Minetest logo

IRC log for #minetest-dev, 2016-10-04

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

All times shown according to UTC.

Time Nick Message
00:18 proller joined #minetest-dev
00:25 torgdor joined #minetest-dev
00:34 proller joined #minetest-dev
00:36 Tmanyo joined #minetest-dev
00:52 diemartin joined #minetest-dev
01:03 ssieb joined #minetest-dev
01:16 paramat one for hmmmm #4490 bbl
01:16 ShadowBot https://github.com/minetest/minetest/issues/4490 -- VoxelArea: faster iter function by HybridDog
02:08 torgdor joined #minetest-dev
02:23 paramat joined #minetest-dev
02:53 lordfingle joined #minetest-dev
02:55 paramat game#1303
02:55 ShadowBot https://github.com/minetest​/minetest_game/issues/1303 -- Default: Add grey sand by paramat
03:05 DI3HARD139 joined #minetest-dev
03:19 paramat left #minetest-dev
03:34 kaeza joined #minetest-dev
04:34 kaeza joined #minetest-dev
05:16 Hunterz joined #minetest-dev
05:18 Hunterz joined #minetest-dev
05:50 nrzkt joined #minetest-dev
06:36 red-999 joined #minetest-dev
06:51 Anchakor_ joined #minetest-dev
07:06 HonoredGlory joined #minetest-dev
07:10 Darcidride joined #minetest-dev
07:28 ensonic joined #minetest-dev
07:36 nore joined #minetest-dev
07:54 nrzkt joined #minetest-dev
08:51 _Megaf sofar: sfan5: #4495
08:51 ShadowBot https://github.com/minetest/minetest/issues/4495 -- Using Zstandard for mapblock compression?
08:51 _Megaf Why not LZMA :)
08:54 Calinou it needs to be fast
08:54 Calinou .tar.xz (LZMA2) is just very slow
08:58 olivattaque joined #minetest-dev
08:59 FabulousGlory joined #minetest-dev
09:15 tenplus1 joined #minetest-dev
09:15 tenplus1 Hi folks...
09:15 tenplus1 could devs please look at pull #4425
09:15 ShadowBot https://github.com/minetest/minetest/issues/4425 -- Assertion failure: sanitize some values obtained from mods (yaw, velocity, etc.) by Rogier-5
09:16 tenplus1 we need this in the game to either print an appropriate error message showing line error OR correct NaN values within mod
09:48 olivattaque is it possible to make the player invisible for mobs ? i don't mean invisible like transparent but that mobs don't attack the player ?
09:49 tenplus1 mobs redo and invisibility mod already does this
09:49 tenplus1 I've tied them both in to work together... take invisibility potion and mobs cant see you to attack
09:49 tenplus1 github.com/tenplus1 <-- get them both here
09:50 olivattaque but i need that the player remain visible and can be pointed by other players, is that the case ?
09:50 tenplus1 if player is still visible then mobs will SEE player and attack...
09:51 olivattaque erf so there is no way ? :/ I'm trying to create my first mod, but i'm running into a lot a modding api limitations :(
09:52 tenplus1 what are you trying to do ?
09:52 tenplus1 stop mobs from attacking players ?
09:52 olivattaque yep i'm trying to implement something like second chance in call of duty
09:53 tenplus1 second chance ???
09:53 olivattaque if you die you stay on the ground for 20 seconds and you can resurrect if you kill a mob or if a player resurect you otherwise you actually die
09:54 olivattaque since i can't seem to do anything with register on die, i use register on hp change with condition player_hp == 1
09:54 tenplus1 hrm, sounds like a pain to programme
09:56 tenplus1 in mobs api we use a table called:  invis[player_name].... if this is set to TRUE for player then they are considered invisible to mobs
09:56 tenplus1 you could use this instead, but it will be overriden by invisibility mod if in use
09:57 tenplus1 e.g.    invis["singleplayer"] = true     ...   hides singleplayer from mobs until it's set to nil
09:57 tenplus1 would that help any ?
