Minetest logo

IRC log for #minetest-dev, 2014-06-30

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

All times shown according to UTC.

Time Nick Message
00:51 us`0gb joined #minetest-dev
01:03 Taoki joined #minetest-dev
01:39 diemartin joined #minetest-dev
01:40 CraigyDavi` joined #minetest-dev
02:00 Kalista joined #minetest-dev
03:28 diemartin joined #minetest-dev
04:57 Nerotic joined #minetest-dev
05:00 Nerotic Is it possible to pull request the minetest_game/default and become a Core Developer? Not sure, because some Core things are there, but I am not sure if that stuff belongs in the forum instead of an entire release.
05:01 Nerotic (if accepted that is)
05:10 ShadowNinja Nerotic: Pull request minetest_game's defaukt mod?  Can you explain what you meant better?
05:12 Nerotic ShadowNinja, Like, if I want to make a change to default, can I pull request it and if accepted, be in a release? Because usually anything in minetest_game that is changed would be a subgame/mod
05:14 ShadowNinja Nerotic: Yes, but we're in a feature freeze ATM, and minetest_game's been in an indefinite feature freeze for a while.
05:14 ShadowNinja So no features, but bug fixes are good.
05:14 ShadowNinja We'll accept features for the core in about a week.
05:14 Nerotic May I have a link to where I can find bug fixes to find and fix?
05:15 Nerotic I really want to contribute instead of make a fork that nobody will really use.
05:15 ShadowNinja Nerotic: The issue list and dev.minetest.net/TODO.
05:15 Nerotic Thanks man.
05:15 ShadowNinja Note that most of the TODO items are big.
05:15 hmmmm http://www.getacoder.com/proj​ects/programming_c_87390.html
05:16 Nerotic Thats alright, if I cant do it, Ill wait the week. I already have some major Core things done.
05:16 ShadowNinja Updating the dev wiki is a important but underappreciated job too.
05:16 Nerotic I have been working on the regular Wiki but not the dev
05:17 ShadowNinja hmmmm: Haha, looks like a great job.  ;-P
05:20 Nerotic Thanks. Lol, I found a simple thing to do thatll take me an hour or two in the TODO: and thats changing two textures for minetest_game. Gold and Bronze. Easy. I have dibs now xP
05:20 * Nerotic goes to sleep.
05:24 asl joined #minetest-dev
05:37 Hunterz joined #minetest-dev
06:08 ShadowNinja joined #minetest-dev
06:17 LemonLake joined #minetest-dev
06:20 kahrl hmmmm: I like the realistic budget!
06:21 kahrl I guess if nobody bids, he can still donate that money to ReactOS :)
06:22 Nerotic +1 kahrl
06:38 ShadowNinja joined #minetest-dev
07:03 nore joined #minetest-dev
08:39 Nerotic If I have something completed on the TODO list, can I simply edit it and delete it so nobody does whats already been done?
08:43 CraigyDavi`` joined #minetest-dev
08:45 RealBadAngel that should make a moderator or some1 responisble for that
08:46 RealBadAngel and ofc after that code has been merged
09:10 ImQ009 joined #minetest-dev
09:12 darkrose joined #minetest-dev
09:29 restcoser joined #minetest-dev
10:14 proller joined #minetest-dev
11:00 vifino joined #minetest-dev
11:10 Jordach joined #minetest-dev
11:30 nore without surprise, the commit causing the lighting bug is a0f78659f31abdce973bc6641b5b5a58f2ba9afc (Improved faces shading with and without shaders.)
11:31 nore RealBadAngel, ^
11:40 PenguinDad joined #minetest-dev
11:55 ImQ009 joined #minetest-dev
12:15 cg72 joined #minetest-dev
13:05 hmmmm joined #minetest-dev
13:33 Garmine joined #minetest-dev
13:51 sfan5 can someone finally delete the minetest/{common,build,survival} repos?
14:00 Megaf joined #minetest-dev
14:22 LemonLake joined #minetest-dev
14:37 celeron55 i'll do that
14:38 celeron55 done
14:57 nore Should https://github.com/minetest/minetest/issues/1431 go into 0.4.10 or 0.4.11 milestone? (since it is a regression, so should I label it as "blocker"?)
15:12 proller joined #minetest-dev
15:18 Megaf celeron55; Have you ever seen this chars?
15:18 Megaf crash(
15:18 Megaf *** glibc detected *** ./minetestserver: double free or corruption (fasttop): 0x027e8cc8 ***
15:18 Megaf Aborted
15:18 celeron55 yes, about 100 times
15:19 celeron55 which means: that tells nothing about the crash
15:19 Megaf I see that
15:19 Megaf A little frustrating
15:20 Megaf Anyway, git pull, cmake and make and move on
15:24 Sokomine celeron55: are you by any chance still familiar with the moving around of inventory items on the c++ side? i'd really love to get #944 fixed; it's bad that allow_metadata_inventory_put is blantantly ignored whenever items are swapped or temporally stored
15:25 Sokomine i havn't found anyone willing/able to fix that yet, and i'm not that familiar with the code yet either
15:30 Krock joined #minetest-dev
15:33 proller joined #minetest-dev
15:36 celeron55 all that logic seems to be in IMoveAction::apply
15:38 celeron55 well actually not all... but that's where the additional check has to go
15:40 celeron55 it has to check that either all the content from the source slot fits in the destination slot, or if that isn't the case, the source slot must allow "put", and that isn't the case, the move is disallowed
15:40 celeron55 +if
15:40 celeron55 currently it does not do that chain of checks
15:43 celeron55 is there a reliable test case for this?
