece of code |
| 17:53 |
sfan5 |
updated diff: https://gist.github.com/sfan5/5944262 |
| 17:53 |
hmmmm |
when people are adding contributors to the list they'd just make sure they write the comma in. i can see how you'd miss some on accident when doing it in bulk though |
| 17:53 |
hmmmm |
ah yes, that looks good sfan. |
| 17:53 |
hmmmm |
does that fix the "sometimes blue" bottom bar for you? |
| 17:54 |
sfan5 |
for some reason, yes |
| 17:54 |
hmmmm |
there must've been a less obvious string parsing issue after fixing the substr calls then... not that it matters now |
| 17:55 |
hmmmm |
also with this you can remove the if(color.size() != 6) return false; |
| 17:56 |
hmmmm |
and please change the code style of that second check to if (!string_allowed(...)) \n \t return false; |
| 17:56 |
sfan5 |
wah, too late |
| 17:56 |
hmmmm |
i seriously don't get where this trend of putting the if body on the same line. |
| 17:57 |
hmmmm |
that's okay, you can force push for about 5-10 minutes |
| 17:57 |
sfan5 |
and its gone again |
| 17:57 |
hmmmm |
the if body on the same line came from* i meant to say before |
| 17:58 |
sfan5 |
I don't get why I should remove the .size() check |
| 17:58 |
hmmmm |
because it's not needed anymore |
| 17:59 |
hmmmm |
you read the entire color in as an integer instead of breaking it up into 3 two-character strings now |
| 17:59 |
sfan5 |
pushed |
| 18:00 |
hmmmm |
ah, great, perfect |
| 18:02 |
sfan5 |
but if someone would inpu "F0F" it wouldn't end up how the most browsers parse #F0F |
| 18:02 |
sfan5 |
*input |
| 18:02 |
sfan5 |
but since Minetest is a game and not a browser.. |
| 18:03 |
hmmmm |
i guess it's because i see hexadecimal colors as not a code but rather just a number |
| 18:05 |
|
proller joined #minetest-dev |
| 18:13 |
|
kaeza joined #minetest-dev |
| 18:35 |
serengeor |
redeclipse |
| 18:35 |
serengeor |
urgh, wrong window |
| 18:36 |
serengeor |
downside of having window manager that autofocuses window under mouse cursor.. |
| 18:47 |
|
rubenwardy joined #minetest-dev |
| 18:58 |
|
neko259 joined #minetest-dev |
| 19:20 |
hmmmm |
the whole hex color code commit really screwed sapier's fixes and there are like 5000000 conflicts |
| 19:22 |
|
sapier joined #minetest-dev |
| 19:36 |
VanessaE |
sapier: your current patches don't apply cleanly against current HEAD. |
| 19:36 |
NakedFury |
what is the hex color code for? colored chat? |
| 19:36 |
sapier |
grrr what did they change again? |
| 19:37 |
VanessaE |
current git HEAD is busted for me anyway - it throws these errors and http://pastebin.com/ckeCHeZL and tries to log me into some server, not sure where, as soon as I launch it. |
| 19:38 |
VanessaE |
s/errors and/errors / |
| 19:38 |
sfan5 |
NakedFury: someone could actually add that |
| 19:38 |
sfan5 |
they are for colored formspecs |
| 19:38 |
hmmmm |
lol colored chat. |
| 19:39 |
sapier |
hmm sounds like someone created a invalid serverlist/favourites entry |
| 19:39 |
sfan5 |
previously it was limited to ~10 colors, with the new format we can use all 16.3M colors |
| 19:39 |
NakedFury |
not just chat but also player names |
| 19:39 |
hmmmm |
every single time there's a game with a colored chat feature, they eventually remove it in a patch because it creates more problems than it's worth |
| 19:40 |
sapier |
guys I'm seriously annoyed ... you add features but don't add bugfixes |
| 19:40 |
NakedFury |
yes when the admins go stupid teen mode and change the colors |
| 19:40 |
sfan5 |
what problems could colored chat create? |
| 19:40 |
hmmmm |
all of my recent commits have been bugfixes |
| 19:40 |
sfan5 |
