Minetest logo

IRC log for #minetest-dev, 2017-02-16

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

All times shown according to UTC.

Time Nick Message
01:21 ptv joined #minetest-dev
01:34 kaeza joined #minetest-dev
02:08 twoelk left #minetest-dev
02:16 Miner_48er joined #minetest-dev
04:23 Hunterz joined #minetest-dev
05:38 ssieb joined #minetest-dev
05:51 YuGiOhJCJ joined #minetest-dev
05:56 lumidify joined #minetest-dev
05:59 DonBatman joined #minetest-dev
06:07 Hunterz joined #minetest-dev
06:56 lumidify joined #minetest-dev
07:35 est31 joined #minetest-dev
07:42 lumidify joined #minetest-dev
07:56 red-001 joined #minetest-dev
07:57 Darcidride joined #minetest-dev
07:57 red-001 joined #minetest-dev
07:57 red-001 joined #minetest-dev
08:09 nerzhul joined #minetest-dev
08:13 Player_2 joined #minetest-dev
08:23 kilbith joined #minetest-dev
08:54 Gael-de-Sailly joined #minetest-dev
10:19 lisac joined #minetest-dev
11:14 xerox123 joined #minetest-dev
11:17 xerox123 joined #minetest-dev
11:24 Fixer joined #minetest-dev
11:48 proller joined #minetest-dev
11:52 Darcidride joined #minetest-dev
11:52 paramat joined #minetest-dev
11:55 paramat nerzhul nore sofar Zeno` please can i merge #5223 later? simple crashfix
11:55 ShadowBot https://github.com/minetest/minetest/issues/5223 -- Objectpos over limit: Avoid crash caused by sector over limit by paramat
11:57 paramat VanessaE "is 5223 also meant to address http://pastebin.ubuntu.com/24003628/" no, seems a different issue
11:58 sfan5 merging #5243 now since it's a trivial regression fix
11:58 ShadowBot https://github.com/minetest/minetest/issues/5243 -- Fix a small regression caused by #5078 by red-001
11:59 paramat +1
12:03 paramat mapgen PR #5237 can anyone approve? i know exactly what i'm doing here :]
12:03 ShadowBot https://github.com/minetest/minetest/issues/5237 -- Cavegen: Place correct biome surface in tunnel entrances by paramat
12:06 sfan5 code looks fine to me
12:06 sfan5 can't comment on the functionality
12:07 paramat ok, it's tested (5237)
12:14 paramat sfan5 so is that a +1 for 5237?
12:14 sfan5 yes
12:16 paramat thanks, will also merge #5234 later
12:16 ShadowBot https://github.com/minetest/minetest/issues/5234 -- Re-add Droid Sans as fallback font (Reverts part of eb490) by sofar
12:19 paramat celeron55 concerning #5239 is there a reason 26 neighbour blocks are added instead of 6?
12:19 ShadowBot https://github.com/minetest/minetest/issues/5239 -- Only add direct neighbors for each MapBlock loaded on the client. by lhofhansl
12:24 paramat ah looks like there is https://github.com/minetest/minetest/commit/5c26972cddb78f3f93d5d6e0428a6b282fde4bfe
12:24 paramat will investigate
12:25 lumidify joined #minetest-dev
12:28 paramat hm there's no PR for it
12:29 paramat celeron55 could you describe the glitch your commit fixed so we can check again whether 26 is necessary?
12:31 celeron55 meshes at the corners depend on those
12:32 paramat ok
12:42 AnotherBrick joined #minetest-dev
12:44 sfan5 wouldn't it be better to do thing depending on how many where the meshes are?
12:50 twoelk joined #minetest-dev
12:53 lumidify joined #minetest-dev
13:00 kilbith joined #minetest-dev
13:32 paramat woo at this rate we'll be at 900 issues soon
13:33 Darcidride joined #minetest-dev
13:35 ircSparky joined #minetest-dev
13:39 rubenwardy joined #minetest-dev
14:17 Darcidride joined #minetest-dev
14:57 octacian joined #minetest-dev
14:57 octacian joined #minetest-dev
15:37 nerzhul joined #minetest-dev
15:47 lisac joined #minetest-dev
15:57 paramat sfan5 please could you look at #5223 ?
15:57 ShadowBot https://github.com/minetest/minetest/issues/5223 -- Objectpos over limit: Avoid crash caused by sector over limit by paramat
16:09 red-002 #5232 is a duplicate of #4087
16:09 ShadowBot https://github.com/minetest/minetest/issues/5232 -- Falling nodes loses their meta info
16:09 ShadowBot https://github.com/minetest/minetest/issues/4087 -- Node metadata is lost on falling nodes
16:11 red-002 Fixer, maybe just make a pr for #5245 ?
