Minetest logo

IRC log for #minetest-dev, 2016-01-16

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

All times shown according to UTC.

Time Nick Message
00:36 est31 joined #minetest-dev
00:52 Taoki joined #minetest-dev
01:13 jin_xi joined #minetest-dev
01:28 Fritigern The new creative inventory looks great guys. However, I also have WorldEdit and its icon shows up on every page of every tab and covers an inventory item. Dunno if CI is responsible for this or WE.
01:28 sofar technically WE
01:28 sofar and yes, I've seen that too
01:34 Taoki joined #minetest-dev
01:37 Player2 joined #minetest-dev
02:25 lezzy joined #minetest-dev
03:19 nolsen joined #minetest-dev
04:06 turtleman joined #minetest-dev
05:48 Lunatrius joined #minetest-dev
06:56 nrzkt joined #minetest-dev
07:10 fling joined #minetest-dev
07:14 sofar I must have hit some weird lua bug....
07:19 sofar I've made a lua object and a function to get a handle to it
07:19 sofar then some methods in the object
07:20 sofar a few of those methods have a parameter
07:20 sofar but the content of the paramater is always the object itself, not what was passed into it
07:44 sofar lua objects are ... bizarre
07:47 AnotherBrick joined #minetest-dev
08:15 Hunterz joined #minetest-dev
08:19 celeron55 sofar: so you are calling the methods using : which always makes the first parameter be the object itself?
08:20 Obani joined #minetest-dev
08:20 sofar the problem is the lua documentation, it mixes : and . and doesn't coherently provide a full example that's consistent
08:20 sofar it was clearly written in a rush by someone
08:20 sofar I wanted to design an API for : use
08:21 sofar but I got something wrong in the method declaration
08:28 sofar celeron55: I'm making a decent API for doors
08:28 sofar e.g.
08:28 sofar local d = doors.get(pos)
08:28 sofar if not d then
08:29 sofar print "not a door"
08:29 sofar else
08:29 sofar d:toggle(clicker)
08:29 sofar end
08:32 sofar I'm testing it in-game with //lua doors.get(pos):toggle() -- thanks worldedit!
08:35 H-H-H joined #minetest-dev
08:41 sofar celeron55: https://github.com/minetest/minetest_game/pull/789
08:53 sofar I don't suppose I could provide my own cracking texture, can't i?
09:16 Megaf joined #minetest-dev
09:42 proller joined #minetest-dev
09:57 VargaD joined #minetest-dev
10:01 Krock joined #minetest-dev
10:14 sapier joined #minetest-dev
10:23 lezzy joined #minetest-dev
10:24 lezzy https://github.com/minetest/minetest/pull/3572
10:24 lezzy fixed this one :P
10:27 blaze joined #minetest-dev
10:28 sapier decimal_places?
10:43 Obani joined #minetest-dev
10:53 Calinou joined #minetest-dev
11:45 CraigyDavi joined #minetest-dev
11:52 red-001 joined #minetest-dev
11:52 Fixer joined #minetest-dev
12:05 jin_xi joined #minetest-dev
12:10 Fritigern joined #minetest-dev
12:40 blaze joined #minetest-dev
12:40 paramat joined #minetest-dev
12:45 sdzen joined #minetest-dev
12:48 paramat will merge 3564 3571 3572 very soon
12:58 rubenwardy joined #minetest-dev
13:00 turtleman joined #minetest-dev
13:02 paramat now merging
13:12 sdzen joined #minetest-dev
13:14 paramat merged
13:17 paramat left #minetest-dev
13:22 Gael-de-Sailly joined #minetest-dev
13:41 RealBadAngel sodfar, have you read my comments on sings?
13:41 RealBadAngel *sofar
14:38 sdzen joined #minetest-dev
15:12 Elinvention joined #minetest-dev
15:25 sdzen joined #minetest-dev
15:25 Elinvention joined #minetest-dev
15:27 kaadmy joined #minetest-dev
15:48 behalebabo joined #minetest-dev
16:46 ElectronLibre joined #minetest-dev
16:49 EvergreenTree joined #minetest-dev
17:05 celeron55 joined #minetest-dev
17:13 Elinvention joined #minetest-dev
17:15 celeron55 joined #minetest-dev
17:19 sofar RealBadAngel: yes, you said the metadata update patch blocks it?
