Time |
Nick |
Message |
00:05 |
|
Lesha_Vel joined #minetest |
00:28 |
|
v-rob joined #minetest |
00:32 |
|
Verticen joined #minetest |
00:42 |
|
Sokomine joined #minetest |
00:47 |
MisterE |
hey, do you have any news for me for the minetest blog post? I know about the new png encoder, but I havnt been following engine development too closely in the last few weeks. |
00:47 |
|
v-rob joined #minetest |
00:48 |
MTDiscord |
<GreenXenith> the png encoder isnt exactly new |
00:53 |
MisterE |
oh, heh, I thought I was PMing sfan5 |
00:53 |
MisterE |
rip |
00:56 |
MTDiscord |
<MisterE> oh, that was added in 5.5 huh, missed that. |
00:58 |
|
AliasAlreadyTake joined #minetest |
02:00 |
|
v-rob joined #minetest |
02:07 |
|
v-rob joined #minetest |
03:02 |
|
LibreLikeGNU joined #minetest |
03:19 |
|
Verticen joined #minetest |
03:24 |
erlehmann |
MisterE when you mention the PNG encoding in the engine, also mention the pure lua thing in modlib, just so it's clear that this can be done backwards compatible. |
03:28 |
|
MisterE joined #minetest |
03:29 |
|
queria^clone joined #minetest |
03:33 |
|
queria^clone joined #minetest |
04:00 |
|
MTDiscord joined #minetest |
04:03 |
|
v-rob joined #minetest |
04:32 |
|
Hawk777 joined #minetest |
04:40 |
|
cation joined #minetest |
05:01 |
|
riff-IRC joined #minetest |
05:23 |
|
Pexin joined #minetest |
05:48 |
|
Yad joined #minetest |
06:02 |
|
jvalleroy joined #minetest |
06:02 |
|
jvalleroy joined #minetest |
06:27 |
|
Pexin joined #minetest |
06:32 |
|
lemonzest joined #minetest |
06:41 |
|
calcul0n joined #minetest |
06:42 |
|
v-rob joined #minetest |
07:28 |
|
panwolfram joined #minetest |
07:40 |
sfan5 |
isn't it kinda obvious that often functions added in the engine can be replicated by hand in Lua? |
07:41 |
|
panwolfram joined #minetest |
07:46 |
|
_Zaizen_[m] joined #minetest |
07:49 |
Oblomov |
sfan5: well, it's often, not always (depends how tightly they depend on stuff not exposed by the engine to lua), so it might be appropriate to specify |
08:29 |
|
grouinos joined #minetest |
09:08 |
|
mazes_83 joined #minetest |
09:26 |
|
appguru joined #minetest |
09:43 |
|
specing_ joined #minetest |
09:55 |
|
toulene joined #minetest |
10:15 |
MTDiscord |
<luatic> MisterE: I thought the blog was supposed to be concerned with recent events? The PNG encoders are both significantly older than the first blog post. Perhaps add some kind of "in case you missed it" kind of thing to cover past developments? |
10:40 |
|
appguru joined #minetest |
10:48 |
|
Taoki joined #minetest |
11:03 |
celeron55 |
i certainly don't see why you couldn't/shouldn't showcase old stuff that wasn't showcased before |
12:04 |
|
tech_exorcist joined #minetest |
12:31 |
|
Norkle joined #minetest |
12:34 |
|
Sven_vB joined #minetest |
12:44 |
|
lumidify joined #minetest |
12:50 |
|
tech_exorcist joined #minetest |
12:55 |
|
proller joined #minetest |
12:58 |
|
kamdard joined #minetest |
13:06 |
erlehmann |
sfan5 the thing with PNG encoding in particular is that you probably want backwards compatibility. also in my measurements (encoding up to 128×128 images), doing stuff in lua is not even meaningfully slower. |
13:06 |
erlehmann |
and since at least luatic claims that the encoders are similar … |
13:09 |
|
mazes_83 joined #minetest |
13:10 |
erlehmann |
MisterE for the blogpost, you could highlight the release of mineclone2 0.72. it has mcl_maps (using tga_encoder), barrels (like chests, but movable and a full node), composters (to make organic matter into fertilizer), corrsbow (yes, it is called that in the game) … and a new bug where pigs, sheep, and villagers spawn on trees (and basically everywhere you can find a solid node, lol). |
13:11 |
erlehmann |
mineclone2 has not had a release for a year or so |
13:11 |
erlehmann |
it's probably a good idea to mention it |
13:12 |
erlehmann |
because that way the advice “don't update from contentdb, update using git directly” can finally go the way of the dodo |
13:30 |
|
appguru joined #minetest |
14:02 |
erlehmann |
