Time Nick Message 00:00 sapier1 now it's "lemme look in chat logs to find some chat where someone told something to be interpreted as approoval" 00:01 kahrl I'd be okay with doing that 00:01 ShadowNinja While we are on pull requests, does this look good? https://github.com/minetest/minetest/pull/856 00:02 sapier1 and yes I know it's a little bit of work but I guess it's gonna reduce some discussions thus it may be even more efficient 00:02 ShadowNinja PilzAdam: ^? 00:03 PilzAdam ShadowNinja, this was brought up many times already and my comment doesnt change: pass a pointed_thing to it and call it for objects too 00:03 sapier1 think about it kahrl I don't have any problems with decision but I hate it if ppl are frightened about decisions and just go on without deciding anything 00:04 kahrl sapier1: I doubt it would help things much, because the problem is getting the approval in the first place, but it wouldn't be a bad thing 00:05 ShadowNinja PilzAdam: This is for node protection. Jou just want to check if one particular node is protected. Objects are irrelivant. 00:05 sapier1 it helps i case you don't get two approvals at same time 00:05 PilzAdam ShadowNinja, why exlude objects from protection mods? 00:06 kahrl sapier1: though if a dev says "I approve of commit 2342cafebabe" in the chat, don't nudge him to comment on the pull request, just add a link to the logs 00:06 sapier1 PilzAdam if you include objects without any difference this may result e.g. in hostile mobs beeing able to attack players while player can't fight back 00:06 ShadowNinja PilzAdam: Because there aren't any mods that protect objects. 00:06 ShadowNinja I don't see why you would want to. 00:07 PilzAdam ShadowNinja, so you are free to walk into others people houses and steal their dropped items? 00:07 kahrl ShadowNinja: protect them from the engine :P 00:07 kahrl to avoid it getting duplicated 00:07 PilzAdam sapier1, thats the mods fault then 00:07 sapier1 no that's a missing feature as interface doesn't differenciate 00:08 sapier1 I agree nodes and entities should be handled similar but imho there should be ways to handle them different too if required for some usecases 00:09 ShadowNinja PilzAdam: When you drop an item you normally want other users to be able to pick it up. It is used for giving items to other users. There are also other entities that will need to be usable with protection. 00:09 sapier1 shadow what about calling a function specified within the entity? 00:10 sapier1 e.g. the can interact checks if entity.interact_check is a function interpreting it's return value 00:11 sapier1 this way a entity could specify itself if it wants to be handled by interact protection or not 00:11 ShadowNinja This can actually already be used that way of you want. Just add if not minetest.can_interact(vector.round(self.getpos())) then return end 00:11 ShadowNinja + To the begining of on_punch. 00:12 ShadowNinja In fact the rounding step probably isn't necessary. 00:12 VanessaE bbl 00:12 ShadowNinja PilzAdam: ^ 00:13 ShadowNinja kahrl: What are your thoughts on it? 00:13 kahrl ShadowNinja: I don't know anything about protection mods so I can't comment 00:14 ShadowNinja Hmmm, sfan5: Are you around to check it? 00:14 sapier1 I guess this shows up another problem ... things none of the core developers is used to are difficult to add at all ;-) 00:14 ShadowNinja kahrl: How about my bind_address pull? 00:14 sapier1 but I need to get some sleep good night guys 00:15 kahrl ShadowNinja: it looked fine last time I briefly looked at it, but I'm not in the state of mind to review it right now 00:15 kahrl sorry 00:15 ShadowNinja PilzAdam: Can you review it? 00:16 PilzAdam ShadowNinja, sry, Im busy with other things currently 00:23 ShadowNinja It's hard to get commits merged when you have to bug devs at the right time (when two are available) to even get them looked at. 00:26 kahrl I know but there's hardly anything that can be done about that 00:26 kahrl except pay someone to be in the right state of mind nine to five, I guess? 