17:20 RealBadAngel not really blocks, without that patch working, you wont get reliable signs
17:21 RealBadAngel current "how-things-works" is a blocker
17:26 celeron55 joined #minetest-dev
17:26 sofar I don't think it'll be unreliable
17:26 sofar even without the patch
17:29 jin_xi joined #minetest-dev
17:33 RealBadAngel sofar, you may try. do you want to see my code for signs?
17:35 nrzkt joined #minetest-dev
17:36 celeron55 joined #minetest-dev
17:55 celeron55 joined #minetest-dev
17:57 sofar RealBadAngel: sure
18:05 celeron55 joined #minetest-dev
18:16 celeron55 joined #minetest-dev
18:16 void7 joined #minetest-dev
18:26 celeron55 joined #minetest-dev
18:31 hmmmm joined #minetest-dev
18:36 Ardonel joined #minetest-dev
18:38 celeron55 joined #minetest-dev
18:39 ElectronLibre joined #minetest-dev
18:48 sdzen joined #minetest-dev
18:48 celeron55 joined #minetest-dev
18:53 louiloui joined #minetest-dev
18:59 celeron55 joined #minetest-dev
18:59 rubenwardy joined #minetest-dev
19:09 celeron55 joined #minetest-dev
19:13 sofar geesh minetest.after(1, ....) can take 5-6 seconds to elapse
19:13 sapier define "elapse"?
19:13 sapier elapse
19:13 sofar before func gets called
19:13 sapier well that's perfectly fine behaviour
19:14 sapier minetest.after is defined as "wait AT LEAST x seconds"
19:14 sofar I need an event to occur 0.2 seconds after it gets scheduled
19:14 sapier well use a realtime os
19:14 sofar lol
19:14 sapier not even plain linux does guarantee a event to be triggered in 200ms
19:15 sofar the accuracy isn't important
19:15 sapier well then 5 seconds are fine aren't they?
19:15 sofar if I ask for 0.2 seconds, I can live with it being 0.3 seconds
19:15 sapier 5 > 1 > 0.2 thus it's ok
19:15 sofar 5 seconds isn't ok for the effect I want
19:15 sapier nothing to be done in 0.2 second interval is reasonable done in lua
19:16 sapier any major mod running in parallel will break whatever you want to do in this interval
19:16 sofar sure, I'll take some variation
19:17 sofar 1 second would be nice
19:17 sapier even network communication will do so, not talking about someone downloading textures from that server
19:17 sofar which is why I tried minetest.after(1, ...)
19:17 sapier I'd guess in regular cases you'll get something about 1s but there will be a lot of cases where you don't get it ... depends on what other mods are running of course
19:18 sofar that's totally acceptable
19:18 sofar I know I can use node timers, but that makes things complex quickly
19:18 sofar it means putting a special node somewhere, saving state before it was placed, etc.
19:18 sapier right now minetest does warn about abms running  >200ms ... and abms are only part of what is done per server step so assume a server step interval of up to 0.5 seconds (without any mod) is still considered ok
19:19 sofar I guess I'll have to ... register_globalstep()
19:19 sapier I doubt this will help
19:20 sapier everything done by lua is done exactly once per serverstep no matter if it's globalstep or timer
19:20 sapier in fact timer is done in globalstep too the minetest.after is just a convenience wrapper
19:20 sofar right, but minetest.after doesn't allow me to do something the NEXT globalstep
19:21 sapier well I haven't looked at code so this could be right
19:23 celeron55 joined #minetest-dev
19:23 sapier but still a requirement of 200ms is quite heavy I'd not expect this to work if you don't have a completely clean minetest game without any other mods
19:24 sofar there are better solutions, but they require patching the game
19:24 sofar I'm looking to add a light effect to my new lightning mod
19:24 sapier I don't know what you're doing but I'd suggest making whatever you do more tolerant to delays
19:24 sapier puhh ... hard
19:24 sofar lighting up the sky for 5-6 doesn't look like lightning
19:24 Warr1024 joined #minetest-dev
19:25 sapier especally as everything you do there has to be sent to client too so network jitter and delay will add to whatever server already has
19:25 sofar particle lighting would be great, as particles can exist for very short
19:25 sofar like I said, 0.5sec variation is fine
19:25 sofar not 5 seconds
19:25 sofar ligtning itself is jittery
19:25 sapier we can't even guarantee 5 seconds via network
19:26 sofar in real life, that is
19:26 sofar yes, that's why particle light would work
19:26 sofar that's all client-side
19:26 sofar the lightning particle exists for 0.2 seconds
19:26 sapier well good luck ;-) but I believe patching game would provide better results ;-)
19:32 Fixer 200 ms? why so slow?