09:59 tenplus1 I'll make it global by changing to mobs.invis for such things
09:59 olivattaque ah yes i guess it would work
10:00 olivattaque btw i forgot that there is no mobs api in core, so my mod could depends or not on mobs redo
10:00 tenplus1 would have to depend on mobs api to get access to mobs.invis table, yes
10:02 olivattaque but it can be optionnal no ? if i put mobs.invis["singleplayer"] = true in my code and your mod is not loaded, there will be no errors no ? or a null exception ?
10:02 tenplus1 mbos redo git updated with global table
10:04 tenplus1 https://github.com/tenplus1/mob​s_redo/blob/master/api.lua#L29  check this
10:04 tenplus1 shows you how to SEE if a global variable exists and use alternative if not
10:04 olivattaque cool thanks a lot !
10:05 tenplus1 :)
10:06 Zeno` joined #minetest-dev
10:40 tenplus1 left #minetest-dev
10:41 troller joined #minetest-dev
10:41 _Megaf folks, obscure bug for you guys, in minetest_game game#1304
10:41 ShadowBot https://github.com/minetest​/minetest_game/issues/1304 -- Weird and serious bug in locked chests
11:03 proller__ joined #minetest-dev
11:15 olivattaque raah is it possible to prevent the death screen ? i tried to return false in on register_ondie but that doesn't seem to work
11:17 SloanOnLinux joined #minetest-dev
11:17 SloanOnLinux joined #minetest-dev
11:23 Fixer joined #minetest-dev
11:35 Calinou nope
11:35 Calinou it's C++ and client-side
11:41 xunto joined #minetest-dev
11:44 olivattaque erf if I can't modify it or prevent it then I can't do my mod :/ that's too bad for a first mod and not encouraging :(
12:06 sfan5 why do we have SER_FMT_VER_HIGHEST_READ and SER_FMT_VER_HIGHEST_WRITE?
12:07 Wayward_One olivattaque, perhaps open a request here: https://github.com/minetest/minetest/issues
12:07 Krock joined #minetest-dev
12:07 Krock joined #minetest-dev
12:11 proller joined #minetest-dev
12:23 olivattaque Wayward_One: yep thx. I'm gonna continue working on this mod how I would want it to work (mostly for training as it's my first) and i'm gonna open a forum thread to discuss the limitations I encountered (see if they are real limitations) and then open an request
12:29 celeron55 sfan5: they weren't there originally; they were added because they were needed due to a weird serialization goof at some point
12:30 sfan5 o.O
12:30 sfan5 what would be the correct way of raising the mapblock serialization version then?
12:30 celeron55 i guess something along the lines of master writing a new version for a while, after which it was dropped before release and people didn't want to mess up worlds due to support being dropped
12:31 celeron55 well, no, because the goof wasn't released
12:32 celeron55 i could be 100% wrong, dig it up yourselfif you really want to know
12:32 sfan5 i guess i'll find out more about it later when i have time
12:47 red-999 joined #minetest-dev
13:15 STHGOM joined #minetest-dev
13:15 STHGOM joined #minetest-dev
13:45 sfan5 so i implemented support for brotli compression (MapBlock node data & metadata)
13:45 sfan5 https://github.com/minetest/minetest​/issues/4495#issuecomment-251391579
13:49 Krock wow nice
13:50 Krock but why do you go over an additional library instead of directly using brotli?
13:50 Krock I can't see more than some (for windows users) useless compiling scripts
13:51 Karazhan joined #minetest-dev
14:01 hmmmm joined #minetest-dev
14:09 Darcidride joined #minetest-dev
14:21 proller joined #minetest-dev
14:58 _Megaf eh, don't people like lzma?