sapier: does your patch cleanly apply? |
| 19:40 |
NakedFury |
well for one using bright colors |
| 19:40 |
sapier |
#22F922 is crap at least add a lookup table |
| 19:40 |
sapier |
no it doesn't as YOUR patch skrewed it up |
| 19:41 |
hmmmm |
maybe do something like COLOR_LIME_GREEN = 0x22F922 |
| 19:41 |
hmmmm |
(there's no way to do constants in lua so it'd just have to be a global variable) |
| 19:41 |
NakedFury |
but addig colors can help. example admins and special moderators having a different colored name so players can know the, |
| 19:41 |
sfan5 |
^ |
| 19:41 |
sfan5 |
wasn't kahrl working on something chat related while ago? |
| 19:42 |
PilzAdam |
yes |
| 19:42 |
VanessaE |
nice. MT is totally busted now because your stupid lua menu doesn't properly sanitize its input >:-( |
| 19:42 |
sapier |
I'm not against colors I just hate that you guys keep making me additional work |
| 19:42 |
VanessaE |
sapier: ^^^ |
| 19:42 |
hmmmm |
if it makes you feel any better, i had no part in this, it was committed totally without my knowledge |
| 19:43 |
sapier |
VanessaE there are lots of bugs already fixed but sfan5 seems to think that 16m colors instead of 10 is more important than bugfixing |
| 19:43 |
sfan5 |
thats not what I intended |
| 19:43 |
sapier |
but that's what happened |
| 19:43 |
VanessaE |
oh ok, so your latest patch already had that covered. my bad. *waits* |
| 19:43 |
sapier |
I'm gonna rebase it again |
| 19:43 |
PilzAdam |
I guess sfan5 has to rebase sapiers pull then and merge it then |
| 19:44 |
hmmmm |
sapier, it's beyond that |
| 19:44 |
sapier |
I'll rebase it and fix that color numbers too |
| 19:44 |
hmmmm |
mainmenu.lua can be merged and the conflicts manually fixed, but other files, it's easier to just redo your changes |
| 19:45 |
sapier |
redo? |
| 19:45 |
sfan5 |
git mergetool with meld is pretty simple |
| 19:45 |
hmmmm |
yeah, like take the fix that you actually applied and re-apply it to the new code |
| 19:45 |
sapier |
ähhm the fixes already are 500 lines again |
| 19:45 |
hmmmm |
i'm just saying |
| 19:46 |
hmmmm |
it's much harder to work with using a mergetool - i tried it this morning and i gave up |
| 19:46 |
PilzAdam |
nobody like the idea that sfan5 has to rebase it? |
| 19:46 |
hmmmm |
pilzadam, he'd have to rewrite history |
| 19:46 |
ShadowNinja |
kahrl and PilzAdam wrote a highlight system. (Like in IRC) |
| 19:46 |
hmmmm |
these changes were already committed |
| 19:46 |
VanessaE |
sapier: recommend when you rebase, fix your pull request summary work: "-others" <--- what others? |
| 19:46 |
hmmmm |
the most recent commit by sfan fixes what was busted in his original commit that was two days ago |
| 19:47 |
VanessaE |
a/work/wording/ |
| 19:47 |
VanessaE |
*sigh* should I just go back to bed? :-S |
| 19:47 |
PilzAdam |
hmmmm, I mean that he has to rebase sapiers branch to current master |
| 19:47 |
sapier |
lot's of things I already have forgotton due to noone merging it |
| 19:47 |
hmmmm |
oh, erm, it'd probably be better if sapier were to do that |
| 19:48 |
sapier |
yes as I know what needs to be done it's best |
| 19:48 |
sfan5 |
hmmmm: 1 day ago* |
| 19:48 |
VanessaE |
brb |
| 19:48 |
hmmmm |
the funny part about all this is that i predicted it to the T |
| 19:48 |
sapier |
that's why I didn't want to add colors YET |
| 19:49 |
hmmmm |
i said, "great, now that there's the lua main menu, people are going to want to add things but it needs to get stable first, and there are plenty of bugs to fix" |
| 19:49 |
|
VanessaE joined #minetest-dev |
| 19:49 |
VanessaE |
there. |
| 19:50 |
thexyz |
