Minetest logo

IRC log for #minetest-dev, 2020-03-26

| Channels | #minetest-dev index | Today | | Google Search | Plaintext

All times shown according to UTC.

Time Nick Message
00:21 turok joined #minetest-dev
00:22 turok hi all. looking to make my first contributions to minetest... followed the contribution messaging, environment is set up, best to just pick a "beginner friendly" issue and go for it?
02:03 fluxflux joined #minetest-dev
02:25 Foz joined #minetest-dev
02:35 nephele_ joined #minetest-dev
03:05 erlehmann joined #minetest-dev
03:40 Kimapr joined #minetest-dev
03:54 Taoki joined #minetest-dev
06:13 Kimapr joined #minetest-dev
06:22 YuGiOhJCJ joined #minetest-dev
06:36 erlehmann_ joined #minetest-dev
07:00 Beton joined #minetest-dev
08:06 ShadowNinja joined #minetest-dev
08:06 ircSparky joined #minetest-dev
08:07 behalebabo joined #minetest-dev
08:07 indiana joined #minetest-dev
08:40 ShadowNinja joined #minetest-dev
08:45 erlehmann joined #minetest-dev
09:05 karamel joined #minetest-dev
10:04 calcul0n joined #minetest-dev
10:58 lisac joined #minetest-dev
10:58 Kimapr joined #minetest-dev
11:04 Kimapr joined #minetest-dev
11:44 Fixer joined #minetest-dev
12:49 calcul0n_ joined #minetest-dev
12:51 kilbith joined #minetest-dev
12:55 Kimapr joined #minetest-dev
12:59 kilbith_ joined #minetest-dev
13:20 kilbith joined #minetest-dev
14:21 indiana joined #minetest-dev
14:29 Krock joined #minetest-dev
14:34 el joined #minetest-dev
14:35 kilbith joined #minetest-dev
14:48 JDCodeIt joined #minetest-dev
14:49 kilbith joined #minetest-dev
14:53 kilbith_ joined #minetest-dev
15:00 kilbith joined #minetest-dev
15:02 kilbith joined #minetest-dev
15:05 kilbith_ joined #minetest-dev
15:22 kilbith__ joined #minetest-dev
15:30 Taoki joined #minetest-dev
15:43 if_gaga0 joined #minetest-dev
15:45 erlehmann joined #minetest-dev
15:48 erlehmann joined #minetest-dev
16:20 kilbith_ joined #minetest-dev
16:25 kilbith__ joined #minetest-dev
17:21 mizux joined #minetest-dev
17:49 erlehmann joined #minetest-dev
17:49 Krock !tell turok Yes, although there are only 6 of them. Best would be to pick something you dislike in Minetest, and check for similar requests. Try&error is also a way to get familiar with the code
17:49 ShadowBot Krock: O.K.
17:55 Krock will merge #9550 in 15 minutes
17:55 ShadowBot https://github.com/minetest/minetest/issues/9550 -- Connection: Fix deadlock in debug mode by SmallJoker
17:55 Krock #9548 needs a second review
17:55 ShadowBot https://github.com/minetest/minetest/issues/9548 -- Fix alignment tags adding unwanted newlines by pyrollo
18:12 Krock merging..
18:46 sfan5 merging #9548 in 10 m
18:46 ShadowBot https://github.com/minetest/minetest/issues/9548 -- Fix alignment tags adding unwanted newlines by pyrollo
18:48 Wuzzy joined #minetest-dev
18:59 Krock rubenwardy: https://github.com/minetest/minetest/issues/8553#issuecomment-604621479
18:59 Krock this is not about what's drawn behind
18:59 Krock but rather poor rendering performance of textures with alpha
19:00 Krock which makes sense - because each transparent overlay needs to be calculated using the underlying pixels
19:01 Krock which is probably also the reason why Freetype rendering is so slow
19:06 sfan5 if you disable antialiasing, shouldn't 1-bit transparency suffice for font rendering?
19:10 sfan5 >     * **Note**: This will be removed or modified in a future version.
19:10 sfan5 extremely useful
19:17 sfan5 https://github.com/minetest/minetest/commit/e297c7391#diff-1305560bd8befb32862f0feeefabd02eR547
19:17 sfan5 I guess it could be marked as stable?
19:21 Krock yes
19:27 rubenwardy OpenGL supports blending and alpha though?
19:32 Krock yes it does, although hardware can implement it poorly
19:32 sfan5 Krock: https://github.com/minetest/minetest/pull/9555
19:32 Krock another question: is it even possible for the server to get the player's password when SRP is used?
19:33 Krock sfan5: not sure why you opened a PR for that. feel free to merge that patch
19:34 sfan5 because I did more than just delete the line, maybe someone would have found something wrong with the additional documentation