15:43 celeron55 can you provide one
15:44 celeron55 (if i am fixing this, i won't go the route of adding a temporary inventory slot; it's more work and probably causes more problems that it solves)
15:44 VanessaE moreblocks circular saw.  take an item out of its output slots and drop it on an item in your inventory.  the item in the destination slot will be deleted.
15:45 celeron55 than*
15:46 Sokomine ah. testcase. my circular saw (comes with moreblocks) is a good test case
15:47 Sokomine vanessa was faster
15:48 Sokomine the reason why the item is deleted is because i intended the output-slots to be output-only. on the lua side, i do everything to ensure that. but there is no way for me to catch those items that are temporally stored there - all i could do would be iterate through all the inventory before updating in order to see if there's something unexpected
15:50 Sokomine something similar also happens with half-stacks etc
15:50 VanessaE doesn't a furnace also exhibit this issue?
15:50 Sokomine it was discussed here (but afaik no result reached) weather in such cases an extra inventory slot for items that are supposed to stick to the mouse pointer ought to be introduced
15:51 Sokomine yes, but it's probably less important there. no big deal if there's something in the output-slot of a furnace that didn't get there the normal way
15:51 celeron55 the furnace only adds item to the output slots so it doesn't matter (it doesn't preview anything there)
15:51 celeron55 items*
15:51 Sokomine it is only important for nodes that rely on allow_metadata_inventory_put/take/move
15:52 Sokomine or, for that matter, on_metadata_inventory_put/take/move (which is not called in the described case either)
16:06 celeron55 if only a partial stack is being moved and the source does not allow "put", i guess the move should be disallowed
16:06 celeron55 otherwise it's a nightmare to manage by whoever manages the source
16:06 celeron55 hmm wait
16:07 celeron55 yes, this interface is unable to handle that in any sane way
16:08 celeron55 umm... actually, that can be disallowed by the "allow take" callback
16:09 celeron55 yes, i'm not going to add any new checks for that case; it's handleable already
16:10 celeron55 only the #944 case is not handleable
16:15 Hunterz joined #minetest-dev
16:16 Calinou joined #minetest-dev
16:19 werwerwer joined #minetest-dev
16:27 kaeza joined #minetest-dev
16:29 zat joined #minetest-dev
16:29 sfan5 who processes the domain registrations for .minetest.net?
16:30 celeron55 freedns.afraid.org
16:30 sfan5 I mean the person who approves the domain registrations done therew
16:30 sfan5 -w
16:30 celeron55 me
16:30 ShadowNinja celeron55: I think he means "wh...^
16:30 celeron55 altough, i probably won't approve anything there anymore, because there is no way to disapprove them
16:31 celeron55 i'll just leave them laying around unapproved no matter what
16:31 sfan5 :(
16:31 celeron55 they work in that state too
16:31 sfan5 they do?
16:31 sfan5 TIL
16:31 sfan5 did not work yesterday
16:32 nore I might have a look at #944 in days to come... btw, do you think #1431 should be labeled at blocker, since it is a regression?
16:32 ShadowNinja Approval only removes it from the queue, disaproval removes it from DNS.
16:33 sfan5 ShadowNinja: in case you want to do some tests with the serverlist I host: http://meow.minetest.net/
16:33 celeron55 ShadowNinja: yes, and once approved, there is no mechanism for removing them from DNS
16:33 Nerotic When the minetest_game isn't frozen anymore, will the current pull requests still be up for conversation?
16:33 sfan5 Nerotic: minetest_game will likely not get unfrozen
16:33 Nerotic Whys that?
16:34 sfan5 ShadowNinja: where is the python serverlist again?
16:34 celeron55 nore: i am working on it now
16:34 nore celeron55, ok
16:34 ShadowNinja Players: 0/undefined I thought I fixed that...
16:34 sfan5 ShadowNinja: I did not update yet
16:34 Nerotic Meh, who cares, I'll go find some bug fixes I am capable of fixing.
16:34 sfan5 wait wut
16:34 sfan5 there are 0 servers but it says 0/1
16:34 nore sfan5, perhaps change the name from minetest_game to minimal ;)
16:35 ShadowNinja celeron55: What do you think about moving the serverlist to a seperate repo?  You shouldn't have to clone and pull all of Minetest just for a small server.
16:35 Nerotic ^
16:35 sfan5 I don't think we need that
16:35 VanessaE nore: and current minimal becomes "dev_testing" maybe? :)
16:35 celeron55 ShadowNinja: i think it should be in a separate repo
16:36 celeron55 ShadowNinja: if it is developed more, that is
16:36 Nerotic I would vote for Vanessa's idea.