rubenwardy given mod incompatibilities, do you see an issue with the fact that mcl5 published an mcl_maps that was compatible with mcl_maps used in mcl2 and mcla (in fact, it was basically a copy of it) and then changed it in an incompatible way recently? or is this not a problem because there is no API exposed that another mod could depend on? |
14:03 |
erlehmann |
i mean in all 3 cases it is part of a game and not a standalone mod on contentdb |
14:04 |
erlehmann |
but if at some point in the future some feature would depend on existence of maps (e.g. treasure maps), then it might be? unclear to me. |
14:05 |
erlehmann |
i mean, i'll probably just add such stuff to mcl_maps once i get to it. it's more a “figure out the gameplay balance” than a technical thing. |
14:06 |
erlehmann |
but i think it could help if there was a clear guideline under which circumstances it is permissible to have a mod named the same and under which circumstances it is not |
14:07 |
erlehmann |
and I think changing an API in an incompatible way could be a good start |
14:07 |
erlehmann |
it would still leave open all kinds of enhancements |
14:07 |
erlehmann |
but prevent the “well, minetest things you have this, but actually you don't” scenario |
14:08 |
erlehmann |
thinks |
14:10 |
erlehmann |
this is not about mcl_maps in particular. i think i have done the majority of development on it after fleckenstein stopped and i simply stopped caring about what mcl5 does. i just think it is a neat example for figuring out some guidelines. |
14:10 |
erlehmann |
and it would apply to many mods in games i think. |
14:26 |
|
GNUHacker joined #minetest |
14:27 |
|
Fixer joined #minetest |
14:29 |
ronoaldo |
not that it is big thing but if there is room you can showcase my mod which lead to a small contribution to another mod along the way: minenews (shameless self promotion) |
14:46 |
|
tech_exorcist joined #minetest |
14:49 |
|
appguru joined #minetest |
15:17 |
|
definitelya joined #minetest |
15:23 |
|
fling joined #minetest |
15:35 |
|
Talkless joined #minetest |
15:44 |
|
specing joined #minetest |
16:04 |
|
olliy joined #minetest |
16:24 |
|
MisterE joined #minetest |
16:27 |
|
v-rob joined #minetest |
16:27 |
|
gargamel joined #minetest |
16:28 |
gargamel |
I haven't seen a witch house in weeks any clues? |
16:40 |
MTDiscord |
<Jonathon> erlehmann: get your facts straight |
16:40 |
MTDiscord |
<Jonathon> "mineclone2 has not had a release for a year or so" https://content.minetest.net/packages/Wuzzy/mineclone2/ look at releases |
16:40 |
rubenwardy |
2022-03-13, 2021-03-1 |
16:40 |
rubenwardy |
checks out for me |
16:40 |
rubenwardy |
over a year |
16:41 |
rubenwardy |
0.72.0 was this month |
16:41 |
MTDiscord |
<Jonathon> yeah, so it has had a release within a year |
16:41 |
rubenwardy |
I think they meant "had not had a release" |
16:42 |
MTDiscord |
<Jonathon> never corrected it ? |
16:42 |
rubenwardy |
you never correct your typos |
16:42 |
MTDiscord |
<Jonathon> anyways, if they want to also complain about the minor thing that is maps, they should first address all mineclon*s massive issue with using the mesecon* names |
16:45 |
MTDiscord |
<Jonathon> also, if im remembering this correctly, maps where only added in the latest release on mineclone2, which mineclone5 has been releasing to cdb much more often, and had the change before cdb ever had a mineclone2 release, so technically from the cdb point of things, mineclone5 got it first |
16:45 |
MTDiscord |
<Jonathon> ^citation needed on this |
16:46 |
MTDiscord |
<Jonathon> also for anyone else that cares, see https://content.minetest.net/policy_and_guidance/ section 3, but it really only applies at approval time |
17:06 |
|
Verticen joined #minetest |
17:49 |
|
Pexin joined #minetest |
18:00 |
|
sobkas joined #minetest |
18:29 |
|
RhineDevil joined #minetest |
18:35 |
|
sobkas joined #minetest |
18:46 |
|
___nick___ joined #minetest |
19:04 |
|
___nick___ joined #minetest |
19:06 |
|
___nick___ joined #minetest |
19:08 |
|
erlehmann joined #minetest |
