Minetest logo

IRC log for #minetest-dev, 2013-12-03

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

All times shown according to UTC.

Time Nick Message
00:00 sapier ok I have to work tomorrow so I need some sleep ;-) good night
00:00 sapier left #minetest-dev
00:08 djdduty joined #minetest-dev
00:46 bas080 joined #minetest-dev
01:14 VanessaE bbl
01:33 EvergreenTree joined #minetest-dev
02:06 NakedFury joined #minetest-dev
02:37 Anchakor_ joined #minetest-dev
02:53 ShadowNinja kahrl: So, you worked on libmtmap. I beleive the usage or C only made it difficult to finish. hmmmm: Does it have to be C, or can it be C++? Perhaps you could help kahrl with it if you really want it to be C.
02:54 ShadowNinja (I just got reminded why my server is still on SQLite...)
02:55 ShadowNinja And 1031 needs checking as soon as one of you comes arround.
02:57 OldCoder joined #minetest-dev
03:02 werwerwer_ joined #minetest-dev
03:08 iqualfragile ShadowNinja: i tested #1031 works
03:09 iqualfragile (and fixes the crash)
03:18 Anchakor_ joined #minetest-dev
03:33 hmmmm hello
03:33 hmmmm kahrl was working on libmtmap?  sweet
03:33 hmmmm I think it would be best to keep it C personally
03:34 hmmmm also I really doubt that kahrl is having trouble with libmtmap
03:34 hmmmm I'm working on more mapgen stuff right now
04:00 hmmmm sapier, good one with the asserts, lol
04:01 hmmmm why wrap (void)(expr) in a block though
04:26 hmmmm oh wow sorry PilzAdam, it looks like I never change the MapgenParams over to the MapgenParamsV7 when the mapgen name is switched
05:02 PavelS joined #minetest-dev
05:33 darkrose joined #minetest-dev
06:50 deltib joined #minetest-dev
07:22 kahrl 1031 looks good
07:24 kahrl what could be done next is get rid of SimpleThread (JThread already does everything that SimpleThread does, and JThread::Wait is better than the sleep_ms loop in SimpleThread::stop
07:24 kahrl )
09:00 VanessaE joined #minetest-dev
09:09 darkrose joined #minetest-dev
09:25 IceCraft joined #minetest-dev
09:27 IceCraft joined #minetest-dev
09:30 Gethiox2 joined #minetest-dev
10:03 proller joined #minetest-dev
10:08 sapier joined #minetest-dev
11:06 sapier joined #minetest-dev
11:07 PavelS joined #minetest-dev
11:38 sapier joined #minetest-dev
12:02 Gethiox3 joined #minetest-dev
12:26 ImQ009 joined #minetest-dev
12:51 Ritchie_ joined #minetest-dev
13:03 naxthesurvivor joined #minetest-dev
13:03 nyuszika7h joined #minetest-dev
13:05 Exio4 joined #minetest-dev
13:22 hmmmm joined #minetest-dev
13:33 BlockMen joined #minetest-dev
13:57 troller joined #minetest-dev
14:05 proller joined #minetest-dev
14:24 zat joined #minetest-dev
14:45 NakedFury joined #minetest-dev
14:48 RealBadAngel joined #minetest-dev
14:49 RealBadAngel hi
14:50 RealBadAngel current sources are broken, anybody experienced instant segfaults?
14:50 Exio4 known bug
14:51 RealBadAngel its not a bug, its a diseaster.
14:51 Exio4 look at #1031
14:51 Exio4 it is the fix
14:51 Exio4 but sapier was waiting for more people to agree so he can merge it
14:52 RealBadAngel such critical bug fix shall not wait
14:52 Exio4 it is a big change in jthread
14:52 Exio4 that is why he was waiting
14:52 proller joined #minetest-dev
14:53 RealBadAngel i wasted two days tryin to rebase my shaders rework, i thought my code was crashing the game
14:53 RealBadAngel something told me to try clean sources and it was still crashing instantly as soon as i started new world
14:54 Exio4 you could read the logs there too, that was said a lot of times
14:57 RealBadAngel what caused that regression in the first place?
14:57 RealBadAngel 0.4.8 was not affected by it
14:57 RealBadAngel something after 0.4.8 is causing game to crash
14:58 RealBadAngel ive seen some weird error with "tables"
14:58 RealBadAngel Stack is over 30:
14:58 RealBadAngel table table 4.02919e+ userdata 16 5.9375 4 4.9375 1 4 1 4.9375 1 1 -48 -16 -32 0 18 userdata 5 18 1 5 -46 -15 4 4 0 -1 4 "default:dirt_with_grass" table table
14:58 RealBadAngel Naruszenie ochrony pamięci (core dumped)
14:58 Exio4 https://github.com/minetest/minetest/commi​t/de0cdbc01cc39e4f89a9d012661031a66ba3294f
14:59 Exio4 err, the one was https://github.com/minetest/minetest/commi​t/04e9a9d5410a151d232a577b46791d2edffba527
14:59 Exio4 the debug build works, though
15:00 RealBadAngel you mean compiles?
15:01 RealBadAngel i cannot play the game anymore
15:03 RealBadAngel after a few seconds game crashes instantly, each try
15:06 RealBadAngel i will try the fix if it works
15:09 Exio4 the debug build works, you can play with it
15:10 Exio4 the problem is that in release-mode those asserts are replaced with "nothing" => no mutex / locking
15:14 RealBadAngel such non tested code should be merged in the first place
15:14 RealBadAngel *shouldnt
15:20 RealBadAngel what other devs said about the situation?