00:27 ShadowNinja Well we do have donations, but I doubt they would be enough. ;-) 03:28 kaeza it seems the place_schematic function has some problems when emerging/generating non-existing chunks 03:29 kaeza sometimes there are lighting errors, sometimes the schematic seems to be partly overwritten by the terrain (or not placed at all) 03:29 ShadowNinja kaeza: Same with VM. hmmmm said it was unfixable. 03:29 kaeza ah 03:30 ShadowNinja Wait, maybe not. 03:30 ShadowNinja Do you mean some blocks are entirely air? 03:31 kaeza well, dunno what it is, but sometimes the ground "sinks" 03:31 kaeza (is that the proper word?) 03:32 VanessaE mud reflow. 03:32 kaeza even in a flat map? 03:33 kaeza hrm 11:32 BlockMen ok, here is what i have done at formspecs so far: you can set a backgroundcolor for each formspec, this backgroundcolor can be drawn over the full screen if wanted, the background-image can be cliped to formspec-size (with offset), the color of item slots can be defined (also different color for mouse hover state), a bordercolor can be drawn 11:33 BlockMen https://github.com/BlockMen/minetest/compare/formspec_rework 11:37 nore BlockMen, that looks a good idea 11:40 Vakadavr|2 hi all 11:42 BlockMen nore, thx 11:43 BlockMen also, here is example code (http://pastebin.com/5LwX8hhH) and image(http://i.imgur.com/1vDRTQv.png) to get inventory like this http://i.imgur.com/UnrmevT.png 11:44 Vakadavr|2 is there anybody who could help me with minetest unitests? 11:44 kahrl Vakadavr|2, simply ask your question 11:45 Vakadavr|2 In debug version of minetest TestSocket test is failed 11:46 kahrl Vakadavr|2, it is the same problem as https://github.com/minetest/minetest/issues/793? 11:46 kahrl is it* 11:47 Vakadavr|2 unfortunately no 11:48 Vakadavr|2 minetest failed to process bind function 11:49 Vakadavr|2 exception thrown in socket.cpp:355 11:49 kahrl does netstat show anything listening on UDP port 30003? 11:54 Vakadavr|2 one second, i'll check it 11:55 Vakadavr|2 nope, there is no record with 30003 port 11:56 kahrl Vakadavr|2, strange 11:56 Vakadavr|2 and i already turn of all firewalls =\ 11:57 Vakadavr|2 what's even stranger that in release all ok 11:57 kahrl release doesn't run the unit tests unless you pass --enable-unittests on the command line 11:59 kahrl BlockMen, didn't somebody make a pull request with a config[] element for options that concern the entire formspec? 11:59 BlockMen kahrl, i dont know of one. 12:00 Vakadavr|2 yes i know, but if unit test failed shouldn't game run with errors too? ) 12:00 BlockMen but beside that celeron55 said i should make it more specific, like bgcolor[] 12:00 Vakadavr|2 and sorry if my english is terrible 12:01 kahrl Vakadavr|2: your english seems fine 12:01 kahrl Vakadavr|2: the game itself uses a different port, which may be the reason 12:02 celeron55 it's probably just some sloppines in timing or throughput or something that the unit tests don't like 12:03 celeron55 or maybe there's some ipv4/6 problem 12:04 kahrl Vakadavr|2: what's your OS? 12:04 Vakadavr|2 win 8 12:06 Vakadavr|2 by the way the error is "156: Bind failed: no such file or directory" 12:07 kahrl I guess it shouldn't use strerror(errno) there 12:08 kahrl something with WSAGetLastError() instead 12:12 Vakadavr error n.10049 "Unknown error" 12:13 Vakadavr wait 12:14 Vakadavr msdn: "Cannot assign requested address." 12:14 Vakadavr from here http://msdn.microsoft.com/en-us/library/windows/desktop/ms740668(v=vs.85).aspx 12:17 kahrl makes me think your winsock somehow doesn't support ipv6 12:17 kahrl but does that ever happen in windows 8? 12:19 kahrl aha! http://forums.codeguru.com/showthread.php?349275-Bind-to-*any*-IPv6-address 12:19 kahrl if you don't zero sin6_flowinfo and sin6_scope_id, it apparently something gives WSAEADDRNOTAVAIL 12:20 kahrl this problem has been showing up in valgrind for forever but it never seemed to cause any actual problems... until now 12:20 kahrl s/something/sometimes 12:23 Vakadavr yes! zeroing helps me =) 12:23 Vakadavr you are hero ) 12:23 kahrl nope just a valgrind devotee :P 12:24 Vakadavr it does not matter ) 12:25 Vakadavr so one error gone and one appear ) 12:26 Vakadavr some trouble with sending package but i will put this problem off until evening 12:26 Vakadavr thank you all