16:36 celeron55 because it will grow much larger than what an "utility" should be
16:36 nore VanessaE, something like that
16:36 sfan5 celeron55: if we use the python thing it will get dev'ed more, nobody understands that perl script
16:36 VanessaE Nerotic: it was a joke, not an idea :P
16:36 Nerotic I still like it
16:36 Nerotic <3
16:37 ShadowNinja sfan5: Did one ever announce?
16:37 Nerotic is there a list of needed translations for the game? I wouldn't mind doing that.
16:37 sfan5 yes
16:37 rubenwardy joined #minetest-dev
16:37 ShadowNinja sfan5: It's currentServers/maxServers
16:38 sfan5 ah
16:38 LemonLake I know that some people use minimal as a game-base, though
16:38 sfan5 ShadowNinja: http://sprunge.us/eRYF
16:39 sfan5 also: will http://translate.minetest.ru ever get fixed?
16:39 sfan5 it.total_max && it.total_max.clients > 0out+='P[...]
16:40 sfan5 seems like there is a ; missing
16:40 Nerotic sfan5; its not that hard, its just a 502 error, if somone had the motivation it could be potentially fixed within an hour
16:40 Nerotic *someone
16:41 ShadowNinja sfan5: Uh, that's not the python serverlist.
16:41 ShadowNinja Guess it wasn't sopposed to be.
16:41 ShadowNinja But can you switch to it?
16:42 werwerwer joined #minetest-dev
16:42 sfan5 ShadowNinja: I'll try it if you answer my question
16:43 sfan5 <sfan5> ShadowNinja: where is the python serverlist again?
16:44 ShadowNinja sfan5: It's in a PR, ShadowNinja/python-server-list branch.  And is that only after proller's "fix"?
16:44 ShadowNinja My branch doesn't have that, but I'll have to update and revert that.
16:44 sfan5 ShadowNinja: only after the "fix"
16:47 proller onse second..
16:48 ShadowNinja proller: Just revert it and fix it properly, it's not a fix, it's a workaround.
16:49 sfan5 proller: you can close that issue https://github.com/minetest/minetest/issues/1432
16:52 sfan5 ShadowNinja: would be helpful if the README.md would tell me how to set this thing up
16:56 ShadowNinja sfan5: It does.
16:56 ShadowNinja sfan5: Install the depependencies, edit config.py, and run it.  Then point nginx at it.
16:57 sfan5 https://github.com/ShadowNinja/minetes​t/tree/python-server-list/util/master
16:57 sfan5 oh wat
16:57 sfan5 wait*
16:57 sfan5 nvm
17:03 Calinou Carbone loads in latest Git but smooth lighting is still BROKEN
17:04 grrk-bzzt joined #minetest-dev
17:04 grrk-bzzt joined #minetest-dev
17:04 Calinou below the 2 dirt blocks, there's a cave. lighting looks bright if you remove the dirt blocks, otherwise a weird shadow appears: https://cdn.mediacru.sh/9eFq7w7HMgMx.jpg
17:04 celeron55 Sokomine: try this http://www.fpaste.org/114402/41478601/
17:05 scientistnik joined #minetest-dev
17:09 celeron55 it needs a bit of regression testing too
17:09 celeron55 maybe someone could quickly put it in use on a server to see if it breaks any regular gameplay
17:12 ShadowNinja We also need a new weblate translations site host.
17:13 Krock https://github.com/minetest/minetest/issues/1433
17:13 ShadowNinja sfan5: I (force) pushed a rebase and revert of 897c2.
17:14 sfan5 ShadowNinja:   Could not find any downloads that satisfy the requirement apschedule
17:14 sfan5 nv
17:14 sfan5 m
17:14 sfan5 I needed the python 3 pip
17:17 ShadowNinja I'd like to merge the client-side parts of my server list pull before 0.4.10, since they're just fixes.
17:17 celeron55 Calinou: is that a new bug or has it always existed?
17:18 celeron55 Calinou: if it is a new bug, shout at RBA when you have a chance, because he has been most closely involved with those things
17:18 celeron55 if not, then don't expect it to get fixed anytime soon :P
17:18 ShadowNinja Well, multi-part is a pretty big fix.  I might be able to exclude that.
17:19 sfan5 ShadowNinja:   Could not find any downloads that satisfy the requirement apschedule
17:19 Calinou new bug: smooth lighting wasn't broken like that before
17:19 Calinou areas between blocks are not darkened at all
17:19 Calinou it's broken since something like HEAD~15
17:19 Calinou https://github.com/minetest/minetest/commi​t/35ec3855f689cf2c6b1504a5813b7c3d9697ae14 → this may be the last working one
17:19 scientistnik1 joined #minetest-dev
17:20 sfan5 ShadowNinja: can flask handle as much requests as nginx?