14:58 _Megaf it's well stabilished already
14:59 _Megaf I think we should benchmark worlds using 3 or 4 different compression libs
15:07 SloanOnLinux joined #minetest-dev
15:07 SloanOnLinux joined #minetest-dev
15:19 sfan5 Krock: linux package maintainers (rightfully imo) obsess over linking everything dynamically, if i don't then they will
15:20 sfan5 also this approach was easier for me
15:20 sfan5 _Megaf: lzma is slow
15:20 T4im joined #minetest-dev
15:21 nrzkt _Megaf: lz4 is better than lzma
15:22 _Megaf oh, ok
15:28 sfan5 _Megaf: if you look at the times here http://blog.fefe.de/?ts=a90c6783 (don't mind all the german text)
15:28 sfan5 xz takes 38.787 with 8 threads so basically 310.296 seconds
15:28 KaadmY joined #minetest-dev
15:28 sfan5 while e.g. brotli with -9 takes 57.761 (single threaded, so not * 8)
15:29 _Megaf ok them I removed my comment already
15:29 sfan5 it's just that lzma is not made for our usecase
15:34 sfan5 celeron55: it was pretty much what you said, the issue was weather params: https://github.com/minetest/minetest/commi​t/06cdce1e1231fb5946ac9750f9b53b53c6f7e5f8
15:35 sfan5 which means the correct way of raising the mapblock version would be to skip 26?
15:36 Krock to keep compatibility with these old builds, yeah.
15:37 Darcidride_ joined #minetest-dev
15:38 Krock nrzkt, you can display the '*' character in markdown by escaping it
15:38 Krock ( https://github.com/minetest/minete​st/pull/4566#discussion_r81783974 )
15:43 _Megaf T4im:  don't get it, was that merged already or not? The fix
15:44 _Megaf game#1304
15:44 ShadowBot https://github.com/minetest​/minetest_game/issues/1304 -- Weird and serious bug in locked chests
15:44 nrzkt Krock: right thanks
15:45 T4im not yet; but some other user wanted itemframe support for saplings and seeds fixed and couldn't find the right issue tracker yet x)
15:45 T4im so i started fixing that; and the CONTENT_IGNORE issues are just in the same lines; so i fixed them too
15:45 T4im but not PR'ed that yet
15:46 _Megaf ok
15:46 T4im because seeds kinda unraveled more issues
15:46 T4im I'll push something up this evening
15:47 T4im generally if something is replaced on placement it's nearly always a CONTENT_IGNORE being ironically ignored :P
15:48 T4im i hope you could roll back the chest? or get the content from a backup?
15:48 T4im metadata would be gone with the set_node :(
15:49 _Megaf T4im: I can't, I have some items to him but there's no way of reverting
15:50 _Megaf trees and saplings are very dengerous and destructive
15:50 T4im well; **all** custom on_place are potentially destructive
15:50 _Megaf nether trees just replace everything on their way
15:50 _Megaf and pine trees used to replace everything too
15:50 _Megaf dont know it that was fixed on pine trees
15:50 T4im they all share the same sapling placment iirc
15:50 T4im so one fix to fix them all
15:51 _Megaf hopefully
15:51 T4im well all saplings*
15:51 T4im not all the other cases ;)
15:54 T4im i might have a potential "overall" minetest fix for this class of issue; but I'd love to get some feedback on that from someone with more intimate knowledge of ignore nodes first
15:57 _Megaf yep, that wouldn't be me
16:03 Hunterz joined #minetest-dev
16:04 _Megaf ok https://github.com/minetest/minetest​/issues/4495#issuecomment-251422192
16:04 _Megaf lz4 it is by me
16:05 _Megaf the cpu free thingy I like, don't know what you mean by that tho
16:06 MoNTE48 joined #minetest-dev
16:06 Krock perhaps that it uses the GPU? sounds strange anyway
16:07 _Megaf I don't thing it uses GPU
16:08 _Megaf maybe it is so efficient and well crafted that it doesn't consume CPU time, it's on the fly
16:08 _Megaf who knows...