19:12 |
|
Verticen joined #minetest |
19:12 |
erlehmann |
Jonathon yes the mesecon stuff is similar to the maps, you are totally on point there! |
19:14 |
erlehmann |
jonathon i doubt “whoever got the stuff on cdb first has the right to trash it” goes down well with people who created mcl_maps (fleckenstein) or who are enhancing it (me) |
19:15 |
rubenwardy |
the policy about getting things first was designed for mods, rather than games |
19:15 |
rubenwardy |
the mod case is just much simpler |
19:15 |
rubenwardy |
It's unfair to a game if mod names were allowed to be stolen |
19:15 |
erlehmann |
well, if i create a mod and someone else puts it up on contentdb and then at some point later makes it incompatible, can i upload a fixed version? |
19:16 |
erlehmann |
of the mod i mean |
19:16 |
rubenwardy |
given that mcl_ is in mineclone2's namespace, I'd say that mineclone2 and mineclone5 should both have the right to it |
19:16 |
erlehmann |
well, mcla also uses it |
19:17 |
erlehmann |
because everything is supposed to be at least roughly compatible with mcl2 |
19:17 |
erlehmann |
(in most cases, it is just a straight up copy) |
19:17 |
erlehmann |
the thing is, jonathon is right, this applies to mesecons/redstone too. |
19:17 |
erlehmann |
so the problem has been lurking there for a while, it is only that no one so far blatantly broke compatibility |
19:18 |
erlehmann |
or maybe they did |
19:18 |
erlehmann |
but i did not notice |
19:18 |
erlehmann |
i bet i find some mesecons problems if i look for them |
19:19 |
erlehmann |
rubenwardy if you hae any guidance how to avoid stuff like this in the future, let me know. but i can only make *my* code a drop-in replacement for other people's code. i can not stop them from butchering stuff. |
19:20 |
rubenwardy |
only way to avoid it is to stop forking projects |
19:21 |
erlehmann |
well, mcl5 was forked because very few people agreed with kays “move fast and break bedrock” idea and mcla was created because at least initially people liked having absolutely no regressions. many regret that now! |
19:21 |
erlehmann |
a friend, who does project management says that extensive testing works 100% but is entirely unfun, so he quit |
19:22 |
erlehmann |
i doubt there is some space between those 3 to make more forks, but i also think that mcl5 will continue the trajectory of not even being compatible with earlier versions of itself. |
19:23 |
erlehmann |
well, i guess i should prepare an mcla release for cdb |
19:23 |
erlehmann |
april's fools release or so ;) |
19:23 |
erlehmann |
rubenwardy is it possible to depend on gamename/modname ? |
19:23 |
erlehmann |
rubenwardy that would solve it cleanly |
19:23 |
erlehmann |
at least for this use case |
19:24 |
rubenwardy |
that would make it so that if someone depends on mineclone2/modname, it won't work with mineclonia |
19:24 |
erlehmann |
indeed |
19:24 |
rubenwardy |
#12034 |
19:24 |
ShadowBot |
https://github.com/minetest/minetest/issues/12034 -- Improve mod dependency system |
19:24 |
rubenwardy |
the related issue |
19:25 |
erlehmann |
i have a simple example why i came up with that. i made crying obsidian and put it in mcl_core in mcla. kay was like “haha why don't you make it a separate mod”. i agreed, i can only depend on it if i make it one. |
19:26 |
erlehmann |
turns out he was only mocking me and put it in mcl_core in mcl5. now the problem is that i have plans to actually make a mod that needs crying obsidian, but i can't depend on mcl_core obviously. |
19:26 |
erlehmann |
i'd like to depend on mcl5/mcl_core or so |
19:26 |
erlehmann |
i mean i don't *need* to be compatible with mcl5, but it would be nice if i just could |
19:27 |
|
MisterE joined #minetest |
20:07 |
|
debiankaios joined #minetest |
20:07 |
|
Sven_vB joined #minetest |
20:56 |
|
proller joined #minetest |
21:15 |
|
Fixer_ joined #minetest |
21:28 |
|
v-rob joined #minetest |
21:42 |
|
specing_ joined #minetest |
22:16 |
|
Alias joined #minetest |
22:32 |
|
z812 joined #minetest |
22:34 |
|