17:20 celeron55 it could be a0f78659f31abdce973bc6641b5b5a58f2ba9afc or it could be some of sapier's performance fixes
17:21 sfan5 oh wait
17:21 sfan5 we can use fcgi
17:22 Calinou performance fixes didn't change CPU usage of the server for me, it's very similar
17:22 Calinou probably helps client more
17:24 celeron55 if 35ec is the last working one, then obviously 334ec4bb1b031cd3423e52fdc7f7ec9e4dec8c84 breaks it
17:24 celeron55 and it's one that very well could break it
17:24 celeron55 actually, maybe not
17:25 celeron55 are you sure? check it properly, nobody is going to do that for you
17:25 Calinou I tried each commit one by one, for the Carbone crash issue, but that crash is now solved
17:25 Calinou for the smooth lighting, I'm not sure, but I think the crash correlates with the smooth lighting being broken
17:25 celeron55 it probably doesn't
17:25 celeron55 i'm not going to look into this more until you have more accurate info
17:26 Calinou will try several commits to see the one breaking it
17:28 sfan5 ShadowNinja: http://flask.pocoo.org/docs/deploying/ "You can use the builtin server during development, but you should use a full deployment option for production applications. (Do not use the builtin development server in production.)"
17:30 ShadowNinja sfan5: Uh, find out how to start a production server then.
17:31 sfan5 ShadowNinja: it's APScheduler not apscheduler
17:31 ShadowNinja sfan5: Alright, I'll fix that.
17:32 sfan5 hm :/
17:32 ShadowNinja client fixes in server-list-client-fixes  removing client stuff from python-server-list...
17:34 sfan5 ShadowNinja: wtf
17:34 sfan5 except FileNotFoundError:
17:34 sfan5 NameError: global name 'FileNotFoundError' is not defined
17:35 Krock lol ^
17:36 celeron55 that's python in a nutshell
17:36 sfan5 ShadowNinja: we can't use python3 in production, flup does not work with python3
17:38 Calinou found the commit
17:38 Calinou latest working commit: fcfd53a “Don't keep temporary images in RAM”
17:39 Calinou the next one is what breaks stuff
17:39 sfan5 what the fuck
17:39 sfan5 File "/usr/lib/python2.7/json/encoder.py", line 353, in _iterencode_dict
17:39 sfan5 newline_indent = '\n' + (' ' * (_indent * _current_indent_level))
17:39 sfan5 TypeError: can't multiply sequence by non-int of type 'str'
17:39 sfan5 everything is broken
17:40 casimir joined #minetest-dev
17:40 proller lol, but why you doing this ?
17:43 Exio4 joined #minetest-dev
17:43 tomreyn joined #minetest-dev
17:45 sfan5 uwsgi should work too...
17:48 ShadowNinja sfan5: It uses indent='\t' and old versions only worked with indent=123(numSpaces)
17:54 diemartin joined #minetest-dev
17:58 sfan5 ShadowNinja: ok, running http://meow.minetest.net/
18:02 sfan5 ShadowNinja: http://sprunge.us/MIFG
18:06 ImQ009 joined #minetest-dev
18:12 sfan5 ShadowNinja: there is zero caching, fix that please
18:13 sfan5 or just directly let the webserver serve those files
18:14 sfan5 ShadowNinja: sorry, not zero caching, but the / page isn't cached
18:31 VanessaE joined #minetest-dev
18:33 ShadowNinja sfan5: send_static_file does caching.
18:33 ShadowNinja See /list, I had to explictly disable caching.
18:37 VanessaE joined #minetest-dev
18:44 sfan5 ShadowNinja: I'm talking about index.html
18:44 sfan5 ShadowNinja: /list is cached to some degree too "" If-Modified-Since:Mon, 30 Jun 2014 18:21:54 GMT\r\nIf-None-Match:"flask-140​4152514.924497-113-3612938977" ""
18:44 scientistnik joined #minetest-dev
18:47 sfan5 ShadowNinja: did you merge http://sprunge.us/MIFG yet?
18:54 Piggybear87 joined #minetest-dev
18:54 Piggybear87 left #minetest-dev
18:54 CraigyDavi` joined #minetest-dev
18:59 nore what's the flag for a debug build?
18:59 nore nvm, just found it
19:01 VargaD joined #minetest-dev
19:07 nore sfan5, I'm trying to reproduce the bug... however, it's quite random
19:07 sfan5 ShadowNinja: http://sprunge.us/bDBh and http://sprunge.us/fYJU
19:07 sfan5 nore: then try again :)
19:07 VanessaE joined #minetest-dev
19:08 nore there
19:11 nore sfan5, still #0 0x00000000014e2030 in ?? ()
19:11 nore (in debug build)
19:12 sfan5 I'm afraid you need an Irrlicht debug build too
19:12 nore argh... how can I get one?
19:17 ShadowNinja sfan5: MIFG merged.
19:20 ShadowNinja bDBh isn't strictly related, and I'll fix fYJU
19:20 sfan5 but bDBh is a good addition
19:20 iqualfragile joined #minetest-dev
19:21 sfan5 ShadowNinja: you can leave out the font change, but the names/descs. in servers.jst should be changed
19:24 ShadowNinja sfan5: Sure, but that isn't related to the server list rewrite.  Add it seperately.  And Mapgenerator -> Map generator or MapGenerator (or just MapGen).