16:08 _Megaf and why isn't #4549 merged yet? :)
16:08 ShadowBot https://github.com/minetest/minetest/issues/4549 -- Add coloured logs by HybridDog
16:09 _Megaf I knw, est wants some changes, but it looks good so far
16:12 Krock2 joined #minetest-dev
16:14 Krock joined #minetest-dev
16:22 olivattaque Here is my forum thread following our discussion :p https://forum.minetest.net/viewtopic.ph​p?f=47&amp;t=15630&amp;p=235374#p235374
16:25 red-999 joined #minetest-dev
16:41 _Megaf !title
16:41 ShadowBot modding around death screen (second chance mod) - Minetest Forums
16:47 MoNTE48 joined #minetest-dev
17:00 Krock2 joined #minetest-dev
17:05 paramat joined #minetest-dev
17:06 Void7 joined #minetest-dev
17:13 Grandolf joined #minetest-dev
17:14 nrzkt joined #minetest-dev
17:14 paramat game#1303
17:14 ShadowBot https://github.com/minetest​/minetest_game/issues/1303 -- Default: Add grey sand by paramat
17:15 est31 joined #minetest-dev
17:16 Krock joined #minetest-dev
17:26 paramat i'm considering reducing the lebgth of the red damage flash, as i have seen videos where in some situations the screen becomes constantly red, which is blinding in low-light
17:26 paramat (length)
17:26 nrzkt if you reduce it you can do some epileptic issues
17:29 Krock it was already reduced maybe two years ago :)
17:30 est31 i like the current damage flash
17:30 est31 if you have pain you are blind too, no?
17:30 est31 or dunno
17:30 nrzkt est31: src/porting.cpp:626:4: erreur : ‘ICON_DIR’ was not declared in this scope did you know this issue ?
17:30 est31 never was severely wounded in RL
17:31 est31 nrzkt, are you on current master?
17:31 est31 and did you re-run cmake
17:31 est31 nrzkt, there were some issues about it in the past
17:31 est31 you are on BSD rightß
17:31 est31 ?
17:32 red-999 It would be nice if the damage flash was stronger at the sides of the screen
17:33 nrzkt no i'm on arch atm
17:33 nrzkt an i'm in master, i dropped my cmakebuild directory and recreate it with default options
17:34 nrzkt est31, i'm working on a PR to backport some chat administrative options from my fork but i'm faced of this tstrange issue
17:34 nrzkt but => ../cmakebuild/src/cmake_config.h:#define ICON_DIR "/usr/local/share/icons"
17:34 nrzkt it's strange that it's not know
17:34 est31 it may be missing an include then
17:35 nrzkt missing include i think
17:35 est31 but it does work for me
17:35 est31 so maybe some of your changes removed the include
17:35 est31 note: don't include cmake_config.h directly
17:35 ssieb joined #minetest-dev
17:35 twoelk joined #minetest-dev
17:35 nrzkt oh
17:36 nrzkt i have a cmake_config.h in src dir, i think it causes problem, rebuilding without it
17:36 est31 cmake_config.h should be there
17:37 nrzkt yes but not in the right place
17:37 nrzkt it should be in cmakebuild/src/cmake_config.h not src/cmake_config.h
17:37 nrzkt that solve the issue, i removed it, i think it was legacy build :)
17:37 nrzkt maybe a fail cmake . at a moment
17:37 nrzkt :)
17:45 paramat i'm not epileptic but i find any visual flash literally painful, the intensity may have to be reduced. anyway it's ridiculous, at the one time you really need to see and fight, blinded by constant red screen
17:45 est31 its punishment for getting hit :)
17:45 paramat heh
17:45 paramat damage-tilt does it too for me
17:46 Calinou yeah, should be tweakable
17:46 paramat don't really need both
17:46 Calinou Enemy Territory lets you tweak damage flash too
17:46 Calinou tilt-only is not really visible in my experience
17:46 Calinou I used to play like that (with source modification)
17:46 Calinou way to go would be damage vignette (redden only the sides of the screen) and compass (if possible)
17:46 paramat hm yeah i guess if you're moving around a lot
17:46 Calinou like all the modern games do
17:47 paramat personally i think we could drop damage-tilt
17:48 HonoredGlory joined #minetest-dev