15:23 RealBadAngel the fix works
15:23 RealBadAngel game is no longer crashing
15:32 RealBadAngel since its critical bug fix im about to merge it, its unaceptable that game remains in non working state
15:41 celeron55 eh
15:41 celeron55 master can be briefly broken, it's unavoidable
15:41 celeron55 that's why we have releases
15:41 proller https://github.com/proller/minetest/commit​/c236ed6df3041af3c31b6190093a49ec51028fde - any objectons ?
15:42 celeron55 anyway, looks like kahrl checked 1031 so it can go in
15:43 celeron55 i'll merge it after testing that it works at least a little bit
15:45 celeron55 proller: seems fine (altough having class name "a" is kind of weird)
15:45 proller celeron55, its for emulating underlined <a>
15:46 celeron55 "clickable" or something would make much more sense
15:46 proller ok
15:46 celeron55 i don't care though, the masterserver is just a small piece of code anyway
15:47 lanxu_ what is the problem of href="#"?
15:47 Calinou joined #minetest-dev
15:49 ShadowNinja proller: Um, should unlimited_player_transfer_distance be held in a field called "names"?
15:49 celeron55 #1031 works for me, pushing
15:49 proller lanxu_, its replace location and jump you at top of page
15:49 RealBadAngel celeron55, it works for me also, i was also testing it before i could merge it
15:50 proller ShadowNinja, unlimited_player_transfer_distance is yooo long
15:50 proller too
15:50 ShadowNinja proller: So you use "names"?  That's a deceptive field name.
15:50 PilzAdam joined #minetest-dev
15:50 proller visible names of players
15:50 ShadowNinja proller: I would expect that to contain a player list.
15:53 celeron55 i guess in practice players call that "seeing names of others"
15:53 lanxu_ proller: ok
15:54 ShadowNinja So, "unlimited_name_see_distance" or even "can_see_names".  But I really don't see a issue with a long field name.
15:55 proller its not for peoples, its for future detail server info
16:01 celeron55 the field name is repeatedly sent over network along with the data, so short names make sense
16:04 ShadowNinja Yes, but there's no need to make them short to the point where they are deceptive. I don't think a JSON string is using a whole lot of bandwidth.
16:08 celeron55 i vote for "can_see_names"
16:08 ShadowNinja aaaand you already commited it.
16:08 celeron55 well i didn't commit it; proller did
16:08 ShadowNinja Yes, I knoe.
16:08 proller ш сфт кутфьу
16:08 ShadowNinja know*
16:08 proller i can rename
16:09 celeron55 a new commit won't hurt
16:09 NakedFury will Can_see_names reduce the nametag of players that we see through everything in the game?
16:09 Jordach joined #minetest-dev
16:09 ShadowNinja And that's not accurate, as you can see names on nearby players.
16:09 celeron55 bah
16:09 celeron55 can_see_far_names
16:09 proller can_see_names_at_long_distance
16:09 proller i win
16:10 celeron55 you don't; it's longer than is needed
16:10 kaeza show_player_names_when_they_are_closer_than_​the_value_specified_in_block_send_distance?
16:10 proller kaeza win
16:10 celeron55 kaeza: +do_not_
16:10 kaeza pretty accurate I think :P
16:10 kaeza ah k :P
16:10 proller can_see_far_names
16:10 celeron55 but seriously, "can_see_far_names"
16:11 proller also i will add flag to http://servers.minetest.net/
16:20 zat joined #minetest-dev
16:26 Gethiox4 joined #minetest-dev
16:27 zat1 joined #minetest-dev
16:43 RealBadAngel PilzAdam, i would like to push to minetest_game "waving = 1" changes for plants and leaves anything against it?
16:43 PilzAdam yes, dont do it until shaders are in
16:43 RealBadAngel i would also like to push shaders today
16:43 PilzAdam they are broken AFAIK
16:44 RealBadAngel ive rebased them
16:44 PilzAdam VanessaE and thexyz reported issues with alpha
16:44 RealBadAngel no, they are not broken anymore
16:44 RealBadAngel ive fixed that already
16:44 RealBadAngel that was simple error
16:44 Jordach where's that commit then :P
16:45 RealBadAngel i was stopped for a while thx to sapier's bug
16:45 RealBadAngel i thought that my code is wrong
16:45 RealBadAngel theres no commit yet, but it is on my git
16:46 celeron55 VanessaE, thexyz: report new test results of shaders here so that everyone knows about them
16:48 RealBadAngel i would like to get that merged at least. since last few days other devs commits altered 5 files already
16:49 RealBadAngel constantly rebasing  such huge commit is simply pain in the ass
16:50 sapier Rba I rebased your code for you to continue. You just should read logs with your name ;-)
16:51 RealBadAngel i was off for 3 days, had night shifts at work
16:52 RealBadAngel but no problem, i made some cleanup also
16:52 PilzAdam you dont have to constantly rebase
16:52 PilzAdam its enough if you do it before a merge
16:52 sapier Ok so I guess my rebased version of your old commit wont help that much
16:53 RealBadAngel i had to, many files not touched for quite long had changed
16:55 RealBadAngel im double checkin the commit now, will pull it in a few minutes (want to play with it for a while)
16:57 sapier Did you fix the defaultsettings regression?
16:58 RealBadAngel https://github.com/RealBadAngel/minetest/com​mit/2bfbd5ff400cd3427b9d7a295c2de33b36a8ab43
16:58 RealBadAngel the patch is here
16:58 RealBadAngel sapier, what regression?