ShadowNinja: link? |
| 19:50 |
hmmmm |
why, oh why, did i not revert sfan's commit and apply sapier's first |
| 19:50 |
PilzAdam |
<hmmmm> obviously. but i thought we'd give him a chance to fix things up |
| 19:50 |
* sfan5 |
will try rebasing sapiers changes onto HEAD |
| 19:50 |
sfan5 |
+now |
| 19:50 |
sapier |
hmmmm actually I guess that'd be mmodt easy way to do |
| 19:51 |
sapier |
sfan rebasing with concurrent changes fails most of time |
| 19:51 |
hmmmm |
erm, you want me to do that right now? |
| 19:51 |
hmmmm |
alright here's my plan |
| 19:51 |
hmmmm |
i revert the two color related commits, apply your fixes, then just redo what the coloring commits changed |
| 19:51 |
sapier |
sounds reasonable |
| 19:52 |
PilzAdam |
wait |
| 19:52 |
PilzAdam |
do that in a seperate branch, and apply the diff against master by hand |
| 19:52 |
hmmmm |
of course |
| 19:52 |
hmmmm |
you think i'd want to make a mess out of the commit log? |
| 19:53 |
PilzAdam |
just wanted to be sure |
| 19:53 |
PilzAdam |
people have strange ideas sometimes |
| 19:54 |
VanessaE |
hmmmm: weblate already did that. :P |
| 19:54 |
sapier |
btw I changed key press handling I guess noone did use "EditBoxEnter" event by now |
| 19:55 |
* sfan5 |
is almost done rebasing the first file |
| 19:55 |
PilzAdam |
sapier, two things Id like to see fixed too: the scrollbar should automatically be set so that the selected world is visble on startup; and mouse wheel control would be nice too |
| 19:55 |
hmmmm |
I know, but it wasn't possible to fix weblate and I guess the pros outweight the cons |
| 19:55 |
hmmmm |
sfan5, nevermind that |
| 19:56 |
hmmmm |
i am doing it on my own |
| 19:56 |
sapier |
I'd like to fix that PilzAdam but irrlicht doesn't support setting scroll position of a listbox |
| 19:56 |
PilzAdam |
ummm... we had that before Lua menu IIRC |
| 19:56 |
sapier |
mouswheel could be possible |
| 19:57 |
sapier |
it should be visible pilzadam |
| 19:57 |
sapier |
but at lowermost position only |
| 19:57 |
PilzAdam |
also, we should sort the world list |
| 19:57 |
|
kaeza joined #minetest-dev |
| 19:57 |
sapier |
sort by what criteria? |
| 19:57 |
PilzAdam |
alphabetically |
| 19:58 |
sapier |
I suggest sorting in c++ as it'd reduce world selection complexity |
| 19:59 |
hmmmm |
how is it currently sorted? by age? |
| 20:00 |
sapier |
don't know as it's read from fs |
| 20:01 |
sfan5 |
almost done merging 3rd file |
| 20:01 |
hmmmm |
sfan5, no, i'm doing it already! |
| 20:01 |
hmmmm |
stop |
| 20:03 |
thexyz |
that's all because of merge-fobia |
| 20:08 |
hmmmm |
and while I'm doing this, would you like me to use color constants instead of the actual hex codes? |
| 20:08 |
hmmmm |
well actually i should be doing that in a separate commit |
| 20:09 |
sfan5 |
sapier: does this look correct? https://gist.github.com/sfan5/5944706 |
| 20:10 |
PilzAdam |
I guess the assembly shaders can be removed |
| 20:11 |
PilzAdam |
since the reason why we changed it to a bool is that assembly shaders wont be done |
| 20:12 |
Exio |
what assembly shaders? |
| 20:12 |
sfan5 |
Exio: those who werent done |
| 20:12 |
Exio |
exactly :P |
| 20:13 |
Anchakor |
do the current MT shaders do anything else then bump-mapping? |
| 20:13 |
PilzAdam |
oh, and the comment "// 0 = off, 1 = assembly shaders only, 2 = highlevel or assembly" needs to be changed too |
| 20:13 |
PilzAdam |
^ thats relevant for your rebase, hmmmm |
| 20:13 |
hmmmm |
yeah |
| 20:13 |
PilzAdam |
Anchakor, sure |
| 20:13 |
sfan5 |
Anchakor: coloring light |
| 20:14 |
PilzAdam |
Anchakor, fog, sunrise/set, brighen topside of a node |
| 20:14 |