17:48 paramat tilt makes me feel sick :)
17:49 _Megaf wolfer, one of the best arcade style fps ever
17:49 _Megaf wolfet*
17:49 _Megaf Wolfenstein Enemy Territory
17:50 Calinou paramat: we can keep tilt, but only for large damage maybe
17:50 Calinou (half-heart damage shouldn't cause tilting probably)
17:51 _Megaf game##1304 getting interessting
17:51 _Megaf game#1304
17:51 ShadowBot https://github.com/minetest​/minetest_game/issues/1304 -- Serious bug in locked chests. on_rightclick vs on_construct. formspec
17:51 paramat my opinion of vignette is we can't assume the edges of the screen are less important (they're not)
17:51 Fixer_ joined #minetest-dev
17:51 Calinou peripheral vision :) if you want a less obstrusive vignette, use a darkening mode
17:51 Calinou Irrlicht should have that
17:52 Calinou but some people may think it's too blood-like, as it's automatically dark red
17:54 Krock I wonder when bugs aren't meant serious (apart from the one issue that was opened because a developer didn't want to contribute for a moment)
17:55 Krock sadly I can't find it anymore :'(
17:55 torgdor joined #minetest-dev
17:56 T4im it's a serious but also very simple to fix bug... not that interesting imo
17:56 _Megaf well, I consider a bug if it made a player to have lost everything it had mined for the past months
17:56 _Megaf a bug serious
17:56 _Megaf for example...
17:56 _Megaf mined*
17:56 Krock _Megaf, if so, call "aix" a bug
17:57 _Megaf he is not here
17:57 Krock hmm yes :/
17:58 T4im sadly one cannot place ignore nodes willingly; or i would test for that class of bug automatically :/
17:58 * T4im did consider mocking the situation though
17:58 T4im get_node function that is
18:01 Krock oh right, the most serious bug ever: game#1176
18:01 ShadowBot https://github.com/minetest​/minetest_game/issues/1176 -- tenplus1
18:01 T4im because paramat is quite right about that being probably an issue in a lot more places
18:17 paramat sofar rubenwardy game#1306 simple and fairly urgent
18:17 ShadowBot https://github.com/minetest​/minetest_game/issues/1306 -- Fix sapling placement handling of on_rightclick and CONTENT_IGNORE by t4im
18:18 paramat i might merge that tonight as it's almost trivial
18:18 sfan5 code lgtm
18:19 sfan5 i'd count it as trivial too
18:19 paramat ok
18:20 proller joined #minetest-dev
18:20 Krock why was the check for sneak added in?
18:20 Krock err nvm
18:20 T4im so you can force placing a sapling against a chest or itemframe
18:22 T4im another trivial one, just a simple nil check #1307
18:22 ShadowBot https://github.com/minetest/minetest/issues/1307 -- Add minetest.register_on_punchplayer by Bremaweb
18:22 T4im eh
18:22 T4im game#1307
18:22 ShadowBot https://github.com/minetest​/minetest_game/issues/1307 -- Prevent possible nil-crash on invalid book metadata. by t4im
18:23 _Megaf what do you mean by lgtm?
18:24 Krock "looks good to me"
18:24 Krock which is kind of an approval
18:26 paramat i'll merge both later
18:30 paramat maybe with grey sand too :}
18:39 DI3HARD139 joined #minetest-dev
18:41 sfan5 *sigh*
18:41 sfan5 why don't do mapnodes have a version number
18:41 sfan5 changing MapNode::deSerializeBulk breaks .mts schematics
18:41 sfan5 because the schematic code always uses SER_FMT_VER_HIGHEST_READ
18:42 sfan5 who thought that would be a good idea
18:42 Krock the developers the developers the developers the developers!
18:43 T4im great, now i will never stop imagining you to look like balmer
18:43 T4im :D
18:44 T4im sfan5: what stops you from adding one?
18:44 Krock :D
18:44 sfan5 backward compatibility
18:44 T4im assume everything without to have a version of 0?
18:44 sfan5 how do i know whether the first byte is data or a version?
18:44 sfan5 that's right, i don't
18:45 sfan5 there's other ways of solving this though
18:49 T4im param0 would be the content id? isn't there an invalid to serialize id you can use to make the separation?
