Minetest logo

IRC log for #minetest-dev, 2016-02-26

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

All times shown according to UTC.

Time Nick Message
00:08 paramat joined #minetest-dev
00:09 paramat will merge 3140 3760 3763 in a moment
00:12 sofar paramat: has #1988 been considered for 0.4.14?
00:12 ShadowBot https://github.com/minetest/minetest/issues/1988 -- Fix #1374 by adding a listpredict piece to the formspec by MinerDad7
00:13 sofar est31: paramat: if needed, I can rebase and remove the FORMSPEC_API_VERSION if that is what it's gating
00:18 est31 yeah good idea
00:18 est31 but i want to rather write some code and get my own prs ready
00:18 est31 they need to get in before branch gets freezed
00:18 sofar I'm happy to rebase it.
00:19 sofar I'll see if I can do so tonight, it shouldn't be too hard
00:19 paramat that PR is stuff i don't much understand
00:19 sofar I'll also check middle-click
00:22 est31 paramat, can you do one thing when merging prs
00:22 est31 sometimes you see when you do the "git am" part  "adds whitespace error" or something
00:23 est31 you can remove the whitespace errors quite easily with git rebase upstream/master --whitespace=fix
00:23 est31 keeps us clean from additional whitespace errors from RBA
00:23 paramat ok. vbo has 2 errors
00:24 paramat useful to know
00:24 est31 thanks
00:33 paramat 3760 has a failing build
00:35 paramat 3763 has exact same unit test fail
00:37 sofar something merged earlier breaking it?
00:37 paramat that's what i thought
00:38 sofar ======== Testing module TestThreading
00:38 est31 paramat, #3786
00:38 ShadowBot https://github.com/minetest/minetest/issues/3786 -- Unreliably failing unittests on mac
00:38 est31 sofar, ^
00:39 paramat ok
00:39 sofar yeah threading bugs don't strike me as something engine changes would consistently do
00:39 paramat so i can go ahead and merge these 3 PRs?
00:40 sofar I would
00:43 paramat now merging
00:49 turtleman joined #minetest-dev
00:49 rockyb joined #minetest-dev
00:58 paramat merged :)
01:00 blert joined #minetest-dev
01:53 est31 ahem, interesting
01:53 est31 didnt know that we support locally stored sounds
01:57 paramat looks like 'allfaces optional' only works if all 6 tiles are identical. when fancy leaves is enabled it seems to use the top tile for all 6 faces. this is causing trouble for 'pine needles with snow'
01:58 VanessaE has RBA's VBO patch been merged?  if not, merge it - even VE-Creative spawn performs better now (more than twice as fast if I am right)
01:58 est31 dont have pine needles with snow a nodebox on top?
01:58 est31 VanessaE, yes it has
01:58 VanessaE yay!
01:59 VanessaE (I've been afk all day, didn't look at the commit log)
02:00 Taoki ...
02:00 Taoki Why was automatic FPS - view range adjustment removed?! :(
02:00 est31 because it didnt really work
02:00 Taoki I seriously don't see the reason for such feature removals.
02:00 Taoki est31: It worked perfectly well for me last time.
02:01 est31 also some people claim stuff is "smoother" now whatever it means
02:01 Taoki I see
02:01 est31 i guess its totally phantasy bullshit, and wouldnt survive A/B testing
02:01 est31 but well, it doesnt really work either
02:01 est31 didn't*
02:02 Taoki At least there's VBO now... at last :)
02:02 est31 yeah and its even a very small patch
02:02 est31 really like that
02:02 Taoki Yep
02:04 paramat auto view range was causing some problems, however it was going to be removed by c55's farmap PR anyway
02:04 VanessaE RBA mentioned some shader-related cleanups by gregorycu that need merged, can't recall the title..  something about superfluous variables.
02:05 paramat yeah that's being worked on
02:05 VanessaE ok
02:05 est31 people didnt like the factory pattern he added
02:05 est31 i havent looked at the code so cant say whether its needed
02:05 Taoki paramat: Farmap? Sounds interesting.
02:05 est31 if its not needed its a dumb idea to add it just for the lulz
02:06 paramat auto view range was a lot of code, so stuff is simpler and very slightly faster
02:06 VanessaE I'd like to see farmap merged in as well, though I'm sure it requires some rewriting now.
02:07 Taoki Any screenshots of it?
02:07 paramat in #dev logs somewhere
02:08 est31 very very slightly
02:08 paramat yeah :)
02:08 est31 its one hundred lines without any loops running once every instruction
02:08 est31 ermm
02:08 est31 once every frame*
02:09 est31 so thats perhaps 200 instructions, 60 times per sec
02:09 est31 not much at all on modern cpus