17:00 sapier The remote_port setting was lost in your last version
17:01 RealBadAngel thats rebase thingy, done
17:01 RealBadAngel ive rebased it today, so its surely up to date with everything
17:02 Calinou joined #minetest-dev
17:06 OldCoder joined #minetest-dev
17:07 PilzAdam https://github.com/minetest/minetest/pull/1034
17:07 thexyz celeron55: well i saw everyone talking about broken shaders and tested them and yes, they were broken, so i thought there's no real need in reporting it because everyone was talking about how broken they were
17:08 thexyz but next time I of course will report it here, sorry for not doing that the last time
17:09 RealBadAngel thexyz, can you test them now?
17:09 thexyz RealBadAngel: of course, gimme link
17:10 RealBadAngel https://github.com/RealBadAngel/minetest/commit​/2bfbd5ff400cd3427b9d7a295c2de33b36a8ab43.patch
17:10 thexyz ok, I see, it segfaults right as I start the game
17:10 RealBadAngel update your game
17:10 thexyz or shall I apply it myself to the current master?
17:11 thexyz I just cloned your repo
17:11 RealBadAngel it doesnt contain the sapier's fix
17:11 thexyz alright, will do in a moment
17:11 RealBadAngel it was added a few minutes after i rebased it
17:11 RealBadAngel apply the patch to current minetest repo
17:18 RealBadAngel https://github.com/RealBadAngel/minetest_game
17:18 RealBadAngel modified game with added waving =1 to plants and leaves
17:20 thexyz meanwhile https://gist.github.com/xyzz/126cf7f​613734c42ebb9/raw/7d0a4fe92982a1636e​42e1f7964b5631f501ab16/gistfile1.txt
17:21 PilzAdam RealBadAngel, have you fixed the flowers waving?
17:21 naxthesurvivor joined #minetest-dev
17:21 PilzAdam RealBadAngel, also its "waving = 1," not "waving =1,"
17:22 PilzAdam and what about other values for faster waving for e.g. fire?
17:22 RealBadAngel flowers are not waving atm
17:22 thexyz yeah, well, they're waving so everything's cool I guess
17:22 thexyz anything else I should test?
17:22 RealBadAngel you can test bumpmapping and parallax
17:23 thexyz how? what should I look for?
17:23 RealBadAngel http://realbadangel.pl/minet​est/HavenNG/HavenNG_256x.zip
17:23 ShadowNinja PilzAdam: Your patch looks good.
17:23 RealBadAngel use this texture pack
17:24 RealBadAngel with bumpmapping and parallax i suggest turning on filtering
17:25 RealBadAngel effect is way smoother
17:27 RealBadAngel PilzAdam, i will work on new features once base will be there
17:27 thexyz yes, seems to work too
17:28 RealBadAngel i do have several ideas to implement
17:28 thexyz how about wielded torch light? I remember here being a pull request
17:29 RealBadAngel it was wrong at one thing
17:29 RealBadAngel effect was only visible for client
17:29 thexyz yes but that's better than nothing
17:29 RealBadAngel so other players couldnt see the player holding the light
17:30 PilzAdam ShadowNinja, maybe its better do get another core dev agree on that pull request? it may be controversial
17:31 RealBadAngel PilzAdam, i will fix the missing space shortly
17:31 thexyz what's controversial here?
17:31 PilzAdam thexyz, https://github.com/minetest/minetest/pull/1034
17:31 RealBadAngel PilzAdam, have you tested the shaders too?
17:31 ShadowNinja PilzAdam: I don't think it's contreversal, but you've got two devs right here...
17:31 thexyz no, I mean, what was done is terrible and if that patch fixes it then great, just let it in
17:32 thexyz RealBadAngel: so do you have plans to reimplement wielded light?
17:32 PilzAdam ok, Ill merge it then
17:33 RealBadAngel thexyz, yes i do, i will add better version of it in the future
17:33 thexyz any time frames?
17:34 RealBadAngel once i have my rework in place i can work on new features
17:34 RealBadAngel atm most of the time i spend on rebasing and bugs that come from rebasing
17:34 thexyz I see. This probably means that this development model is bad
17:35 thexyz I'm not sure who decides that but I'd like them to rethink it
17:35 RealBadAngel before i rebased only defaultsettings and mainmenu.lua
17:35 thexyz if this makes harder for people to contribute then something have to be done
17:36 RealBadAngel now i had to rebase /scripts/common/c_content.cpp,  nodedef.cpp also
17:37 RealBadAngel and minetestconf.example :)
17:37 RealBadAngel thexyz, so as for you, everything works fine?
17:38 thexyz yes
17:38 thexyz it waves and it seems there are no issues with alpha channel anymore
17:39 RealBadAngel ok
17:39 RealBadAngel PilzAdam, and you?
17:42 ShadowNinja RealBadAngel: You removed the direct3d check. (lthough it should probably be a opengl whitelist check)
17:43 RealBadAngel ShadowNinja, lemme check that
17:44 ShadowNinja RealBadAngel: You also moved the port setting and removed the remote_port setting from minetest.conf.exmple
17:44 RealBadAngel https://github.com/RealBadAngel/minetest/com​mit/2bfbd5ff400cd3427b9d7a295c2de33b36a8ab43
17:44 RealBadAngel no i havent
17:44 RealBadAngel all is in place
17:44 ShadowNinja RealBadAngel: Update your pull request then.
17:44 RealBadAngel forget the pull requests
17:45 RealBadAngel theyre outdated thx to rebasing constantly
17:45 smoke_fumus joined #minetest-dev
17:46 proller maybe renmae game to "Rebasetest" ?
17:46 RealBadAngel after releasing 0.4.8 several commits broke my pulls
17:46 RealBadAngel one was especially annoyin. ONE character change in mainmenu.lua :)
17:47 RealBadAngel that was spelling of Nore's real name
17:47 ShadowNinja RealBadAngel: Alright, the commit you linked looks good.