19:32 sofar lol
19:32 sapier slow?
19:32 sofar ahahaha
19:33 sapier I always wonder what ppl believe their crapy pc can do :-)
19:34 celeron55 joined #minetest-dev
19:34 sapier nothing against your pc fixer ;-) it's just about ANY PC having that issue ;-)
19:37 Krock I don't like "crapy pc".. I could be one of those who uses such a computer
19:38 sapier Well if you need 200ms this is a realtime requirement and there's no pc no matter if it's linux windows macos or any other general purpos os which will guarantee this
19:38 Fixer joined #minetest-dev
19:39 sapier on linux you can at least add the RT patches those most likely will enable your machine to fullfill 200ms requirements
19:39 sapier but you'd have to run minetest as root ;-)
19:42 H-H-H joined #minetest-dev
19:44 celeron55 joined #minetest-dev
19:46 sofar sapier: nonsense, we're not audio idiots who need nano second accuracy
19:46 sofar sapier: I never said it had to be lag-free
19:48 sofar oh wow globalstep works nice
19:48 sapier Well I just said 200ms guarantee ain't possible with general purpose os ;-) that's pure sense! but it was not targeted at your requirement as you already said you don't need a guarantee
19:49 sapier I suggest testing on server, localhost is often missleading :-) know this from own experience ;-)
19:50 Calinou joined #minetest-dev
19:51 void7 joined #minetest-dev
19:51 sapier But there's one important point if globalstep works in 0.2s and minetest.after doesn't in 5s this may indicate a bug in minetest.after
19:59 sofar I should make a timer and see what the jitter is
20:30 Elinvention joined #minetest-dev
20:53 void7 joined #minetest-dev
21:02 Amaz joined #minetest-dev
21:18 Skyduskguy joined #minetest-dev
21:19 Skyduskguy request to make liquid movment consistant or the settings to be
21:44 kaeza joined #minetest-dev
21:46 luizrpgluiz joined #minetest-dev
21:54 Fritigern joined #minetest-dev
21:58 luizrpgluiz joined #minetest-dev
21:59 luizrpgluiz hello, I read on github that Parana had a significant loss of even frames with disabled shaders, this is a mistake of some file in minetest?
21:59 luizrpgluiz hehe Paramat
22:17 red-001 joined #minetest-dev
22:41 BrandonReese joined #minetest-dev
22:54 Ardonel joined #minetest-dev
23:02 sofar ShadowBot: game#790
23:02 sofar I must be doing that wrong...
23:23 DogePony joined #minetest-dev
23:24 Megaf RealBadAngel, perhaps you stop implementing new stuff on minetest and begin fixed this
23:24 Megaf [730709.159806] Out of memory: OOM killed process 1710 (minetestserver) score 0 vm:1360700kB, rss:901000kB, swap:212764kB
23:24 Megaf I run minetest servers for 4 years, and it always had this bloody memory leaks
23:25 Megaf as I said days ago, begin cleaning the code, removing old cold
23:25 Megaf code
23:25 RealBadAngel Megaf, ive used to code client side stuff rather
23:25 RealBadAngel idk how such things can be related to server out of ram
23:26 Megaf I'm not blaming you, I'd like time to be used in fixing bugs and leaks rather than implementing new features
23:26 RealBadAngel you just did
23:28 proller joined #minetest-dev
23:32 RealBadAngel and when talking bout fixing bugs: https://github.com/minetest/minetest/commi​t/da686160c33069b77a7b4c31fd97da5fa5b59baa
23:32 RealBadAngel i do fix things when i can
23:32 red-001 xyz
23:32 red-001 hi; so the minetest.ru server that hosts IRC logs might be getting disposed of soon-ish and so I wondered if somebody here wanted to take over the logs? (I've talked to c55 but he's not interested)
23:32 red-001 thats what xyz said on #minetest
23:33 Megaf RealBadAngel, also I think we could drop JPEG support
23:33 Megaf I dropped JPEG on my fork and didnt notice a thing of difference
23:33 Megaf (in gameplay)
23:35 RealBadAngel theres lotsa other really dead code to remove rather that dropping something that actually can be used
23:35 Megaf RealBadAngel, and that was a bug introduced by someone, my fork is older than the bug
23:35 RealBadAngel what bug?