sapier |
looks ok sfan5, at least I don't see any obvious error |
| 20:15 |
kahrl |
will enable_shaders = 2 now mean that shaders are disabled? |
| 20:15 |
sapier |
yes |
| 20:15 |
kahrl |
well this will break existing configs |
| 20:16 |
kahrl |
but it can be done I guess |
| 20:16 |
Exio |
i guess anything that isn't "0" or "false" or similar should be true |
| 20:16 |
hmmmm |
this is ridiculous |
| 20:16 |
Exio |
s/guess/think/ |
| 20:16 |
hmmmm |
something as simple as changing enable_shaders shouldn't break configurations |
| 20:16 |
sapier |
what does getbool result for "2"? |
| 20:16 |
kahrl |
sapier: false |
| 20:17 |
sapier |
so only 1 is true? |
| 20:17 |
kahrl |
yeah |
| 20:17 |
PilzAdam |
kahrl, it uses is_yes(), wich only accepts "y", "yes", "true" and "1" |
| 20:17 |
sapier |
:-) strange behaviour normaly everything except 0 is true |
| 20:18 |
sapier |
I realize this isn't quite good but as shaders are a very obvious feature I guess anyone will notice they're disabled and check in settings |
| 20:18 |
Anchakor |
if you want a bloom shader I did some successful experimentation a year ago: https://github.com/Anchakor/minetest/commits/bloom_shader |
| 20:18 |
kahrl |
would changing is_yes break anything? |
| 20:19 |
PilzAdam |
kahrl, only used in settings.h and test.cpp |
| 20:19 |
PilzAdam |
so I guess no |
| 20:21 |
PilzAdam |
hmm... its used in Settings:getStruct() and I dont know how that works |
| 20:21 |
ShadowNinja |
thexyz: Somewhere in the logs. kahrl: Do you have that highlight patch gist? |
| 20:22 |
sfan5 |
hmmmm: I got a complete rebase now |
| 20:22 |
hmmmm |
great, thanks, but i already did it ... |
| 20:22 |
PilzAdam |
ShadowNinja, https://github.com/kahrl/minetest/tree/filteredchat |
| 20:22 |
hmmmm |
grr |
| 20:23 |
hmmmm |
alright, sfan5, merge yours |
| 20:23 |
hmmmm |
s/merge/commit or whatever/ |
| 20:23 |
PilzAdam |
sfan5, have you removed the .asm shaders and fixed the comment? |
| 20:23 |
sfan5 |
i need to test it first |
| 20:23 |
ShadowNinja |
PilzAdam: Wasn't there a colored one? |
| 20:23 |
sfan5 |
PilzAdam: no |
| 20:23 |
sfan5 |
I just merged it |
| 20:23 |
sfan5 |
*rebased |
| 20:24 |
PilzAdam |
ShadowNinja, that one has colores |
| 20:24 |
PilzAdam |
sfan5, then do that before pushing |
| 20:24 |
kahrl |
but only in the console |
| 20:24 |
ShadowNinja |
Ok. |
| 20:24 |
kahrl |
adding colors to the ingame chat is less trivial |
| 20:24 |
PilzAdam |
well, we need to merge this console-only pull request, so that doesnt really matter |
| 20:25 |
PilzAdam |
ehm, nvm |
| 20:25 |
sfan5 |
wat. git changed history but it didn't require force-push |
| 20:26 |
hmmmm |
uhh.. |
| 20:26 |
hmmmm |
why not push it to your own branch first and we'll see |
| 20:26 |
PilzAdam |
" Merge branch 'formspec_mainmenu_fixes' of https://github.com/sapier/m… |
| 20:26 |
PilzAdam |
…inetest" |
| 20:26 |
PilzAdam |
not good... |
| 20:26 |
hmmmm |
squash it. |
| 20:26 |
sfan5 |
PilzAdam: wasn't avoidable |
| 20:26 |
VanessaE |
hmmmm: ..before it multiplies? |
| 20:26 |
VanessaE |
;) |
| 20:27 |
hmmmm |
git cherry-pick |
| 20:27 |
PilzAdam |
its not hard to not use "git merge" |
| 20:27 |
hmmmm |
and you pushed it already |
| 20:27 |
sfan5 |
PilzAdam: git am doesn't allow you to merge it manually |
| 20:27 |
hmmmm |
christ |
| 20:27 |
hmmmm |
someone fix it before it gets pulled |
| 20:27 |
PilzAdam |
oh man, you messed everything up |
| 20:28 |
hmmmm |
why |
| 20:28 |
hmmmm |
the hell |
| 20:28 |
hmmmm |
are people so trigger happy |
| 20:28 |
sfan5 |
:( |
| 20:28 |
kaeza |
I already pulled :( |
| 20:28 |
sfan5 |
whats wrong now? |