17:48 rubenwardy joined #minetest-dev
17:48 RealBadAngel so, 2 core devs have nothing against it by now
17:49 PilzAdam I wonder why RealBadAngel and proller have so much troubles with rebasing
17:49 proller rebase - useless operation
17:49 Exio4 why?
17:49 RealBadAngel because i dont like github and using different way of maintaing my sources
17:49 proller and sometimes harmful
17:50 RealBadAngel im really old school coder and propably wont change that ;)
17:50 RealBadAngel PilzAdam, so, do you agree on merging my pull?
17:54 PilzAdam https://github.com/minetest/minetest/pull/1035
17:54 PilzAdam RealBadAngel, Ill test it gimme a second
17:56 sapier joined #minetest-dev
17:56 thexyz Exio4: http://irc.minetest.ru/min​etest/2013-12-02#i_3468460
17:57 RealBadAngel PilzAdam, sure
18:01 PilzAdam RealBadAngel, Id say its in state where its reasonable to merge it
18:01 PilzAdam but you still have to work on it
18:03 RealBadAngel ofc, with plenty of new features
18:04 PilzAdam water waving should become part of the nodedef too
18:04 RealBadAngel water needs total rework
18:04 RealBadAngel both in the engine and in game
18:04 PilzAdam do you want to push it, or should I do it? since I already have it rebased now
18:04 RealBadAngel it should be surface + other tiles
18:05 RealBadAngel i have pull ready, waited for votes
18:08 RealBadAngel but ok, you can push it. i have not merged last commits so it shows me some weird warnings
18:09 john_minetest joined #minetest-dev
18:11 Zeitgeist_ joined #minetest-dev
18:11 Zeitgeist_ joined #minetest-dev
18:12 john_minetest PilzAdam: Is sneak priv for sneaking at all or for glitch-sneaking?
18:12 PilzAdam john_minetest, sneaking at all
18:14 john_minetest So now there is a priv needed so people don't fall off their house's roof when doing works there?
18:15 PilzAdam RealBadAngel, done
18:15 PilzAdam lets wait for the shitstorm ;-)
18:16 RealBadAngel lol
18:16 john_minetest I am not sure if it is good to make sneak a priv. When not given by default it will make building a lot more difficult.
18:16 RealBadAngel ive updated minetest_game
18:17 RealBadAngel bot reported both at once :)
18:17 thexyz yes, i agree, there's totally no need for a sneak privilege
18:17 thexyz there's a need for a "sneak_glitch" privilege but I guess it's a bit harder to implement
18:18 thexyz or maybe
18:18 thexyz just disallow jumping when sneaking?
18:18 thexyz and call it sneak_glitch
18:19 john_minetest sneak priv makes games on servers partially impossible.
18:19 thexyz yes, I agree
18:19 thexyz PilzAdam: ^
18:19 john_minetest jumping games can sometimes only be won when going to the edge and a bit over while sneaking and then jump to the next platform
18:20 john_minetest And I'd like to have that feature without the need to allow sneak glitching at the same time.
18:23 john_minetest Also sneaking is an integral part of the building process. I can't imagine any situation where sneaking wouldn't be useful. It is all about fixing a bug that is not critical and that should be fixed other wise, without destroying the gameplay.
18:23 PilzAdam john_minetest, "When not given by default" it is given by default in all games; just mine will revoke it
18:23 PilzAdam the point is that not all games like sneaking, since as you said: it makes building easier
18:24 john_minetest In real life you can also lean over the edge a bit to build a cantilever roof or something like that.
18:24 PilzAdam my games doesnt want to simulate real life
18:24 john_minetest You would have to stand on a lower layer (e.g. the ground) to build anything. Like bridges.
18:24 thexyz please stop using "in real life" argument
18:25 RealBadAngel PilzAdam,thexyz, sfan5, so i guess its time to make new windows builds
18:25 RealBadAngel :)
18:25 PilzAdam john_minetest, no, its actually possible to build a bride while standing on the same layer, you just have to be more careful
18:26 john_minetest okay, the "IRL" argument might be invalid in this case. Still building bridges would be nearly impossible. It would take so much care not to fall down :/ Not sure if this is optimal.
18:26 thexyz if it's given by default to players then I guess it's fine although it shouldn't be advertised as a fix to the "sneak glitch" problem
18:26 PilzAdam thexyz, yes, you are right
18:26 thexyz this'll make more hardcore servers possible so why not do that
18:26 john_minetest Exactly. That was my point.
18:27 thexyz then stop arguing and test/merge it
18:27 PilzAdam if somebody wants to test it: minetest.registered_privileges["sneak"] = nil -- this disables sneaking for all players
18:30 sapier left #minetest-dev
18:32 sapier joined #minetest-dev
18:41 john_minetest Question: Would more grass in the world decrease the performance a lot?
18:44 sapier performance predictions are very difficult for our engine ... there's no pressing reason why more grass should degrade performance very much (unless more is meant as factor 100+)
18:45 sapier but if a grass node was 10 faces for example this would have a lot of impact
18:45 john_minetest a grass node with only 2 faces.
18:46 sfan5 RealBadAngel: made a new build
18:46 john_minetest I think it would be cool to have more grass, now that it is waving in the wind.
18:46 sapier waving grass?
18:47 sapier most likely that'd be way more then 10 faces per grass then ;-)
18:47 john_minetest waving shaded grass.
18:47 john_minetest rba is in master
18:48 sapier ok so pseudo waving
18:48 john_minetest yeah.
18:55 Calinou plantlike nodes look bad since they use less faces now
18:55 Calinou they are symmetrical from 2 points of view...
18:56 sfan5 who broke the leveldb backend?