for your help 12:29 Vakadavr btw should i post some bugreport or something else? 12:30 kahrl I'll definitely fix it anyway 12:31 kahrl but not until the bind_address pull request is merged 12:31 Vakadavr ok 15:20 ShadowNinja Reproducable segfault. Load a singleplayer world that crashes after loading, then load the world again. (Try something like minetest.after(1, function() minetest.crash() end)). Full backtace: http://ix.io/7tM 15:25 celeron55 run it in valgrind 15:29 proller https://github.com/proller/minetest/compare/heat https://github.com/proller/minetest/compare/liquid63 15:35 ShadowNinja bind_address is rebased over master now. 15:57 ShadowNinja I am trying valgrind, but I don't think the world is going to finish loading this hour... (The one that causes this error is my mod world, including technic.) 16:01 PilzAdam ShadowNinja, I can not reproduce that 16:02 ShadowNinja I also noticed another bug where loading stalls at "Item definitions" sometines. 16:02 ShadowNinja PilzAdam: It might have something to do with how I an crashing it or the mods. 16:03 PilzAdam I tried assert(0 == 1) and indexing a nil value 16:03 PilzAdam fun fact: assert(0) does not crash, 0 is true in Lua 16:04 * ShadowNinja knew that 16:04 ShadowNinja Loaded! 16:05 ShadowNinja FPS=0 We should have support for fractional FPS. 16:20 ShadowNinja I guess I got the "Item definitions" freze on this run... 16:27 ShadowNinja Alright, guess I can't reproduce it in valgrind because of the slowness, and things like this should provide more details: ERROR[main]: Something failed 16:29 sapier Pilzadam 0 isn't exyctly true but boolean evaluation is defined as nil and false beeing false while everything else isn't false ... of course result is same 16:54 proller https://github.com/proller/minetest/tree/next -- test branch with some unmerged pulls 17:17 sapier1 http://postimg.org/image/rvqrkqenr/ this is how the modmanager is meant to look like any comments to this? 17:19 thexyz I think the description should be where "Depends" now is, also you don't really need a list of dependencies, just make it a string of deps separated with ", " 17:20 sapier1 if I'm gonna make a string next one having 1o depencencys will tell "this is looking ugly" ... so what about removing the deps completely? 17:21 BlockMen sapier1, looks good. but why the gap at depends? i would use the same gap like at the image 17:21 thexyz maybe 17:21 sapier1 yes BlockMan I realized same after posting 17:23 thexyz offtopic, but why don't you use freetype? 17:24 sapier1 but I can't fix the real modmanager problem as long as kahrl doesn't finish his http async work ... what I planed to do prior he started this task is about 60-90% conflicting depending on his actual implementation 17:24 sapier1 I guess because I havent installed the dev headers ;-) 17:25 sapier1 and of course using different fonts will disort the gui ... so I want at least default to look ok 17:25 thexyz oh 17:25 PilzAdam the default is freetyep 17:25 PilzAdam *type 17:25 thexyz yes, I think it should be freetype 17:26 thexyz I guess most distributions package Minetest with this flag enabled 17:26 sapier1 if it's default it should at least post a warning on cmake 17:26 thexyz so there really no point in using it without freetype (unless you're using different font) 17:26 PilzAdam "default" as in: all releases have it 17:28 sapier1 "default" is git to some ppl but I don't want to argue again its useless to discuss to some ppl 17:28 thexyz "default" is default, not "default to some ppl" 17:28 sapier1 btw cmake doesn't tell anything about freetype or not so I can't tell if I have freetype enabled or not 17:28 PilzAdam "default" is not defined, thats why I said what I meant 17:29 PilzAdam sapier1, cmake . -LH | grep FREETYPE 17:29 sapier1 ok it's off 17:30 sapier1 and I have freetype headers so minetest default is to turn it off unless you manually enable it 17:31 PilzAdam changing freetype and gettext to automatically be enabled when they are present is on the TODO list 17:32 sapier1 sane decision ... at least if it's done ever 17:32 sapier1 but back to the modmgr gui 17:32 sapier1 depends or not or something completely different? 