| 20:28 |
kaeza |
(if you know what I mean) |
| 20:28 |
PilzAdam |
sfan5, you just merged |
| 20:28 |
sfan5 |
PilzAdam: yeah |
| 20:28 |
PilzAdam |
we. dont. merge. |
| 20:28 |
PilzAdam |
we rebase |
| 20:28 |
sfan5 |
... |
| 20:29 |
hmmmm |
amazing how something so cute and simple like colors turned out to be a nightmare |
| 20:29 |
hmmmm |
pilzadam, are you going to squash the merge? |
| 20:29 |
hmmmm |
and force push it |
| 20:29 |
hmmmm |
or should i |
| 20:29 |
PilzAdam |
lemme try |
| 20:30 |
sfan5 |
its gone now |
| 20:30 |
sfan5 |
now what |
| 20:30 |
hmmmm |
alright |
| 20:30 |
hmmmm |
at this point it's back to the way it was |
| 20:30 |
hmmmm |
maybe i should do it |
| 20:30 |
PilzAdam |
sfan5, push your thing to your fork |
| 20:30 |
PilzAdam |
hmmmm, I can do it |
| 20:31 |
hmmmm |
alright |
| 20:33 |
PilzAdam |
sfan5? |
| 20:35 |
sfan5 |
I got 3 patch files from the commits I just removed, none of them apply cleanly |
| 20:37 |
hmmmm |
how about i just push my own version |
| 20:38 |
sfan5 |
wait |
| 20:38 |
sfan5 |
I got the diff I put into the gist to apply, its one commit now |
| 20:42 |
sfan5 |
I have a commit by "sapier <Sapier at GMX dot net>" with everything merged now |
| 20:43 |
sfan5 |
can I push that? |
| 20:43 |
sapier |
hmm I guess I need to check my git settings again ... thought I removed my email address some time ago :) |
| 20:43 |
PilzAdam |
push to your fork first and look if you havent messed it up |
| 20:44 |
kahrl |
just to make everything more complicated, I propose: https://gist.github.com/kahrl/5944866 |
| 20:47 |
PilzAdam |
kahrl, seems good |
| 20:55 |
sfan5 |
PilzAdam: git isn't pushing it to the branch for some reason, hmmmm can do it |
| 20:58 |
sfan5 |
the commit looks like this: https://gist.github.com/sfan5/5944706 |
| 21:08 |
sfan5 |
good night now |
| 21:19 |
|
PilzAdam joined #minetest-dev |
| 21:26 |
|
mrtux joined #minetest-dev |
| 21:27 |
|
Miner_48er joined #minetest-dev |
| 21:29 |
PilzAdam |
hmmmm, you can push your version |
| 21:50 |
|
salamanderrake joined #minetest-dev |
| 22:15 |
PilzAdam |
sapier1, Lua version of world sorting: https://github.com/PilzAdam/minetest/commit/3b9a511f94264c1205d7c031b5ea271dbc186d64 |
| 22:17 |
sapier1 |
did you check this isn't messing up world selection? |
| 22:18 |
PilzAdam |
oh... |
| 22:18 |
sapier1 |
starting of worlds depends on indexes beeing same anytime ... not sure if it's a problem but at least it needs to be checked |
| 22:19 |
PilzAdam |
it breaks it |
| 22:19 |
sapier1 |
I guess if sorting is done in core you don't need to verify correct indexing everywhere |
| 22:20 |
PilzAdam |
yep |
| 22:21 |
sapier1 |
you could also put a _raw function beween original core fct and lua fct ... my personal preference would be sorting done in core |
| 22:21 |
PilzAdam |
yea, sorting in core would be best I think |
| 22:37 |
ShadowNinja |
hmmmm: https://github.com/minetest/minetest/pull/812 |
| 22:38 |
ShadowNinja |
PilzAdam, kahrl: ^ |
| 23:20 |
hmmmm |
that still wasn't fixed!? ugh |
| 23:20 |
hmmmm |
i need the last commit, i don't get why that was removed |
| 23:21 |
hmmmm |
my patch reverts the last one, the original color commit, applies sapier's patches, then i manually applied the color patches |
| 23:21 |
hmmmm |
can someone please move history forward one? |
| 23:22 |
hmmmm |
this was so stupid, playing with git history especially to this extent, and it's all caused because some people can't learn how to push things to their own goddamn fork before screwing upstream |
| 23:26 |
VanessaE |
https://github.com/minetest/minetest/commit/773b7fb6cf1782535808dafffe6158b75d06f598 |