18:56 sfan5 a non-existing block counts as blank block
18:56 sfan5 creating new worlds with leveldb no longer works
18:57 john_minetest Calinou: When they wave they just look so awesome. And... http://www.souleaterwallpaper.com/images/wallpap​ers/Soul-Eater-Death-TheKid-Symmetry-308638.jpeg
18:57 sapier guess that's the issue with too many different supported things doing same ... you can't test everything prior commit
18:58 PilzAdam sfan5, hmmmm changed something about this
18:58 sfan5 wow, how intelligent
18:58 RealBadAngel sfan5, super. i will later add to it texture packs. maybe you could add a link for it on your page
18:58 sfan5 first checking the length of the datastr and then checking where the returned leveldb::status is ok
18:59 sfan5 whether*
18:59 sfan5 RealBadAngel: "on your page" where exactly?
18:59 RealBadAngel with the links to your builds
19:00 sfan5 >where
19:00 RealBadAngel in the forums?
19:01 RealBadAngel i would like to prepare rdy to play package for windows with all the eye candies
19:01 sfan5 I already posted it in my builds topic
19:01 Jordach ...0.4.9 anyone :)
19:01 sfan5 anyway, someone should do a News post about shaders
19:01 sfan5 leveldb is completly broken in 0.4.8, this annoys me
19:02 thexyz sapier: that's what tests are for
19:03 RealBadAngel i have to write some guide on how to prepare proper normal maps
19:03 RealBadAngel gosh, i have lotsa to do now
19:04 celeron55 thexyz: there is an implementation for a sneak glitch privilege made by me in the sneak glitch issue (or was it a pull request, dunno)
19:04 john_minetest I think I will try to do that one: https://github.com/minetest/minetest/issues/1033
19:05 thexyz celeron55: I see; it got ignored for some reason
19:05 RealBadAngel btw, i was thinkin on one idea. since Bumped texture pack doesnt change any default textures and provides "render like" feel
19:05 sfan5 may I push https://gist.github.com/sfan5/aae5ee59bef7bf4ff23c ?
19:05 celeron55 thexyz: well it was a suggestion, i'm fine with people not caring as long as they don't come up with inferior solutions later
19:05 RealBadAngel couldnt it be made default with game?
19:05 thexyz celeron55: well, that's what happened, people didn't come up with inferior solutions
19:06 sapier yes thexyz but did you ever think about what we need to test for minetest? if each separate commit would have to be testet for everything you'd need 2 days to only do the tests for it ;-)
19:06 thexyz I guess both PilzAdam's and celeron55's fixes should be in
19:07 sfan5 sapier: well... the leveldb issue would've been noticed if you open a leveldb world and are near ungenerated chunks
19:07 celeron55 PilzAdam: if you add a sneak glitch (or disable sneak glitch) privilege, it should result in sneaking being enabled in by default in all possible current situations which don't explicitly set it to not be allowed
19:07 celeron55 -in
19:07 sapier yes as the openbsd issues occur if you build in openbsd as sql issues occur if you open with sql ....
19:08 celeron55 sfan5: if it works, yes
19:08 sfan5 it does
19:08 sfan5 ok
19:08 john_minetest sneak_glitch should be set per server, not per person
19:08 celeron55 this is my sneak_glitch privilege: https://github.com/minetest/minete​st/pull/760#issuecomment-20988912
19:10 celeron55 john_minetest: probably, but due to KISS principles it can be better to set it for each player instead of adding a system for overlaying abilities for a whole server
19:10 celeron55 because we need per-player stuff in any case
19:12 john_minetest okay. I didn't know it was player specific.
19:12 thexyz sapier: I'm not saying we should test everything
19:12 thexyz 100% code coverage is inane
19:13 john_minetest Most as possible should be tested.
19:13 john_minetest Else we will have crashes like the recent one again.
19:14 john_minetest RealBadAngel: How big are the bumpmaps for default texture pack?
19:14 sapier 100% coverage is (for non trivial cases) impossible
19:15 Warr1024 joined #minetest-dev
19:15 celeron55 the basic functionality of everything can be tested beforehand; other than that, bugs just have to be fixed fast when they happen to be found
19:15 celeron55 and given long enough feature freezes, that will result in stable releases
19:16 celeron55 that's how everything has worked before and that's how everything will work in the future
19:16 sapier celeron55 that's not true you can't test (linux win32msvc10 msvc11 mingw)exp(lua/luajit/sql/leveldb​/curl/noncurl)exp(release/debug)
19:17 sapier at least you can't for each single commit
19:17 celeron55 basic functionality means the most common combination
19:17 celeron55 and of course code must be made so that it's most likely to work for every platform if it works for one platform
19:18 celeron55 that's why we have jthread and stuff; to keep the dirty work away from the productive code
19:18 sapier :-) actually jthread now is our own implementation ;-)
19:19 celeron55 it doesn't matter whose it is
19:19 celeron55 it matters that it's in it's own files that don't interfere with minetest itself
19:19 sapier yes once it's done it's less likely to break again
19:19 ShadowNinja PilzAdam: I'm against the sneak priv as existing players won't have their privset updated. (If I understand Minetest's priv management)
19:19 celeron55 same must be reached with the database backends
19:20 celeron55 ShadowNinja: yeah, for that reason it should probably be a privilege that disables sneaking 8)
19:21 celeron55 or alternatively it should be added to the "player physics override" stuff
19:22 ShadowNinja Well my bind_address patch is fixed... (It tried to bind to <server address>:0 when connecting to other servers)
19:26 RealBadAngel john_minetest, by now theyre 256x, but i think i can make them way smaller
19:26 john_minetest shouldn't 16x16 be possible?