19:34 sfan5 but sure
19:49 Wuzzy what's the point of ephemeral sounds?
19:50 Wuzzy it seems completely pointless to me. If i dont need the handle, i just ignore the return value
19:53 sfan5 less network traffic and memory usage, purely engine internal reasons
19:54 Wuzzy ah its a performance thing
19:55 Wuzzy Please can you add this (that it's a performance thing) to the docs? Otherwise most modders will consider this feature to be useless, and not use it (although they should)
19:57 sfan5 the documentation does recommend usecases for ephemeral sounds, though without explaining the background
19:59 Wuzzy true
19:59 rubenwardy might be worth deprecating playing a sound with `ephemeral` as nil, so that mods opt into using non-ephemeral sounds
20:00 Wuzzy this would generate MASSIVe log spam tho
20:00 rubenwardy deprecation notices only occur once iirc
20:00 rubenwardy although, that may only apply to strict.lua
20:00 Wuzzy not for set_sky...
20:00 rubenwardy and you can fix the deprecation by adding the argument explicitly
20:00 rubenwardy which will work in previous versions
20:01 Wuzzy set_sky currently generates log warnings although paramat suggested to consider the new set_sky API as experimental.....
20:01 Wuzzy i mean, if you use the 5.1.0 syntax of set_sky you get punished with log warnings
20:01 rubenwardy is the API for the new set_sky sound?
20:01 sfan5 deprecation notices are logged each time
20:01 Wuzzy IMO yes
20:02 sfan5 the bigger problem with them is that I get the impression that no modders use them
20:02 sfan5 since they are disabled by default
20:02 rubenwardy aw, it's just the strict.lua global variable checking that doesn't log multiple times
20:02 rubenwardy I swear deprecations do log by default?
20:02 rubenwardy log to warnings
20:02 Wuzzy there is a PR that will scream all warnings to chat (if you enable it). this is VERY important for dev. still not merged :((((
20:03 Wuzzy rubenwardy: yes they are logged but they are not screamed into chat. that's annoying when you're a gamedev like me ^^
20:04 Krock only renamed functions (getpos -> set_pos) are logged once per line
20:04 sfan5 #    Handling for deprecated Lua API calls:
20:04 sfan5 #    -    legacy: (try to) mimic old behaviour (default for release).
20:04 Krock but that deprecation wrapper could be re-used for other functions as well
20:04 Krock problem is the tail calls in LuaJIT
20:04 sfan5 the others being "log" or "error"
20:05 Wuzzy sfan5: yes I have set it to "log" but I want it to be screamed at. I always forget to check the logfile afterwards. It's better if I *instantly* see all warnings while testing 🙂
20:05 sfan5 why afterwards, do you not have a console window open to run minetest?
20:07 Krock also print() is shorter than minetest.chat_send_all
20:07 Krock regarding the debugging potential in console/terminal
20:08 Wuzzy print doesn't seem reliable/official tho
20:08 Wuzzy is print() logged?
20:08 Wuzzy i.e. goes into logfile?
20:08 Krock no, but for testing you don't need it logged anyway
20:09 Wuzzy good point
20:09 Krock also - the print output is thread-safe, since it wraps the minetest.log() function
20:09 Wuzzy i always do minetest.log("error", msg) because it's shorter than minetest.chat_send_all and its in RED 😆
20:09 Wuzzy (for temporary testing, ofcourse)
20:10 rubenwardy ha, real modders use   error(msg)
20:10 rubenwardy that's even more read
20:10 rubenwardy -a
20:11 Wuzzy oh right
20:12 rubenwardy although, it's not actually red
20:12 rubenwardy but it's more in your face
20:12 rubenwardy (I'm joking)
20:12 Wuzzy red is a lie. colors are a lie. everyhing is an illusion
20:12 rubenwardy (obviously real modders log by writing to a default:book in a chest at  0,0,0)
20:13 Krock rubenwardy: oh yes, enslaved node item
20:13 Wuzzy (implying that real modders use MTG? no, that doesn't make sense)
20:14 Wuzzy paramat: I feel the cloud density has been massively nerfed in MTG btw
20:14 Wuzzy in most regions clouds are much less dense. its rare to see a 5.1.0-style density now
20:14 Wuzzy and i think i never saw a density actually higher than in 5.1.0
20:16 sfan5 well it's now dependent on the biome(?)