16:11 ShadowBot https://github.com/minetest/minetest/issues/5245 -- Clarify loading of minetest.conf in Readme.txt
16:12 Fixer i can try it
16:13 red-002 btw I think that "...configuration file in
16:13 red-002 $location_of_exe/../minetest.conf ..." will sound better
16:14 red-002 instead of 'from' use 'in'
16:15 VanessaE neither.
16:15 VanessaE "at" is more appropriate in this instance.
16:15 VanessaE if you were just giving the directory/path but not the filename, "in" is more appropriate
16:15 red-002 oh good point
16:16 red-002 yeah that's a better way to write it
16:16 VanessaE "from" works too since the file is being pulled "from" a specific location.
16:16 red-002 It just kinda sounds strange
16:18 VanessaE "from" would work better if the word "file" were dropped.
16:18 VanessaE i.e. "loading configuration from $minetest_path/minetest.conf"
16:18 red-002 yeah it would work find
16:18 red-002 ehh
16:18 red-002 fine*
16:19 red-002 options on #5242 ?
16:19 ShadowBot https://github.com/minetest/minetest/issues/5242 -- Dumping/saving console history
16:20 nerzhul when do you think we should merge CSM ? development is flat these days and we have the base for making it available into master
16:21 nerzhul its security is sufficient
16:21 VanessaE does what exists work?  will it break anything by being merged?
16:21 red-002 It shouldn't break anything
16:21 nerzhul nothing will break as it's a pure client side thing which only connects like server to current client events
16:22 nerzhul it works great, the death formspec has been replaced with a pure CSM
16:22 nerzhul for example
16:22 VanessaE then I'd say to merge it
16:22 nerzhul yeah but you are not a core dev, i ask for core devs :p
16:22 VanessaE :P
16:24 red-002 I think there is a small graphical glitch in the death formspec. The text seems to be not centred correctly so it's off a bit to the side
16:24 red-002 in csm
16:24 red-002 I haven't seen any other issues
16:29 red-002 I think it is caused by the removal of the fixed size part of the size tag
16:30 red-002 so that it will work correctly on android
16:36 garywhite joined #minetest-dev
16:46 YuGiOhJCJ joined #minetest-dev
16:47 Hunterz joined #minetest-dev
16:57 twoelk|2 joined #minetest-dev
17:22 lumidify joined #minetest-dev
17:58 paramat joined #minetest-dev
18:04 Darcidride joined #minetest-dev
18:10 paramat "development is flat"? as in slow?
18:15 red-002 sounds like it
18:16 red-002 #5248
18:16 ShadowBot https://github.com/minetest/minetest/issues/5248 -- Move warnings about not being able to load the node to infostream. by red-001
18:18 lumidify joined #minetest-dev
18:29 paramat yeah it's slow because est and hmmmm are inactive, and so-far is busy with a big project, there's still a lot of work that needs doing, for example reviewing #4967 #5227
18:29 ShadowBot https://github.com/minetest/minetest/issues/4967 -- New bulk node light update by juhdanad
18:29 ShadowBot https://github.com/minetest/minetest/issues/5227 -- [squashed] Refactor content_mapblock by numberZero
18:31 paramat that refactor is delaying other work in that file
18:40 paramat i intend to have a look but i'm not good with C++ structural stuff (classes etc.) so that needs reviewing by other devs
18:44 red-002 what about merging client-sided modding?
18:55 paramat i can't comment, i don't understand it at all, but since it's a big feature it should be done well before a release for testing
18:56 paramat i remember nerz intended it to be merged soon before a release? the next release is not soon, probably june
18:57 paramat perhaps May
18:58 red-002 it would be easier to work on it if it's in master
18:58 red-002 wouldn't need to keep rebasing the branch
18:59 paramat maybe ask sapier for his opinion since he seems knowledgeable about it
18:59 paramat avoiding rebasing isn't a reason for merging, it has to be good enough to merge first
19:13 ptv joined #minetest-dev
19:20 Fritigern joined #minetest-dev
19:21 ptv joined #minetest-dev
19:25 sofar paramat: I may have some extra time next week to review a bunch of PRs in depth more
19:33 twoelk left #minetest-dev
19:34 paramat nice
19:48 OldCoder sofar, red-002 are either of you familiar with LBMs ?
19:55 nerzhul sofar, can you review the CSM to make it mergeable and usable for everybody as a preview ?
19:55 nerzhul and permitting everybody to add features easily ?