02:09 sofar VanessaE: it was mergable 2 weeks ago (had to liven it up in a few spots, nothing bad), but broke windows somehow
02:09 est31 if you have megaherz cpu, you have a problem
02:10 VanessaE est31: actually I think that was part of the problem - wouldn't the code responsible for measuring the frame rate be unable to take a measurement if it can't run as often as it should have?
02:12 paramat the snowy pines issue https://github.com/minetest/minetest_game/pull/872#issuecomment-189077270 will look into drawtypes
02:12 est31 nah, the code adjusted
02:12 est31 man these pines look ugly
02:12 paramat may have to use opaque tiles for snowy pines
02:13 est31 i really hate what became of mtgame
02:14 est31 but yeah the engine should enable the use case
02:17 kaeza joined #minetest-dev
02:18 turtleman joined #minetest-dev
02:27 paramat for now it could be an opaque texture, since it's snowy, would also help FPS in taiga
02:27 paramat left #minetest-dev
02:29 kaadmy est31: same
02:29 kaadmy seems everyone has suddenly realized MT is heading towards a hacked-together engine
02:29 kaadmy or just voicing it now
03:12 sofar what exactly made or makes it go from an engine to a "hacked together" engine?
03:13 VanessaE well isn't MT officially developed by "a random bunch of lunatics"? :)
04:00 kaeza joined #minetest-dev
04:28 kaeza yey
04:30 est31 ?
04:30 kaeza one sec
04:34 kaeza sooo, you can "steal" the insecure env using `getupvalue` or related functions
04:35 est31 its only accessible under debug.
04:35 est31 and i think the secure env doesnt include it in the whitelist does it?
04:36 kaeza it does, at least here
04:36 kaeza https://gist.github.com/kaeza/ab039940f9ef786d64ca
04:44 est31 oh
04:44 est31 it is in the whitelist indeed
04:47 kaeza I don't think there's any reason to allow `debug`, except for maybe `getinfo`
04:49 est31 here is the whitelist https://github.com/minetest/minetest/blob/master/src/script/cpp_api/s_security.cpp#L108
04:49 est31 remove what you think is not required and make a PR :)
04:52 est31 #3788
04:52 ShadowBot https://github.com/minetest/minetest/issues/3788 -- Bypassable security using upvalues
04:53 ssieb joined #minetest-dev
04:55 kaeza I see many functions there that really should not be whitelisted
04:56 kaeza est31, meanwhile: https://github.com/minetest/minetest/pull/3765#issuecomment-188675453
05:11 est31 seems that assert indeed removes everything
05:11 est31 it not just ignores the return value it also remvoes the stuff that makes us generate it
05:11 est31 well then... usage of assert is wrong here
05:18 est31 will merge that PR as trivial fix
05:38 leat joined #minetest-dev
05:44 est31 somehow he used a leet method to convert an issue into a pull request
05:44 est31 1ee7 very 1337
05:46 leat joined #minetest-dev
06:10 Hunterz joined #minetest-dev
07:09 gregorycu joined #minetest-dev
07:18 DevBox joined #minetest-dev
07:31 kaeza joined #minetest-dev
07:42 Calinou joined #minetest-dev
08:14 nrzkt joined #minetest-dev
08:20 mwamba joined #minetest-dev
09:20 Amaz joined #minetest-dev
09:34 ssieb_ joined #minetest-dev
09:41 kaeza2 joined #minetest-dev
09:45 misprint joined #minetest-dev
10:23 Obani joined #minetest-dev
11:24 proller joined #minetest-dev
11:37 Samson1 joined #minetest-dev
11:52 Fixer joined #minetest-dev
11:56 Taoki joined #minetest-dev
12:13 Taoki Hmmm. Quite weird that even with VBO enabled, I get 30 FPS at 100 node view range. I only seem to achieve 60 FPS at about 70 view range.
12:14 Taoki Considering I just bought a new video card last month (Radeon R7 370), that is a little weird.
12:14 Fixer Taoki, AMD/ATI *cough*
12:14 Fixer Taoki, VBO gives me +20-30% fps boost, for my AMD/ATI HD6870
12:14 Taoki Fixer: Don't get me started... Nvidia isn't all that great and awesome :P
12:15 Taoki Cool. And the 6870 is the card I just switched from, sellint it soon :) It is a very good one though, despite running on the r600 chipset.
12:17 Taoki Well, this is weird: I seem to get no FPS boost from VBO at all o_O
12:19 Player_2 joined #minetest-dev
12:19 Fixer Taoki, wait for superflo shaders commit
12:19 Fixer Taoki, it gives much bigger boost on my ATI
12:19 Taoki Sounds nice :)
12:23 VanessaE I was about to advise turning off shaders entirely.
12:24 Taoki I lose lots of great stuff that way. Especially once bloom and DoF will be in.
12:24 VanessaE on mine, there's really not much boost (+15%?) with shaders on, but off, +80% easily.
12:24 Taoki Shaders do use up a lot of FPS it seems. But I couldn't live without them :)
12:28 Calinou joined #minetest-dev