19:26 sfan5 ShadowNinja: the rewrite is a partial redesign anyway
19:27 ShadowNinja sfan5: Actually only the backend is changed.  I even left pop_v in place (instead of "popularity")
19:27 sfan5 if possible I'd like to get the serverlist moved to a seperate repo today and the python server list rewrite/redesign merged
19:27 sfan5 ShadowNinja: you can't change that because backwards compat.
19:27 ShadowNinja sfan5: So it works?
19:27 sfan5 the python one, yes!
19:28 sfan5 it has been running on http://meow.minetest.net/
19:28 ShadowNinja O.K.  But the client fixes first.
19:28 sfan5 I haven't tested annouce yet, but I'd be surprised if it didn't work
19:28 sfan5 lemme test announcing
19:28 sfan5 there are no client fixes to be done
19:28 sfan5 (not today at least)
19:29 sfan5 hm :/
19:29 sfan5 21:29:12: ERROR[CurlFetchThread]: meow.minetest.net/announce?json=%7B%22action​%22%3A%22start[...]%3A%220.4.9-dev%22%7D%0A not found (HTTP response code said error) (response code 400)
19:30 ShadowNinja sfan5: I mean ShadowNinja/server-list-client-fixes I split it from py-s-l.
19:30 sfan5 ah
19:31 ShadowNinja Uh, 400 might be the "OK, but this needs some more background processing" code that I used.
19:31 sfan5 >Invalid JSON data.
19:31 sfan5 wut
19:31 ShadowNinja Hmmm.
19:31 sfan5 it seems to be perfectly valid json
19:32 sfan5 yep
19:32 sfan5 python can load it
19:32 ShadowNinja sfan5: Yes, that meand a field is missing or of the wrong type.
19:32 ShadowNinja +s
19:32 nore sfan5, there: #0 0x00007ffff4421d18 in main_arena () from /lib/x86_64-linux-gnu/libc.so.6
19:33 sfan5 o.O
19:33 ShadowNinja It accepts anything in bool fields because of dad design.
19:33 ShadowNinja +already
19:33 sfan5 ShadowNinja: how do I fix this?
19:33 sfan5 nore: whole bt full?
19:34 nore nah, that's just the change now that I have irrlicht symbols
19:34 ShadowNinja sfan5: Compare the request with the requirements in checkRequest/fields.
19:34 sfan5 it would be better if it would tell me
19:34 ShadowNinja Or just add print()s before returns and try again.
19:34 nore btw: I found another segfault while trying to reproduce this one
19:36 ShadowNinja Or app.logger.warn() or so.
19:36 scientistnik1 joined #minetest-dev
19:38 sfan5 missing field: uptime
19:38 sfan5 ^ ShadowNinja
19:39 ShadowNinja Hmmm, odd.
19:40 ShadowNinja Is it a start or update request?
19:41 ShadowNinja Might not be sent on start.
19:41 sapier joined #minetest-dev
19:41 sfan5 ShadowNinja: start
19:49 ShadowNinja sfan5: Re-pull and try again.
19:51 RealBadAngel joined #minetest-dev
19:51 sfan5 hm
19:51 sfan5 ShadowNinja: it announced but does not appear on the list
19:52 ShadowNinja sfan5: What's the response?  202?
19:52 RealBadAngel i have figured out problem with drawtypes
19:52 sfan5 [pid: 2458|app: 0|req: 24/24] 79.235.52.52 () {30 vars in 2176 bytes} [Mon Jun 30 15:52:40 2014] GET /announce?json=%7B[...]%0A => generated 36 bytes in 1 msecs (HTTP/1.1 202) 2 headers in 85 bytes (1 switches on core 0)
19:53 sfan5 yes
19:53 RealBadAngel gonna fix it
19:53 ShadowNinja Check caching and add print()s.
19:53 RealBadAngel but definitely i will need day or two
19:53 RealBadAngel its broken to the bone
19:54 sfan5 ShadowNinja: add print() s where
19:54 ShadowNinja sfan5: Is it accessible (!up-style ping) from it's address (or IP if none provided)
19:55 sfan5 yes it should be
19:55 sfan5 <MinetestBot> sfan5.duckdns.org:30000 is up (590ms)
19:55 ShadowNinja sfan5: asyncFinishThread.
19:56 sapier calinou  what was broken by removal of temporary images?
19:57 sfan5 ShadowNinja: the deannounce it does generates a 500 btw
19:58 sapier performance fixes most likely give most of it's benefit to clients as those operations where things have been fixed are mostly client only ... or at least way more often used there
19:59 sfan5 ShadowNinja: could you try announcing to it? also I still don't know for what to add print()'s
19:59 sapier "Calinou  I tried each commit one by one, for the Carbone crash issue ..." that's a very precise error report ... I'd love to check it but with information like that it's gonna be impossible
20:00 ShadowNinja sfan5: Add one before every return.  I'll check the 500.
20:01 ShadowNinja The 500 is because the server wasn't added.
20:01 sapier hmm guess it's good we have a full week feature freeze ;-)
20:03 sfan5 ShadowNinja: app.logger.warning is printed to console, right?
20:03 sapier yes but there's different behaviour for different devices ... seems manufacturer can disable console
20:04 Calinou Carbone crash is fixed, sapier.