17:37 sapier1 anyone to approove this pull request? https://github.com/minetest/minetest/pull/640 17:39 sapier1 iqualfragile are you online? 17:39 PilzAdam sapier1, why have you closed https://github.com/minetest/minetest/pull/696 ? 17:39 iqualfragile sapier1: yes i am 17:39 nore sapier1, I would, but I am no core dev... ;) 17:39 iqualfragile i was gone for a week, came back today 17:39 nore 696 was good, too 17:39 sapier1 because this needs to be redone as it's go a lot of code duplication and I don't have time to do it 17:40 sapier1 no it's not good nore as of code style it's crap 17:40 PilzAdam sapier1, then write "Needs to be redone" instead of "Nobody is interested" 17:40 sapier1 I can't uploade mobf due to filesize limitation iqualfragile what's the file limit? 17:40 sapier1 it's same 17:41 iqualfragile thexyz: ^ 17:41 sapier1 I don't have any interest in that function anymore and others wont push it ... I won't gonna delete it as I never delete anything so it'll be there if someone wants to use it 17:43 sapier1 iqualfragile? :-) whats size limit? :-) is it mentioned somewhere? 17:44 sapier1 Fun fact with freetype enabled I seem to be the only contributor to be visible on credits page without scrolling 17:45 sapier1 as of reading did anyone hear of jeija recently? 17:47 thexyz iqualfragile: what should size limit be? 17:48 thexyz MOBF.SAPiER.r00 … 17:48 sapier1 minetest doesn't support multifilemods by now 17:48 thexyz sapier1: how much do you need? 17:49 sapier1 it's 8.4 mb right now but I'd not have a problem to split it up as it's a modpack I just need to know what limit is set 17:50 PilzAdam thats almost 6 times bigger than minetest_game 17:51 thexyz filesize limit is being decided right now right here! 17:51 thexyz 100 mb should be fair, right? 17:51 thexyz oh, no, 4k texture packs… 17:51 sapier1 mobf contains lots of models minetest_game doesn't 17:52 sapier1 is there a way to set per user filesize limits? 17:52 PilzAdam thexyz, for mods 100 mb should be enough 17:52 thexyz okay 17:52 sapier1 e.g. ppl that have legitimate reason to add bigger mods/texturepacks need to ask for increasing their limit first? 17:52 iqualfragile sapier1: nobody heard of jeija, he just disappeared some while ago 17:53 thexyz you think someone will need more than 100 mb? 17:53 nore I heard him say he preferred real electronics and was not interested in minetest anymore 17:53 iqualfragile yes, 100 mb should be enought for any mod, especialy as mods should try to integrate with the default texturepack 17:54 iqualfragile hmm… he could at least said goodbye 17:54 sapier1 I'm fine with 100mb too ... maybe this is even too big for mods ... but I guess it's more common for texture packs 17:58 proller sapier1, ! auto name creating broken! 17:58 proller when you try connect to server without name 17:58 sapier1 what happens? 17:58 proller error message 17:59 sapier1 that's fine? 17:59 proller no, name must be auto created like Guest1234 17:59 sapier1 oh you're not talking about the servername 18:00 proller no, player name 18:00 proller with fresh install 18:00 sapier1 hmm it should create a name itself unless you delete it on purpose 18:01 proller 1583 menudata.name = std::string("Guest") + itos(myrand_range(1000,9999)); 18:01 proller but doesnt work 18:01 PilzAdam these guest autonames should be removed 18:02 proller PilzAdam, why ? 18:02 PilzAdam proller, https://forum.minetest.net/viewtopic.php?id=6213 18:03 PilzAdam basically people should choose a nick themselves, to identify them 18:03 sapier1 lol 1583 is plain wrong it should set playername not menudata.name 18:04 proller ok, but it must show error with input for name 18:05 sapier1 of course not a big deal to add it 18:09 proller or not error, just window with input, and continue to join game on press ok 18:10 thexyz sapier1: should work now 18:16 sapier1 https://github.com/minetest/minetest/pull/884 18:23 sapier1 thexyz thx it's working 18:39 Sokomine hmm. rollback_check either takes (almost) forever, crashes my client, and/or crashes the server... 18:40 Sokomine any new developments wouldn't even help there - not without redcrab returning :-(