20:15 est31 joined #minetest-dev
20:19 sofar nerzhul: maybe next week, it's certainly something I want to test/review
20:19 sofar OldCoder: yes, I am familiar with LBM's... I was the first one to use them and tested them with est
20:22 betterthanyou711 joined #minetest-dev
20:33 OldCoder sofar, the load every time flag does not work. So I use a random name; this does work. But LBM install times are saved somewhere....
20:33 OldCoder If I use random names, will a file somewhere grow without limit?
20:33 sofar yes, that's not a solution
20:33 OldCoder It is the only one I have found
20:33 OldCoder Is the problem with the load every time flag known ?
20:33 OldCoder sofar, ^
20:34 sofar `run_at_every_load = true,` ?
20:34 OldCoder Yes
20:34 sofar not `1` ?
20:34 OldCoder It does not work; this is definitive
20:34 OldCoder Hm; I will confirm the last part
20:34 OldCoder Wait
20:35 OldCoder sofar, I am using 1. Most conf-file settings take true or 1 for boolean, correct? Does this setting not do so?
20:35 sofar no, it has to be `true`
20:35 OldCoder This is a bite developers on the *ss thing. Out of curiosity, why is this?
20:36 OldCoder Just asking... it seems inconsistent with settings in general
20:36 VanessaE OldCoder: several of my mods use LBMs and they work just fine with run-at-every-load.
20:36 sofar well minetest.conf isn't parsed as lua
20:36 OldCoder See above, VanessaE
20:36 sofar lua_pushboolean is the problem
20:36 OldCoder VanessaE, sofar, thank you
20:36 OldCoder This is helpful
20:37 * OldCoder goes off to look at booleans
20:37 OldCoder
20:37 sofar src/lua/src/lapi.c:503
20:37 sofar maybe we can wrap it somehow, to allow `1` to mean `true`
20:40 sfan5 patching lua to enable non-standard behaviour that helps a few random people in rare situations is a bad idea
20:41 sofar that I agree with
20:41 sofar but in the case of parsing registrations
20:42 sofar node/lbm etc. flags
20:42 sfan5 then minetest should use a replacement for parsing "bools"
20:42 sfan5 tho i'm still against that
20:42 sofar hmm, come to think of that I don't see an easy way to do that anyway
20:44 sofar sfan5: it would be nice though if we could throw an assert if people try to pass an `int` to a parameter that requires a `bool`
20:44 sfan5 yeah that would be nice
20:48 sofar maybe a "check_bool(L, index)" ?
20:49 sofar I could probably throw asserts for hours at our lua api
20:49 paramat joined #minetest-dev
20:55 paramat celeron55 the author of #5239 cannot find any glitches when using 6 block neighbours, could you describe a specific situation that causes glitches with 6, so that we can test?
20:55 ShadowBot https://github.com/minetest/minetest/issues/5239 -- Only add direct neighbors for each MapBlock loaded on the client. by lhofhansl
20:58 paramat devs, i will merge #5223 in an hour (with some other PRs). sorry to be impatient but it's causing crashes and the commit is very simple and tested
20:58 ShadowBot https://github.com/minetest/minetest/issues/5223 -- Objectpos over limit: Avoid crash caused by sector over limit by paramat
20:59 paramat will merge #5234 #5237 with it
20:59 ShadowBot https://github.com/minetest/minetest/issues/5234 -- Re-add Droid Sans as fallback font (Reverts part of eb490) by sofar
20:59 ShadowBot https://github.com/minetest/minetest/issues/5237 -- Cavegen: Place correct biome surface in tunnel entrances by paramat
21:09 VanessaE paramat: can we give #5186 some attention also?