20:04 Calinou <Calinou> latest working commit: fcfd53a “Don't keep temporary images in RAM”
20:04 Calinou <Calinou> the next one is what breaks stuff
20:04 sfan5 I'm not talking about android
20:04 Calinou the smooth lighting is still broken
20:04 Calinou I tested each commit one by one.
20:04 Calinou (about 15)
20:04 sapier how has it been fixes?
20:04 sapier -s+d
20:04 Calinou I don't know, but it runs fine now
20:05 ShadowNinja sfan5: I don't know.
20:05 sapier ah so no crash but broken lighting?
20:05 sfan5 ShadowNinja: wonderful
20:05 ShadowNinja :-|
20:06 sapier "don't know" bugs usually come back again
20:06 sapier https://github.com/minetest/minetest/commi​t/d5de0cbf71d65ce586b3ab2469f2061ccc4b506a this one breacks lighting?
20:07 Calinou probably, since I tested each commit one by one.
20:07 Calinou mapblock_mesh.cpp maybe
20:07 sapier I'll check if I did miss a case on resorting it
20:08 Calinou https://cdn.mediacru.sh/9eFq7w7HMgMx.jpg → this happens, too: if there's air below a block, a random shadow will appear for no reason
20:08 Calinou remove the dirt blocks and the shadow is gone
20:08 sfan5 ShadowNinja: nothing
20:08 Calinou there's air below the dirt (a cave)
20:08 ShadowNinja " The default configuration is to log to stderr if the application is in debug mode."
20:08 ShadowNinja Production isn't mentioned.
20:09 ShadowNinja vim /usr/lib/python3.4/site-packages/flask/app.py
20:09 sfan5 ----------------------------------------​----------------------------------------
20:09 sfan5 WARNING in server [./server.py:294]:
20:09 sfan5 ping missing
20:09 sfan5 ----------------------------------------​----------------------------------------
20:12 sfan5 ^ ShadowNinja
20:14 RealBadAngel i have figured out whats wrong with the drawtypes
20:14 sfan5 RealBadAngel: you're repeating yourself
20:14 RealBadAngel propably
20:15 sapier is this calinous problem?
20:15 RealBadAngel but now i know
20:15 RealBadAngel and i can say one thing
20:15 RealBadAngel we are fucked sideways
20:16 RealBadAngel current approach to rendering is just wrong
20:17 ShadowNinja sfan5: The server isn't accessible then.
20:17 RealBadAngel without a revolution here we wont get too far
20:17 sfan5 ShadowNinja: wouldn't it return False then?
20:17 ShadowNinja sfan5: Is server_address set properly?
20:17 RealBadAngel first of all, we shall return to render a node
20:17 ShadowNinja sfan5: Yes, doesn't it?
20:18 RealBadAngel fucking single node
20:18 RealBadAngel which is not whole world
20:18 RealBadAngel but a cube
20:19 ShadowNinja sfan5: If it's None remove the wildcard except: and you'll probably get something more descriptive.
20:19 sfan5 wtf
20:19 sfan5 why is port a string
20:20 ShadowNinja ...
20:21 sfan5 yay
20:21 sfan5 worked
20:21 sfan5 (hotfix: port=int(port))
20:22 sfan5 ping: 5542
20:22 sfan5 uhhh
20:22 sfan5 <MinetestBot> 79.235.52.52:30000 is up (258ms)
20:22 sfan5 :/
20:25 sfan5 ShadowNinja: you should disable pretty indent if debug is disabled
20:28 ShadowNinja Fix pushed.  sfan5: Maybe, but it isn't that much bigger.
20:28 sfan5 oh are you sure?
20:28 ShadowNinja sfan5: Does the in-menu client serverlist work?
20:29 ShadowNinja sfan5: ~4 bytes per field.
20:29 ShadowNinja Since it uses tabs.
20:31 sfan5 >>> len(data)
20:31 sfan5 78784
20:31 sfan5 >>> len(json.dumps(json.loads(data), indent=1))
20:31 sfan5 108870
20:31 sfan5 of the actual serverlist json data
20:32 sfan5 ShadowNinja: yes, works
20:32 proller joined #minetest-dev
20:33 ShadowNinja Good.  I'll fix the indent.  Can you test server-list-client-fixes now?  (it doesn't have to work with the old server list because the server can be updated easily)
20:34 sfan5 >>> ((big * 1.0) / small) * 100
20:34 sfan5 138.18795694557272
20:34 sfan5 138%
20:37 sfan5 ShadowNinja: https://github.com/ShadowNinja/minete​st/commit/b4c2e79ea49fb7d2c2ac673c004​54070462ef1ad#commitcomment-6846419
20:40 celeron55 sapier, RealBadAngel: make sure to figure out a fix for Calinou's problem; it worked before so it should work now too
20:40 ShadowNinja sfan5: Anything else?