19:27 RealBadAngel definitely not
19:27 RealBadAngel effect needs way more texels to look good
19:28 RealBadAngel i was testing some textures with 64x bumpmaps, that was absolute minimum
19:29 john_minetest I don't want to download textures plus double the size in bumpmaps when I am not using them
19:30 celeron55 that's a valid point; there needs to be a way for a client to not download them if it doesn't use them
19:31 celeron55 it's just immensely wasteful otherwise and everyone will cry for reasons that they don't understand!
19:34 john_minetest minetest_game TP should be the most minimal TP that can be used.
19:34 john_minetest When people want bumpmaps they can download it from the (future) TP browser.
19:35 john_minetest Actually "Bumpmaps" could be a filter option there.
19:44 sapier jmutext isn't exactly clean on current version too ... if a lua error within a worker thread occurs deletion of mutex will cause a follow up error
19:44 sapier deletion of locked mutex isn't a good idea
19:53 NakedFury joined #minetest-dev
19:55 VanessaE *poking in for a sec while I read the backlog*  the current state of RBA's shaders is good for me now.  Worlks with current master.  lava, water flowing lighting is good, colors are good.  Light sources look right -- with one caveat:  they are not well-lit in-hand (they could be brighter).  Leaves and plants wave around fine as does water.  No problems with transparency.  No problems with inventory images.
19:55 VanessaE *goes back to reading backlog*
19:56 thexyz VanessaE: it's merged already
19:58 VanessaE RealBadAngel: a suggestion: use shaders to make the water and lava flow rather than relying on the animation.  Take the first frame of the "source" and make it move around?
19:58 VanessaE (for the _flowing nodes I mean)
19:59 sapier did anyone ever try the dropdowns?
20:00 VanessaE thexyz: yep I just now saw.
20:00 VanessaE but, c55 asked for my opinion.  there it is :)
20:00 VanessaE but that does reveal a glitch:  light-emitting objects used to be brighter in-hand than they are now.  This probably should be fixed.
20:03 celeron55 ehm... the current version seems to segfault reliably when i go fly underground in a certain world
20:04 sapier if it's reproducable debugging might be easy
20:07 celeron55 well this is the slow laptop of mine, so you're probably going to get used to -O1 -g backtraces
20:07 Calinou joined #minetest-dev
20:08 PilzAdam ehm, set_phyiscs_override() seems to be broken
20:08 PilzAdam player:set_physisc_override(2, nil, nil) shouldnt change jump or gravitiy, but it sets them to 0
20:08 celeron55 except that now i moved in the world and it doesn't segfault anymore
20:09 PilzAdam and I know it did work some time ago
20:09 celeron55 and the only core dump i have actually is a release build... ok, i give up
20:10 john_minetest It is always like that. Release crashes, debug runs fine.
20:10 john_minetest Stuff ran in GDB also seems to crash less likely than stuff running normally :D
20:11 celeron55 anyway here's for the logs: http://pastebin.ubuntu.com/6516305/
20:11 Calinou "debug". it removes bugs. :)
20:12 john_minetest Exactly. And "release" releases the bugs onto the program
20:12 celeron55 that could be a mapgen v6 bug, relaly
20:12 celeron55 really*
20:13 VanessaE bbl
20:13 celeron55 as nothing else is happening other than it running
20:14 celeron55 i'm somewhere around 62,18,152 in there and seed = 1142874489878115732
20:24 ShadowNinja PilzAdam: That code looks correct.
20:32 sapier https://github.com/minetest/minetest/pull/1030 updated search toolbar now includes selection of specific version ... same for unsorted tab
20:36 ShadowNinja sapier: Can you make the text black for better contrast?
20:37 PilzAdam seems like its an issue of set_override(2) vs. set_override(2, nil, nil)
20:37 sapier nope labels don't support colors
20:37 sapier already tried :-)
20:38 ShadowNinja sapier: Then make them support colors.
20:38 john_minetest Colored text would be awesome, too
20:38 sapier https://github.com/minetest/minetest/pull/1027 rebased for current master
20:40 sapier I don't think it's that bad to be worth the work if you feel it's bad you can add it too ;-)
20:42 ShadowNinja But I'm trying to fix something else right now... And I don't know much about the Irrlicht GUI system.
20:43 sapier then it's most likely not important enough to be fixed
20:44 e1z0 joined #minetest-dev
20:49 e1z0 joined #minetest-dev
20:51 PilzAdam celeron55, https://github.com/minetest/minetest/pull/1035 updated
20:51 PilzAdam it uses set_physics_override() now
20:53 celeron55 "for legacy reasons" what?
20:53 PilzAdam writeU8() reads false if the server sends nothing
20:54 celeron55 well how about you write it there
20:55 PilzAdam "// these are send inverted so we get true when the server sends nothing" better?
20:55 celeron55 i'm not really comfortable with the array nature of set_physics_override(), but i guess it's fine now and we'll replace it with something similar to set_object_properties or whatever it is
20:56 celeron55 PilzAdam: yes (s/send/sent/)
20:57 PilzAdam same comment on server side?
20:58 celeron55 doesn't hurt
20:59 PilzAdam is it ok then?
21:00 sapier kahrl could you check 1027 if it's what you intended it to be?
21:01 celeron55 PilzAdam: if it works 8)
21:01 celeron55 this is bound to create problems with servers that want to disable sneak glitch but use a mod that does something else with set_physics_override though
21:02 PilzAdam it doesnt change anything if 4th and 5th parameter are nil/not set
21:02 Exio4 what about a table parameter? .gravity, .sneak_glitch, .sneak, etc?