23:36 Megaf the wield mesh
23:37 RealBadAngel im not yet sure what is causing the mesh in question to be displayed fault, yet im able now to reproduce it under certain conditions
23:37 RealBadAngel so i will propably be able to fix it
23:37 sfan5 <Megaf> RealBadAngel, perhaps you stop implementing new stuff on minetest and begin fixed this
23:37 sfan5 <Megaf> I'm not blaming you
23:37 sfan5 wat
23:38 Megaf oh, I really didn't mean to
23:38 Megaf sorry
23:39 Megaf What I mean is, I'd like people to use time to fix bug, not to add new stuff
23:40 RealBadAngel Megaf, working on new stuff usually comes with fixing already extistin problems
23:40 RealBadAngel for example using meshes in inventory eliminates need for preloading item visuals, speeds up displaying lists and reduces ram usage
23:41 RealBadAngel if the items do spin at the same time you can call it added value ;)
23:42 Megaf RealBadAngel, with similar thinking I'm just removing lots of code I don't want
23:42 RealBadAngel also  the fix we have made together with kahrl is made only thx to work on inventory meshes
23:42 Megaf if we have a memory leak (IT IS JUST AN EXAMPLE) on the display of JPEG, now we don't do anymore
23:44 RealBadAngel wanna remove something? wait a sec
23:45 RealBadAngel then delete https://github.com/minetest/minetest/​blob/master/src/content_cao.cpp#L135 to line #534
23:45 RealBadAngel thats completely dead code
23:47 Megaf lol, 400 lines of dead code?
23:47 kahrl RealBadAngel: it is not dead, it is used when connecting to older version servers that still have the corresponding SAOs
23:47 kahrl granted, it might be the case that there is no server that old, but I haven't checked (and you can't check all private servers)
23:48 RealBadAngel kahrl, can you point me to single use of test or item caos?
23:48 Megaf kahrl, thats why git master is used for, its the bleeding edge not worried about compatibility
23:48 Megaf if you dont want your server to break, stick to 0.4 for instance
23:48 Megaf that code removal could go to 0.5
23:49 Megaf 0.4 to 0.5 is major change
23:49 RealBadAngel kahrl, theres no single use for those CAOs
23:49 Megaf 0.4.x to 0.4.y is small change
23:49 Megaf either way, code should get removed
23:49 Megaf RealBadAngel, I will indeed try to remove that
23:49 Megaf hold on
23:49 RealBadAngel just remove it and compile
23:49 kahrl RealBadAngel: https://github.com/minetest/minetest/​blob/master/src/clientobject.cpp#L56
23:50 RealBadAngel well hidden indeed :)
23:51 RealBadAngel when itemcao was last used?
23:51 Megaf what is that anyway?
23:51 kahrl you mean ItemSAO?
23:51 kahrl it was removed from the codebase around February last year by nerzhul IIRC
23:51 RealBadAngel i mean ItemCAO
23:51 kahrl ItemCAO is still used, like I said
23:52 kahrl when you connect to older servers
23:52 RealBadAngel i mean how old?
23:52 kahrl not exactly sure, but <kahrl> it was removed from the codebase around February last year by nerzhul IIRC
23:53 kahrl of course servers were using lua entities for items long before that
23:53 kahrl (but old mapblocks could still contain ItemSAOs up to that point, now they are ignored when deserializing mapblocks)
23:55 RealBadAngel i see

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