RhineDevil joined #minetest |
22:50 |
|
erlehmann joined #minetest |
22:54 |
|
erle joined #minetest |
23:00 |
|
Sven_vB joined #minetest |
23:04 |
|
Lesha_Vel joined #minetest |
23:14 |
|
LibreLikeGNU joined #minetest |
23:32 |
|
panwolfram joined #minetest |
23:35 |
|
Desour joined #minetest |
23:42 |
|
proller joined #minetest |
23:44 |
Desour |
lately I've often seen in the logs that people refer to issues via their link, ie. https://github.com/minetest/minetest/issues/1234 . It would be nice if ShadowBot would look up the title for those |
23:46 |
wsor |
will be redundant for people that have link embeds in there irc client |
23:47 |
erle |
they can ignore the bot |
23:47 |
wsor |
well yes, but actually no |
23:47 |
wsor |
because then it would also be ignoring up, # numbers, etc |
23:48 |
erle |
oh |
23:48 |
wsor |
anyways, just putting it out there, doesnt matter to me either way |
23:48 |
erle |
a separate botis needed |
23:48 |
|
urli_65709 joined #minetest |
23:48 |
urli_65709 |
This program is free software; you can redistribute it and/or modify it under the GNU Affero General Public License as published by the Free Software Foundation, either version 3 of the License, or (at your option) any later version. To download and verify the source code, execute: “git clone http://daten.dieweltistgarnichtso.net/src/urlinfobot-irc.git/;cd urlinfobot-irc;git verify-commit HEAD” |
23:48 |
erle |
https://github.com/minetest/minetest/issues/1234 |
23:48 |
urli_65709 |
https://github.com/minetest/minetest/issues/1234 Title: Client side caching of mapblocks · Issue #1234 · minetest/minetest · GitHub |
23:49 |
erle |
Desour, this is a bot i have written for exactly this purpose. do you like it? |
23:49 |
erle |
https://google.cn https://google.jp |
23:49 |
urli_65709 |
https://google.cn Title: Google |
23:49 |
urli_65709 |
https://google.jp Title: Google |
23:49 |
erle |
damn lol |
23:49 |
erle |
https://zh.wikipedia.org/wiki/Wiki |
23:49 |
urli_65709 |
https://zh.wikipedia.org/wiki/Wiki Title: wiki - 维基百科,自由的百科全书 |
23:49 |
erle |
yeah so it can do more than latin 1 |
23:50 |
erle |
Desour can you make it run somewhere? |
23:50 |
erle |
i have no server rn |
23:51 |
rubenwardy |
erle: bots aren't allowed in this channel without permission |
23:51 |
erle |
rubenwardy i didn't know that, please forgive me. |
23:52 |
Desour |
(don't have server either, kind bot though :) ) |
23:52 |
wsor |
#1 |
23:52 |
ShadowBot |
https://github.com/minetest/minetest/issues/1 -- GlowStone code by anonymousAwesome |
23:52 |
wsor |
oddly similar style |
23:52 |
erle |
the reason is that putting the URL at the front is the only way this can work |
23:53 |
|
Samson1 joined #minetest |
23:54 |
erle |
i have gotten tons of feedback from people who can't parse their way out of a wet paper bag that they dislike it a lot, but everyone has their own cute idea how to do it differently that doesn't really associate a URL with a title. |
23:54 |
Samson1 |
Hello again! I need a hand with something. 3d_armour, 'attempt to call field 'player_register_model' (a nil value.).' |
23:54 |
Samson1 |
The line it failed to call is this: local def = stack:get_definition() |
23:55 |
Samson1 |
Any ideas? |
23:55 |
wsor |
let guess, still running the same stack of half outdated mods? |
23:55 |
erle |
usually every “not putting the URL in front” approach has issues with multiple URLs in one chat message and/or hosts that take long to respond |
23:55 |
Samson1 |
wsor No, I updated everything. |
23:56 |
wsor |
*let me guess |
23:58 |
wsor |
noonenone_of_your_business grep -r player_register_model 3d_armor/ |
23:59 |
wsor |
3d_armor/3d_armor/init.lua:default.player_register_model("3d_armor_character.b3d", { |
23:59 |
wsor |
well, first things first, the mod should be updated out of the darkages |
23:59 |
Samson1 |
It's the latest version of the mod. |
23:59 |
wsor |
https://github.com/minetest/minetest_game/blob/42baede13fdf855773cc44ce10f4a3ea4e239404/mods/default/legacy.lua#L33 |
23:59 |
wsor |
ik |
23:59 |
|
proller joined #minetest |