21:02 celeron55 Exio4: that's what i mean by "similar to set_object_properties"
21:03 sapier1 joined #minetest-dev
21:03 Exio4 ah
21:03 celeron55 except that that's not the name of it... it's called set_properties
21:03 PilzAdam so the mod that disables sneak_glitch would call set_physics_override(nil, nil, nil, nil, false), and other set_physics_override(x, y, z) wont change it
21:03 ecube joined #minetest-dev
21:04 ShadowNinja PilzAdam: The sneak glitch should be disabled by default.
21:04 PilzAdam Exio4, a table is obviously better, but thats something todo in another commit
21:04 PilzAdam ShadowNinja, no, the community has decided against that
21:05 celeron55 PilzAdam: hmm, well that works... until one more parameter is added to it and then a mod calls nil, .... nil, <something for the last one>; it's just not going to work in the long run :P
21:05 ShadowNinja PilzAdam: Then what's the point of fixing it? Nobody is going to manually re-enable it.
21:05 Gethiox joined #minetest-dev
21:06 PilzAdam people dont want it to be fixed for mt_game, but I dont want it in Nodetopia (pilztest)
21:07 PilzAdam maybe we can add a setting in mintest.conf for mt_game or so
21:11 mrtux joined #minetest-dev
21:12 john_minetest So the glitch is now disabled by default in the new versions?
21:13 john_minetest Independant from the game it should be disabled because that bug might confuse new players. On servers it could be enabled because fast ladders can be found there.
21:13 ShadowNinja Why do we include sqlite3? I think I heard that it was for windows, if so shouldn't it be provided as a seperate download?
21:27 PilzAdam celeron55, https://github.com/minetest/minetest/pull/1035 happy now?
21:28 PilzAdam ShadowNinja, we can change the default value later maybe
21:31 ShadowNinja PilzAdam: The non-table sneak settings were never added, so there's no need to support them.
21:31 PilzAdam right
21:32 ShadowNinja PilzAdam: Otherwise it looks good, although I still think sneak_gritch should default to false.
21:33 ShadowNinja And the setting should be conpletely removed in a year or two.
21:33 PilzAdam now why doesnt it play footstep sounds with sneak=false?
21:37 BlockMen i agree that sneak_glitch should be false by default
21:38 proller joined #minetest-dev
21:39 PilzAdam has anybody an idea why it doesnt play footstep sounds when sneak is false?
21:39 BlockMen PilzAdam, currently or with your pull?
21:39 PilzAdam with my pull request
21:40 celeron55 PilzAdam: it should still check for nil just like before for the old fields
21:40 celeron55 otherwise old mods that rely on it break horribly
21:41 PilzAdam hm?
21:41 celeron55 ah, you used _default
21:42 celeron55 ah lol wtf, i don't even know what i thought i  saw
21:42 celeron55 it looks pretty good (won't test it though)
21:47 NakedFury joined #minetest-dev
21:48 Miner_48er joined #minetest-dev
21:56 PilzAdam the problem seems to be getStandingNodePos()
21:58 PilzAdam it only works if m_sneak_node_exists is true
22:00 PilzAdam changing it to return floatToInt(getPosition() - v3f(0, BS, 0), BS); works
22:02 PilzAdam who wrote this? :-p
22:04 PilzAdam updated the pull request
22:06 PilzAdam can I merge it now?
22:15 PilzAdam the change in getStandingNodePos() doesnt break anything, so Ill merge just merge it now
22:22 john_minetest How is minetest_game managed? The mod farming is in it, for example. Does that mean that there is always the newest version of farming in it or just some random stable version?
22:22 john_minetest Who decides what comes in it?
22:22 PilzAdam there is no other farming mod
22:27 PilzAdam john_minetest, we currently have the system to not change mt_game
22:27 PilzAdam we leave it up to the community to write mods and/or (even better) games
22:28 RealBadAngel btw, i can see some pulls now about "tables"
22:28 RealBadAngel i got today one error message with them
22:29 RealBadAngel stack is over 30. table table 4.02919e+ userdata 16 5.9375 4 4.9375 1 4 1 4.9375 1 1 -48 -16 -32 0 18 userdata 5 18 1 5 -46 -15 4 4 0 -1 4 "default:dirt_with_grass" table table
22:29 RealBadAngel and segfault
22:30 sapier someone didn't cleanup lua stack correctly
22:30 sapier or a endless loop
22:30 sapier https://github.com/minetest/minetest/pull/1036 removed SimpleThread ... JThread now provides same features in a better way
22:37 john_minetest sapier: Your mmdb integration work is awesome.
22:38 john_minetest RealBadAngel: Your shaders are awesome.
22:38 sapier yes but iqualfragile should add versions to downloadable files ... using date isn't best to do
22:45 john_minetest yeah, true.
22:46 sapier and he should rename the rating :-) guess that one has a strange name in json
22:46 john_minetest What is the highest rated mod atm?
22:46 john_minetest Still "test"?
22:47 sapier think so but ingame its 0 :-) for not present
22:47 john_minetest https://forum.minetest.net/mmdb/mod/test/
22:47 john_minetest there is a download for it
22:47 sapier still 0 in game
22:48 sapier field "rating" doesn't exist in json ... no idea how he called it
22:49 RealBadAngel john_minetest, ty
22:50 hmmmm sfan5, my apologies
22:50 hmmmm actually that change was going to be for sqlite3 only but somebody suggested that I do it for both
22:50 john_minetest <tr class="mod_rating_positive">
22:50 Exio4 json isn't html
22:50 hmmmm look, this isn't too bad
22:50 zat joined #minetest-dev
22:50 sfan5 it's "sad" that leveldb is broken in 0.4.8 but we can't change that
22:51 hmmmm we have a really good motivator for getting 0.4.9 out of the door really soon now
22:51 john_minetest I want a christmas release!