| 23:26 |
VanessaE |
this? |
| 23:31 |
hmmmm |
erm, no, the one that fixed the original color patch |
| 23:31 |
hmmmm |
i mean the next commit in history after what's currently in upstream |
| 23:31 |
VanessaE |
dammit, I know I saw it in here somewhere. |
| 23:31 |
hmmmm |
my local version's history is different so i don't have it |
| 23:33 |
VanessaE |
only other thing I see that came after the late-media fix was an off-by-1 fix |
| 23:34 |
VanessaE |
https://github.com/minetest/minetest/commit/7e73b7cae9faf203e369dc5cf2b85cff6e6efbad |
| 23:34 |
VanessaE |
(the color fix) |
| 23:35 |
hmmmm |
looking for 7e73b7cae9faf203e369dc5cf2b85cff6e6efbad. |
| 23:35 |
hmmmm |
oh yup, that's it |
| 23:35 |
VanessaE |
lot of tsuris over a dozen-line patch huh? |
| 23:35 |
hmmmm |
yep |
| 23:36 |
hmmmm |
oh wait a minute nevermind, i do have it in my history |
| 23:36 |
hmmmm |
i didn't squash yet |
| 23:36 |
VanessaE |
see? I told you to squash it before it multiplies :P |
| 23:36 |
hmmmm |
I wonder how celeron would've handled this crazy situation |
| 23:39 |
PilzAdam |
hmmmm, done |
| 23:39 |
hmmmm |
thanks |
| 23:39 |
hmmmm |
although it's not going to matter because i'd need to force push anyway because that commit id isn't the same |
| 23:41 |
hmmmm |
grr, how do i change the author... |
| 23:43 |
hmmmm |
alright, this is what i'm going to push: https://github.com/kwolekr/minetest/commit/13d68dbd3dc83a2dcd427258426e70d156b88520 |
| 23:43 |
hmmmm |
agh shit, the lua_api.txt |
| 23:43 |
VanessaE |
hmmmm: I'd recommend a detailed commit msg if possible. |
| 23:43 |
hmmmm |
not needed |
| 23:44 |
hmmmm |
this mess is going to be, for all intents and purposes, invisible to git history |
| 23:44 |
hmmmm |
this never even happened. |
| 23:44 |
VanessaE |
ok :) |
| 23:45 |
hmmmm |
right, so after this commit we need somebody to patch is_yes() in Settings |
| 23:45 |
hmmmm |
(remember, this would break configs as-is) |
| 23:46 |
PilzAdam |
kahrl already linked a patch for that |
| 23:46 |
hmmmm |
great |
| 23:47 |
PilzAdam |
I can commit that together with the shader comment fix and removal of .asm shaders |
| 23:47 |
VanessaE |
https://gist.github.com/kahrl/5944866 |
| 23:48 |
VanessaE |
the aforementioned patch, for reference. |
| 23:48 |
hmmmm |
okay |
| 23:49 |
VanessaE |
I look at some of those variants...wouldn't it be a good idea to pass the string through a "to_lower" sort of function before comparisons? |
| 23:49 |
PilzAdam |
are there lower-case numbers? |
| 23:50 |
VanessaE |
well of course not, but, UASSERT(is_yes("FAlse") == false); |
| 23:50 |
VanessaE |
that, for example |
| 23:50 |
hmmmm |
alright, this is what i am going to push https://github.com/kwolekr/minetest/commit/88d43af276a4ed85500863b427bf370d8eca2095 |
| 23:50 |
VanessaE |
(obviously not one of karhl's changes) |
| 23:50 |
PilzAdam |
VanessaE, std::string s2 = lowercase(trim(s)); is already there.... |
| 23:50 |
hmmmm |
everything looks about right |
| 23:51 |
VanessaE |
PilzAdam: evidently it's not being used if "YeS and FAlse" had to be explicitly checked for :P |
| 23:51 |
hmmmm |
and done |
| 23:52 |
hmmmm |
what the hell is with is_nonzero_number() |
| 23:52 |
VanessaE |
O.o |
| 23:52 |
hmmmm |
is there something wrong with "return atoi(s.c_str())" ? |
| 23:52 |
hmmmm |
well, atoi(s.c_str() != 0 if you want to be more pedantic |
| 23:53 |
hmmmm |
well nevermind, if it works then it works, it just seems convoluted to me |
| 23:59 |
PilzAdam |
https://github.com/PilzAdam/minetest/commit/6bc3a43d52c04b6a91787bcb49d0044f1452fc4e |
| 23:59 |
hmmmm |
yup |