20:40 celeron55 just revert the commit or something
20:41 sapier I don't fully understand what his problem is ;-)
20:41 celeron55 then ask him
20:41 celeron55 you broke it, so you fix it
20:41 sapier I did :-) for what I understand it's shadows
20:42 Calinou smooth lighting is broken; is this not obvious on this screenshot? https://cdn.mediacru.sh/9eFq7w7HMgMx.jpg → there should be some ambient occlusion, there's none
20:42 celeron55 i figure just reverting the "Minor fix in check order" commit will fix the problem
20:42 Calinou it's not shadows exactly, it's ambient occlusion, which doesn't work in most cases, except when it shouldn't! see the problem with air blocks below other blocks, causing a strange shadow to appear
20:42 Calinou you can keep the whitespace removals
20:42 sapier maybe but I don't understand why
20:42 Calinou the check is likely how smooth lighting is correctly done
20:43 sapier I've got no problem on reverting it but I'd like to know why ;)
20:43 celeron55 i might be able to figure out but i would have to re-study the code and i wasn't involved in making the "fix" that broke it either so i'm not bothering to
20:44 sapier lol
20:44 sapier calinou can you change the == to != ?
20:44 sapier stop
20:44 sapier forget about it
20:44 Calinou OK
20:45 sfan5 ShadowNinja: no, it's fine
20:46 sapier that fix is that broken ... I'll give you a fix for it to try
20:47 sfan5 ShadowNinja: you can merge server-list-client-fixes
20:47 cg72 joined #minetest-dev
20:47 sapier https://gist.github.com/ano​nymous/cfd76a9217bd533ff7e9 calinou can you try this one?
20:50 Calinou this works, sapier
20:50 Calinou used patch command to apply the patch
20:50 ShadowNinja sfan5: Ready to switch servers.minetest.net to your server?  Will it have 99%+ uptime?
20:50 Calinou what's wrong with servers.minetest.net?
20:51 ShadowNinja Calinou: xyz doesn't want to maintain it?
20:51 ShadowNinja -?
20:51 sfan5 ShadowNinja: it won't have 100% percent uptime, but I will try to keep it working
20:51 sfan5 but no move yet
20:51 sfan5 (not yet = still today if possible)
20:52 sfan5 I need to make some stuff permanent
20:52 sfan5 ShadowNinja: can we move the serverlist to a seperate repo first?
20:52 ShadowNinja sfan5: Sub-100% is O.K.  but it should be 99% or above (that kives you almost 4 full  days per year).
20:53 ShadowNinja sfan5: Sure.  celeron55: Can you make it?
20:53 cg72 joined #minetest-dev
20:53 proller joined #minetest-dev
20:54 sfan5 ShadowNinja: I should be able to get that going
20:54 sfan5 left #minetest-dev
20:54 sfan5 joined #minetest-dev
20:57 celeron55 ShadowNinja: make what?
20:57 sfan5 celeron55: create a minetest/masterserver repo
20:57 sapier calinou I pushed that fix do you have other issues?
21:00 celeron55 sfan5: done
21:00 sfan5 ok, thanks
21:00 sfan5 ShadowNinja: can you move the masterserver?
21:02 werwerwer_ joined #minetest-dev
21:02 ShadowNinja Yep, working on it...
21:04 ShadowNinja Done.
21:07 Calinou with your patch, shading of standalone blocks is worse than before: https://cdn.mediacru.sh/JHtxMCVeBb4r.jpg
21:07 Calinou (compared to before the fact you broke lighting)
21:08 Calinou but we can do with that, it's not very important
21:08 Calinou (to make it obvious: place a single cobble block in front of you)
21:08 Calinou start placing a group of it and see what happens
21:09 ShadowNinja I'll push a master server removal patch to minetest in a minute.
21:09 ShadowNinja celeron55: Can you rename it to master-server please?  :-)
21:10 celeron55 i think masterserver and master-server are both stupid names
21:10 ShadowNinja server-list?
21:10 celeron55 do these kinds of servers ever have any proper names
21:10 ShadowNinja That's more specific though.
21:10 Calinou https://cdn.mediacru.sh/NC8KFJVYljxN.jpg and https://cdn.mediacru.sh/VllUPf1uCg7_.jpg (look at the left)
21:11 Calinou celeron55, master server is often used in gaming
21:11 Calinou all the Cube and Cube 2 games use that term, Stunt Rally too…
21:11 ShadowNinja celeron55: master server or server list.
21:11 Calinou server list makes it sound like it's not really a server, while it is
21:11 ShadowNinja I think AoS uses the same terminology.
21:11 Calinou the master server could do more things: handle global bans, etc
21:12 sfan5 >global bans
21:12 sfan5 >open source game
21:12 celeron55 i renamed it to master-server then
21:12 Calinou sfan5, yes, exactly
21:12 sfan5 sapier would never accept that
21:12 Calinou Cube 2 games do have global bans and server bans (servers that never appear in list, for various reasons)
21:12 Calinou only way to not be affected, legally, is to make the server not announce to master, sadly
21:12 ShadowNinja sfan5: Global bans aren't bad, as long as they aren't abused.
21:13 Calinou (if you get spotted not obeying global bans, the server would be blacklisted anyway)
21:13 Calinou global bans should remain an eventuality, in order to bans people whose only purpose is to cause havoc
21:13 Calinou griefers, DDoS…
21:13 ShadowNinja sfan5: Oh, and add those two bad servers to the ban list.
21:13 Calinou for custom master servers too
21:13 sfan5 ShadowNinja: which ones?