22:51 Exio4 0.4.{6,8}
22:51 Exio4 two broken releases, ;P
22:51 hmmmm well
22:51 hmmmm we're doing what Linux does
22:51 sapier atm 0.4.9 target date is christmas keep on working guys it's only httpfetch that is absolutely required for it
22:52 hmmmm except instead odd releases are stable and even releases are broken
22:53 john_minetest How difficult would it be to get TP integration when mod integration is done? Any idea, sapier?
22:53 hmmmm toilet paper integration?
22:54 RealBadAngel rotfl
22:54 john_minetest texture pack. But it would be interesting to have a texture pack that is handdrawn on toilet paper...
22:54 john_minetest I think I will do that.
22:54 Exio4 i hope you won't make all the textures brown-colored
22:54 sapier modstore?
22:55 sapier not for 0.4.9 ... sorry that's a major redesign prone to add new errors
22:55 john_minetest sapier: A modstore for textures in the texture tab.
22:55 hmmmm by the way
22:55 john_minetest Of course It doesn't have to be now.
22:55 hmmmm there's still a week of feature freeze so that means we're done adding bullshit on the 18th
22:56 hmmmm this does NOT mean "oh crap i've gotta rush my feature in
22:56 Exio4 the only feature that is a yes-yes to get merged now is the httpfetch code
22:56 hmmmm hold your fucking horses, why don't you wait until it works and is stable.  we're doing monthly-ish releases from now on so there's no reason why anybody should want to shove something half-assed in
22:56 sapier requires a lot of changes in  basic menu design ... nothing completely new of course but moving around things always breaks some things
22:56 john_minetest "It is the 18th, we got 1000 new, unstable features. Wait for 2 years until we fixed the bugs"
22:57 sapier I'd suggest the search tab too as it fixes a visible shortcoming
22:57 BlockMen left #minetest-dev
22:58 john_minetest yes, the seachtab is nessecary.
22:59 john_minetest The tab is even implemented in 0.4.8 but it doesn'T work :D
22:59 sapier no it's not implemented :-) it's a dummy only
23:00 sapier I was hoping someone would take the offer and write it ... obviously not
23:00 john_minetest yeah. Quite annoying when starting 0.4.8 for the first time and - OMG! Search tab!... ummm....
23:01 sapier btw search tab fixes a crash in dropdown too ;-)
23:02 john_minetest now we can drop the bass down.
23:04 john_minetest I wonder how to do the mapgen menu best.
23:04 john_minetest It would be cool to have an infobox there that shows information about the mapgens.
23:05 sapier add it ;-)
23:05 john_minetest oh, and: What exactly is v7? Seems there are biomes missing. v7 is ingame, so terrains should be in minetest_game, too-
23:05 john_minetest sapier: adding it is my intention.
23:06 hmmmm i would really like if people didn't use v7
23:06 hmmmm and stop making biomes for it when i'm going to change all of that stuff soon
23:06 john_minetest Why is it in the game if it just doesn't work?
23:06 sapier for testing purposes
23:06 hmmmm but it does work
23:06 john_minetest "v7" sounds "this is the new, stable mapgen that you should use" to me.
23:07 hmmmm i originally left it commented out when it came to pushing but i forgot to comment it back out one time
23:07 john_minetest "v6" sounds to me like "Don't use that one because there is something new"
23:07 sapier maybe you should call it indev
23:07 hmmmm and i figured, "ah people will be alright"
23:07 hmmmm sapier, that name was already taken
23:07 sapier :-/
23:07 john_minetest And "indev" sounds to me like "Woah, leave it alone, it will burn and crash"
23:07 sapier v-1
23:07 hmmmm v7 won't burn and crash
23:07 iqualfragile joined #minetest-dev
23:07 hmmmm it's just prone to changing
23:08 sapier that's what I call indev
23:08 john_minetest Then it should be called indev
23:08 hmmmm v6 has the guarantee that the terrain that it has generated since it stabilized in 0.4.3 will always be exactly the same
23:08 hmmmm indev is v6, that has experimental modifications done to it
23:08 hmmmm it's the v6 that's in development
23:08 john_minetest Then call it "v6_indev"
23:09 Exio4 v6-dev, v7-dev
23:09 john_minetest or v6-experimental
23:09 hmmmm yeah, well guess what
23:09 Exio4 add -dev prefixes everywhere
23:09 hmmmm it's already said and done
23:09 hmmmm you can NOT change the name now
23:09 Exio4 yeah
23:09 sapier call it v6+abadoned
23:09 hmmmm if you do, things become broken
23:09 Exio4 lol sapier
23:09 hmmmm believe me, I'd change it to v7-dev if I could
23:10 john_minetest nobody uses v7 and it will break anyways.
23:10 hmmmm and I never intended for the choice of mapgen to be in a prominent dropdown list on the world creation menu
23:10 hmmmm that was somebody else's doing
23:11 hmmmm well hey, look, we're better than minecraft at this because they don't even give a single crap about map compatibility
23:11 hmmmm they'll change the mapgen from one version to another and have nice big inconsistent chunks
23:13 hmmmm yo yo yo hold on wtf
23:13 hmmmm https://github.com/minetest/minetest/commi​t/2330267d2207208799ad347ea0d129c0b4551d61​#diff-18513665750ef5adf42b5ec29e14162eR824
23:13 hmmmm I told you to change this before it gets merged
23:18 RealBadAngel hmmmm, ? are you talkin bout my changes?
23:28 Gethiox joined #minetest-dev
23:48 john_minetest woah, that's hard: no mesecons in moddb
23:57 iqualfragile VanessaE: upload mesecons to mmdb plz

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