21:09 ShadowBot https://github.com/minetest/minetest/issues/5186 -- Soft node overlay by juhdanad
21:11 proller joined #minetest-dev
21:11 celeron55 paramat: smooth lighting IIRC
21:12 celeron55 irc logs from the time of that commit might help
21:15 celeron55 looks like we have public logs available from that day
21:15 celeron55 http://irc.minetest.net/minetest-dev/2012-11-27
21:15 celeron55 https://github.com/minetest/minetest/issues/307
21:15 celeron55 the glitch report and fix came from doserj, it looks like
21:16 celeron55 and performance was a concern back then too, but considered good enough
21:17 celeron55 apparently rails might glitch too; no idea how exactly
21:17 paramat thanks
21:19 paramat lighting has changed a little since then so we can test this again
21:20 celeron55 posted on the PR also
21:21 paramat soft node overlay isn't a priority compared to the light bug fixing PR, but there are devs enthusiastic about it so i'm sure it won't be neglected
21:27 paramat *=/ i should have checked logs for that day
21:29 Fixer https://i.imgur.com/5O5bR.jpg this also demonstrates cubic shape of \_\ of light near torches, very annoying shape in some situations
21:29 Fixer wonder how it looks those days
21:30 sofar try it, should be much better
21:31 paramat new lightcurve improves that
21:31 Fixer yeah, will check
21:31 Fixer i have a world with this problem
21:32 Warr1024 joined #minetest-dev
21:33 sofar http://i.imgur.com/75IQSXg.png
21:35 Fixer just checked - MUCH BETTER
21:37 Fixer looks nice now
21:38 Fixer i even posted an issue about this, need to close it down
21:39 Fixer no more this crap: https://cloud.githubusercontent.com/assets/16494741/16126482/0a40ce92-3401-11e6-96c3-530290f6d7ad.png
21:40 Fixer yay
21:40 Fixer that was ridiculous
21:45 Fixer i will play some time with 5239
21:48 Fixer https://github.com/minetest/minetest/pull/4967
21:48 sofar http://i.imgur.com/tlR8vAf.png
21:48 sofar still a little bit
21:50 Fixer sofar: just a little, still a huge improvement
21:50 Fixer sofar: immerse 1 block deep and it goes away completely imo
21:51 Fixer immerse mese lamp*
21:52 betterthanyou711 joined #minetest-dev
21:56 sofar it's the first light step that is the largest
21:56 sofar that makes it easy to spot
21:57 paramat on this subject game#1574
21:57 ShadowBot https://github.com/minetest/minetest_game/issues/1574 -- Torches: Reduce light source level from 13 to 12 by paramat
21:57 sofar we need more devs with unpronouncable github usernames
21:57 sofar like lhofhansl and juhdenad
21:57 sofar maybe I can come up with some
21:57 sofar berflagzr
21:57 sofar grubzgews
21:58 sofar kragzmendi
21:58 paramat i like these
22:02 Fixer i'm trying to spot some bugs with 5239
22:03 Fixer on vanessa creative building at night
22:04 red-002 Fixer, could you change 'configuration file' to just 'configuration'?
22:04 Fixer red-002: no, this will take a lot of time to figure out for me right now
22:05 Fixer %)
22:05 red-002 git you mean?
22:05 Fixer yes
22:05 * red-002 uses a GUI
22:10 Fixer finding bugs with 5239 will be tough :(
22:11 Fixer everything looks fine from first look/fly around
22:14 Fixer anyone measured fps with 5239?
22:17 proller joined #minetest-dev
22:19 paramat how to reproduce the bug is explained here #307
22:19 ShadowBot https://github.com/minetest/minetest/issues/307 -- smooth lighting has a glitch on mapblock boundaries
22:22 paramat will merge 5223 5234 5237 in a few mins
22:26 Fixer oh, will look
22:30 Fixer i only see this: https://i.imgur.com/Uk3mAUg.png will check on build without 5239
22:32 Fixer build without it https://i.imgur.com/6Rp5eMO.png
22:35 paramat merging
22:35 betterthanyou711 joined #minetest-dev
22:38 paramat done
22:40 paramat hm subtle difference
22:42 paramat might, possibly, even be bearable for the advantages
22:43 paramat (but unsure about that) best check at a mapchunk corner too
22:44 Fixer and thats why I asked analog of MC "show mapchunk borders" :/
22:44 Fixer much easier in those situations
22:44 Fixer or mapblock borders
22:45 sofar could just make a mod that does that
22:45 Fixer paramat: fps looks the same, meh
22:45 Fixer sofar: please do :trollface:
22:45 sofar spawn a quick entity that is just a vertical sprite at boundaries
22:49 Fixer sofar: smth like that https://media-curse.cursecdn.com/attachments/173/760/def11af2a10a616f2e80ca295225e76b.png
22:50 sofar lines would already work, yes
22:51 Fixer lines are enough
22:51 Fixer detailed lines are not needed probably
22:51 Fixer but will be a nice option
22:55 sofar I would strongly suggest a mod for that
22:55 sofar that shouldn't go in the engine
22:55 paramat max corner is at 47, 47, 47
22:55 sofar write it and I'll put it in minetest-mods ;)
23:03 Fritigern http://www.47.net/47society/
23:05 Fixer https://i.imgur.com/8tQ0KF4.png
23:07 Fixer paramat: https://cloud.githubusercontent.com/assets/16494741/23045494/7a231d10-f4ad-11e6-9a0a-dd322f3aacd6.png
23:10 lordfingle joined #minetest-dev
23:12 lisac joined #minetest-dev
23:15 paramat oh crumbs another thing that is 47
23:16 paramat aha worse at corner
23:49 betterthanyou711 joined #minetest-dev

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