20:20 JDCodeIt joined #minetest-dev
20:21 Wuzzy humidity*
20:22 Wuzzy it's semi-dependent on the humidity actually
20:22 Wuzzy there's a separate noise as well
20:24 Wuzzy Please somebody look at #9453
20:24 ShadowBot https://github.com/minetest/minetest/issues/9453 -- Document builtin entities by Wuzzy2
20:25 Wuzzy ah, and also this small bugfix PR: #9302
20:25 ShadowBot https://github.com/minetest/minetest/issues/9302 -- Fix z-fighting of large stacked plantlike nodes by Wuzzy2
20:40 p_gimeno joined #minetest-dev
21:17 kilbith joined #minetest-dev
21:18 kilbith sfan5: we have had some disruptions at kidscode owing to the get_connected_players porting to C++
21:19 sfan5 did you pull in the later fixes for it too?
21:19 kilbith it seems to behave differently under the hood than than the Lua version
21:19 kilbith yes
21:20 sfan5 these two? https://github.com/minetest/minetest/commit/7a7bfdca7c2dcbfe40467f64544112551abd7316 https://github.com/minetest/minetest/commit/1761d7d0f8a778e25ca70216520472a61610cbe5
21:20 sfan5 how does the "disruption" manifest?
21:21 kilbith oh nevermind, sorry
21:21 kilbith we don't have pulled these
21:22 kilbith we have 1761d7d0f8a7 (Fix core.get_connected_players() returning unusable ObjectRefs)
21:23 kilbith not 7a7bfdca7c2dcb
21:23 erlehmann_ joined #minetest-dev
21:23 sfan5 7a7bfdca is for a niche situation
21:30 kilbith_ joined #minetest-dev
21:33 kilbith_ joined #minetest-dev
21:45 calcul0n joined #minetest-dev
21:47 Wuzzy oh no! i just realized you are forced to set pixel size for img in hypertext 🙁
21:47 Wuzzy isn't it possible to enter relative "inventory slot" sizes, just like the normal image[] formspec element?
21:48 Wuzzy this is a very odd design decision btw. ALL formspec elements use relative sizes, not absolute pixel counts. why make img in hypertext the exception? 😮
21:55 kilbith joined #minetest-dev
22:47 kilbith_ joined #minetest-dev
23:07 kilbith_ HTML standards man
23:09 kilbith__ joined #minetest-dev
23:20 nephele html in formspecs?
23:20 turtleman joined #minetest-dev
23:21 Wuzzy no
23:23 Wuzzy formspecs will get a hypertext[] formspec element. It's a simple markup language, but it's NOT the same as HTML. It's also not SGML-compatible.
23:24 nephele is there a design spec for that?
23:26 Wuzzy idk
23:27 Wuzzy but you can look into lua_api.txt to see how it works
23:29 Wuzzy nephele: I have recently added a test for the hypertext[] in devtest:
23:29 Wuzzy https://git.minetest.land/Wuzzy/devtest/src/branch/master/mods/testformspec/formspec.lua
23:29 Wuzzy to test in-game, say "/test_formspec" then click "Hypertext"
23:30 nephele Oof, that heavily looks like html
23:31 nephele I kind of wonder why it isn't just normal formspec syntax?
23:40 rubenwardy because normal formspec syntax is terrible
23:44 nephele I disagree, but even if that were the case copying html and using both is surely worse .-.
23:45 kilbith joined #minetest-dev
23:45 Wuzzy html was not copied.
23:45 Wuzzy its NOT HTML
23:45 Wuzzy it just looks kinda like it
23:46 fluxflux joined #minetest-dev
23:46 nephele It looks very much like a subset of it
23:46 kilbith WUzzy always liking to argue
23:46 Wuzzy its not a subset either
23:46 kilbith is there anything that relax you in life
23:47 nephele If the goal is to be /not/ html then copying it's brace style isn't a good idea either imo, if it's different atleast make it properly different
23:47 Wuzzy tbh i dont care
23:47 Wuzzy btw rubenwardy how do you escape "<" and ">" in hypertext?
23:51 kilbith_ joined #minetest-dev
23:52 kilbith_ nephele: do you have a better idea?
23:58 nephele I'd likely just do it with the formspec syntax as is... i don't really like having to switch context like that :)
23:58 kilbith_ Wuzzy: \<\>
23:58 kilbith_ just checked
23:59 nephele (that would be regardless of whether the formspec syntax changes or not=
23:59 rubenwardy Wuzzy:  formspec_escape
23:59 kilbith_ nephele: was that you which I had an argument on this topic few months ago?
23:59 nephele I don't know
23:59 Wuzzy rubenwardy: wat.

| Channels | #minetest-dev index | Today | | Google Search | Plaintext