21:14 Calinou server blacklist for terrible servers full of kids and guests :p
21:14 proller also dont forget ban all freeminer servers
21:14 ShadowNinja sfan5: sektori and nwp.us IIRC...
21:14 sfan5 proller: I won't :)
21:15 ShadowNinja proller: Not until FM's illegal or smth.
21:15 proller ShadowNinja, how you measure "bad server"&
21:15 proller ?
21:15 Calinou FM may already be illegal, the relicense was done without asking anybody
21:16 Calinou proller, bad server quality in general: purposeless players, lag, faking player count or ping…
21:16 ShadowNinja proller: Is it appropriate for children, legal, etc.
21:16 proller BAN BAN BAN fm!!
21:16 sfan5 Calinou: you can't fake ping
21:16 proller Calinou, then ban ALL servers except VE
21:16 ShadowNinja Unless you fake a high ping.  ;-)
21:17 Calinou the most populated ones are the worst maintained ones, the admins are never there, it's full of guests, kids, mobile clients who have no clue…
21:17 Calinou “just test” is a great a sad example
21:17 proller and?
21:17 Calinou you can go there and spend hours killing people, basically nothing else :P
21:17 Calinou people who connect to these servers for the first time will be deceived
21:17 proller you spend weeks for griefing my server
21:19 proller and now you want to ban all servers where playing peoples like you
21:20 kaeza [citation needed]
21:20 kaeza proller, mind being less childish?
21:20 proller also all my servers is "worst maintained ones, the admins are never there, it's full of guests, kids, mobile clients who have no clue" so ban all!
21:21 proller kaeza, ask Calinou yourself
21:21 proller also griefed map still online
21:22 kaeza no, the only servers that are getting banned are those that have inappropriate content in the MOTD
21:22 proller define "inappropriate"
21:23 sfan5 ShadowNinja: the README.md is broken
21:23 khonkhortisan not suitable or proper in the circumstances.
21:23 sfan5 ShadowNinja: did you fix port being a string?
21:26 ShadowNinja sfan5: 1. I'll check, but not mission-critical.  2. Yes, along with clients_max.
21:26 sfan5 ok
21:27 proller time to open second "free" masterserver..
21:30 Megaf_ joined #minetest-dev
21:30 ShadowNinja README fixed.
21:33 sfan5 23:32:57: ERROR[CurlFetchThread]: meow.minetest.net/announce?json=%7B%22ac​tion%22%3A%22delete%22%2C%22address%22%3​A%22%22%2C%22port%22%3A%2230000%22%7D%0A not found (HTTP response code said error) (response code 500)
21:33 sfan5 ^ ShadowNinja
21:35 sfan5 [pid: 3685|app: 0|req: 10/10] 79.235.52.52 () {30 vars in 592 bytes} [Mon Jun 30 17:35:15 2014] GET /announce?json=%7B%22action%22%3A%2​2delete%22%2C%22address%22%3A%22%22​%2C%22port%22%3A%2230000%22%7D%0A => generated 17 bytes in 1 msecs (HTTP/1.1 500) 2 headers in 98 bytes (1 switches on core 0)
21:38 ShadowNinja sfan5: That should only happen if the announce failed or the previous update was missed and it timed out.
21:39 sfan5 the announce worked
21:39 sfan5 http://meow.minetest.net/list
21:42 ShadowNinja sfan5: It sent {"action":"delete","address":"","port":"3000"}
21:42 Taoki joined #minetest-dev
21:42 ShadowNinja address is messed up.  IP should probably be used though.
21:43 ShadowNinja IP is used...
21:43 sfan5 maybe something else was broken
21:43 sfan5 I need to sleep now
21:44 sfan5 let's do the move tomorrow
21:44 sfan5 (or later)
21:45 ShadowNinja If it was on the list at that time it should have worked.
21:46 sfan5| joined #minetest-dev
21:47 sfan5| ShadowNinja: please test http://meow.minetest.net
22:03 proller https://github.com/minetest/minetest/pull/1390
22:03 sapier 0.4.11
22:07 Jordach joined #minetest-dev
22:50 ImQ009 joined #minetest-dev
23:03 paramat joined #minetest-dev
23:08 paramat Calinou's smooth lighting problem may be due to his/her use of the 'debugging in dark' lighting table, from logs: "Calinou > I use the “debug” light table, but why is ambient occlusion broken (sides of nodes)?"
23:09 sapier paramat: according to calinou it's already fixed by latest commit
23:16 paramat duh! thanks i just read logs again and realised :}
23:18 ShadowNinja sfan5: I can't announce to it. It might need to wait for the server to start listening before pinging it.  Try with time.sleep(3) before the ping.
23:21 cg72 joined #minetest-dev
23:22 sapier left #minetest-dev
23:25 paramat im using yesterday's commit 7f9990, and noticed side texture brightness had changed with smooth lighting off, side textures are equally dark, this looks worse: less depth, less 3D looking
23:46 Nerotic joined #minetest-dev
23:53 paramat i'll chat to RBA about it too https://github.com/minetest/minetest/issues/1436
23:57 VanessaE joined #minetest-dev

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