18:50 T4im ok probably too hacky, nvm
18:50 sfan5 i don't know
18:50 Void7 joined #minetest-dev
18:50 sfan5 instead i can just fix loading/saving of schematics https://github.com/sfan5/minetest/commit/​10569a3c8b4ff935f03066ef757e940a062684da
18:51 betterthanyou710 joined #minetest-dev
19:39 Taoki joined #minetest-dev
19:42 twoelk left #minetest-dev
20:08 nrzkt sfan5, a little but hacky but works
20:08 sfan5 well i am forced to deal with it
20:09 sfan5 unless we remove support for .mts right now :P
20:09 nrzkt it's not possible to read version and then do sth depending on that version ?
20:39 VargaD joined #minetest-dev
21:07 sfan5 nrzkt: i could do that
21:07 sfan5 but that doesn't make any sense
21:08 sfan5 because no .mts files with brotli exist
21:08 sfan5 so we don't need to support that
21:08 sfan5 i just need to make sure that it reads mts files with a serfmtver before brotli was introduced
21:22 nrzkt sfan5, not now but in the future ?
21:24 rubenwardy joined #minetest-dev
21:32 sfan5 yes in the future this is how it would work
21:40 nrzkt sfan5, TimeGetter is only in clientlauncher.h
21:40 nrzkt time(NULL) is portable, no problem i think then :)
21:40 troller joined #minetest-dev
21:40 sfan5 nevermind then
21:48 nrzkt sfan5, sofar, i added fixes based on your comments on my pr can you review ?
21:49 sfan5 sure
21:50 sfan5 nrzkt: +1 remains
21:50 nrzkt grat
21:50 nrzkt great, thanks
21:53 kaeza joined #minetest-dev
21:58 nrzkt sfan5, for lz4, install lz4 lib include <lz4.h> and here are comprssion/decompression wrappers
21:58 nrzkt http://pastebin.com/k70YVk8V
21:58 nrzkt they are not the most optimized i think, maybe we can do better but it works
21:59 nrzkt i'm using it in my own MT fork since long time then you can bench safely :)
22:02 Fixer joined #minetest-dev
22:10 SloanOnLinux joined #minetest-dev
22:10 SloanOnLinux joined #minetest-dev
22:10 nrzkt rubenwardy, where ?
22:11 rubenwardy commit message / title
22:11 rubenwardy unimportant atm, first thing I saw
22:14 nrzkt i can improve the description if you want
22:14 nrzkt Chat: new settings to prevent spam
22:14 nrzkt :)
22:15 nrzkt sfan5, i will merge it then i have your approval and mine + sofar approved without the last fixes but they are minor
22:17 nrzkt done
22:21 nrzkt sfan5, rebased and squashed with your comment fixed, okay for merge ?
22:24 nrzkt go to bed, if okay you can merge if you want sfan5 :)
22:27 betterthanyou710 joined #minetest-dev
22:29 paramat game#1285 seems good now
22:29 ShadowBot https://github.com/minetest​/minetest_game/issues/1285 -- Destroy flammable items when in fire or lava by Ferk
22:34 Fixer nice, nerzul PR is very useful
22:34 Fixer admins will like it a lot
22:44 _Megaf I don't like this that much
22:44 _Megaf I prefer if anything that is droped on lava gets destroied
22:44 _Megaf but not on fire
22:57 Void7 joined #minetest-dev
23:07 paramat searching for the code that creates damage flash, anyone know where?
23:08 _Megaf dmaage flash?
23:09 _Megaf what do you mean by that?
23:13 paramat red screen
23:25 T4im paramat: game.cpp#L4277
23:26 paramat well done, thanks
23:28 paramat i also found some code earlier in that file, so will experiment
23:29 paramat L3366
23:32 paramat ok i see, 'damage_flash' controls red alpha, and is reduced per-frame
23:41 paramat alpha hits a max of 180, no wonder then
23:43 jin_xi joined #minetest-dev

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