12:30 Fixer with superfl shadaers you will have nice performance with shaders
12:31 Fixer Taoki, are you on windows or linux?
12:32 Taoki Linux. Also I use the free video driver (mesa)
12:32 Fixer Taoki, ahah
12:32 Fixer Taoki, i think open source drivers has itty support of 370 at the moment, try catalyst crap
12:33 Fixer ok, and it looks like we are on a -dev channel
12:33 Taoki Not planning to instally the proprietary driver. R7 works as well as the R600 otherwise.
12:33 Taoki Ah, right :P
12:33 Fixer Taoki, R7 should be much faster
12:34 Taoki I noticed little improvement in most games, like Xonotic. Though there is one good enough that you can sense it.
12:34 * Taoki hears it will be a lot better once the RadeonSI scheduler will be introduced.
12:34 Taoki That's mainly the reason I got a new video card BTW: To switch to RadeonSI.
13:05 gregorycu joined #minetest-dev
13:05 gregorycu hi
13:07 est31 hello to australia
13:08 gregorycu It's dark over here
13:08 est31 its not dark where I am
13:08 est31 its cloudy though
13:10 gregorycu How is the things? How is all the things?
13:11 est31 going well
13:11 est31 merged some prs
13:11 est31 paramat too
13:11 est31 vbo patch got merged
13:11 gregorycu Such exciting times, best release ever coming up I think
13:11 est31 users like it#
13:37 turtleman joined #minetest-dev
13:41 blaze joined #minetest-dev
13:56 Wuzzy joined #minetest-dev
13:57 Wuzzy Hello! The environment returned by minetest.request_insecure_environment does not contain the minetest table. Is this intentional or a bug?
14:13 STHGOM joined #minetest-dev
14:41 Obani Is #3739 ok ?
14:41 ShadowBot https://github.com/minetest/minetest/issues/3739 -- Add support for Mumble positional audio by ShadowNinja
14:45 kaadmy joined #minetest-dev
15:03 proller joined #minetest-dev
15:30 blaze joined #minetest-dev
15:35 Megaf joined #minetest-dev
16:44 Darcidride joined #minetest-dev
16:46 Hunterz joined #minetest-dev
16:48 Indohmo joined #minetest-dev
17:13 sofar_ joined #minetest-dev
17:14 dulrich_ joined #minetest-dev
17:14 Amoebaa_ joined #minetest-dev
17:14 Ritchie_ joined #minetest-dev
17:14 VirusJon1s joined #minetest-dev
17:14 Ardone1 joined #minetest-dev
17:15 rubenwardy joined #minetest-dev
17:15 Kray_ joined #minetest-dev
17:16 AnotherBrick joined #minetest-dev
17:18 nrzkt2 joined #minetest-dev
17:23 Icedream joined #minetest-dev
17:24 hmmmm joined #minetest-dev
17:48 ud1_ joined #minetest-dev
17:56 Krock joined #minetest-dev
18:09 ElectronLibre joined #minetest-dev
18:20 ShadowNinja Weblate is spamming me with merge failure notices.  It seems someone force-pushed a commit right about when weblate pulled.
18:21 ShadowNinja Weblate may have issues pulling even if I push the rebased translations though.
18:22 paramat joined #minetest-dev
18:22 ShadowNinja And I can't push to weblate's copy of the repo.
18:22 sofar_ can you make it reset after a pull?
18:22 sofar joined #minetest-dev
18:23 ShadowNinja sofar: Well, I have a "reset all local changes" button, which may do the trick.
18:23 sofar that may do the trick
18:24 ShadowNinja But I'll have to push translations to the main repo for this, as I con't write to the weblate copy.
18:29 troller joined #minetest-dev
18:29 ShadowNinja Also, this would be helpfull for making irc_commands usable with SRP: https://github.com/est31/minetest/commit/9afb37b81d4891cd413bb04f79f9daa189c8c2ad
18:29 ShadowNinja The current alternative is having players use /setpassoword to disable SRP completely.
18:30 Gael-de-Sailly joined #minetest-dev
18:31 ShadowNinja As long as you're using TLS it's still pretty secure, and it's definitely better than just disabling SRP for your account.
18:32 ShadowNinja Actually, I could probably implement that commit in Lua, although it would be horribly slow...
18:33 ssieb joined #minetest-dev
18:37 troller joined #minetest-dev
18:46 nrzkt joined #minetest-dev
18:49 Darcidride joined #minetest-dev
19:41 dzho joined #minetest-dev
20:05 Fixer_ joined #minetest-dev
20:06 Gael-de-Sailly joined #minetest-dev
20:22 rubenwardy joined #minetest-dev
21:14 leat joined #minetest-dev
21:21 Indohmo joined #minetest-dev
21:37 troller joined #minetest-dev
22:01 Dragonop joined #minetest-dev
22:19 rubenwardy_ joined #minetest-dev
22:37 paramat #3792
22:37 ShadowBot https://github.com/minetest/minetest/issues/3792 -- Nodedef: Restore smooth lighting to water by paramat
22:44 iqualfragile joined #minetest-dev
23:02 Taoki_1 joined #minetest-dev
23:58 Indohmo joined #minetest-dev

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