Time Nick Message 03:53 octacian #5627 should be good to go. I've made all requested changes. And apologies for being unable to make them sooner, I am not around until the late evening on Saturdays. 03:53 ShadowBot https://github.com/minetest/minetest/issues/5627 -- Menu: Merge singleplayer and server tabs by octacian 04:04 octacian I'd also really appreciate seeing game#1592 in before release, though I guess it's unlikely considering feature freeze tomorrow. 04:04 ShadowBot https://github.com/minetest/minetest_game/issues/1592 -- Add workbench and nametag to rename items by octacian 06:31 Krock https://github.com/minetest/minetest/commit/90808a 06:31 Krock wat 06:32 Krock !tell nerzhul in the recent commit 90808a you also merged the enabling/disabling CSM PR by accident (?) 06:32 ShadowBot Krock: O.K. 06:39 Zeno` 404 07:14 Krock oh, only copied 6 digits. should be 90808a4 07:16 nerzhul #90808a4 07:16 nerzhul :( 07:16 nerzhul it's too late to force push, i should finish my test on the PR and close it this morning... 07:30 nerzhul Krock, i think we cano now merge #5627 07:30 ShadowBot https://github.com/minetest/minetest/issues/5627 -- Menu: Merge singleplayer and server tabs by octacian 07:31 nerzhul can* 07:31 nerzhul are you okay ? 07:34 Krock testing.... 07:48 Krock is there any simple way to change the mainmenu to android mode? 07:51 Krock nvm, found it 07:52 Krock it has its own code, so looks good to me 07:53 Krock LGTM, can be merged 07:54 nerzhul nice merging it 07:54 nerzhul i'm finishing the test on the mods enable/disable but i can't find how preview mod is not loaded whereas CSM is enabled 07:58 nerzhul oh... i'm stupid i modified the minetest.conf in my home instead of run in place 07:59 nerzhul it works as intended 08:05 nerzhul i'm pushing a gitignore fix to ignore clientmods/mods.conf and client/mod_storage when we are in run in place, to prevent them to be commited 08:40 red-hunter-002 #5791 08:40 ShadowBot https://github.com/minetest/minetest/issues/5791 -- Prevent fall damage from overflowing when falling from a large hight. by red-001 08:51 nerzhul red-hunter-002, lgtm 08:51 nerzhul in next release we really should fix all those variables which are typed differently it's a pain, example health which is u16 or s16 server side... 08:52 red-hunter-002 ^ 08:53 Krock to make this somewhat safe, (u16)MYMIN(damage_f + 0.5, 255) could be used 08:54 nerzhul yeah red-hunter-002 can you do the Krock point ? 08:54 nerzhul better than rewriting the variabl 08:54 nerzhul does someone test #5767 ? 08:54 ShadowBot https://github.com/minetest/minetest/issues/5767 -- Don't add damage flash while punch texture modifier is active by stujones11 08:57 red-hunter-002 done 09:00 nerzhul Krock, #5747 has a last chance to be merged too :) 09:00 ShadowBot https://github.com/minetest/minetest/issues/5747 -- [CSM] Add send_chat_message and run_server_chatcommand API functions by paly2 09:02 DS-minetest sending chat messages and chatcommands to server would be nice. it would also give a way to communicate from client to server 09:04 nerzhul DS-minetest, don't use it as a message channel for CSM, you will just polute your chat :p 09:04 nerzhul except if you have a mod to consume them, but wait for next release, we will provide a channel bridge for communications 09:04 DS-minetest ok 09:05 nerzhul i will rewrite the whole chat protocol with break to be better adding message types, messages sender, channel ids, etc 09:05 DS-minetest but it would be nice for making a teleporting interface 09:05 nerzhul :o 09:05 nerzhul i don't see why but maybe :p 09:05 DS-minetest and something like that 09:06 nerzhul merging #5791 in ~10 mins 09:06 ShadowBot https://github.com/minetest/minetest/issues/5791 -- Prevent fall damage from overflowing when falling from a large height. by red-001 09:07 nerzhul our float serializaiton is just shit... why we x 1000.0f 09:08 nerzhul it's why in debug build this breaks 09:08 nerzhul 1000 * 2293670 overflows signed float, gg kwoelr 09:09 nerzhul it's ridulous we cannot send a float or store it to map if exceed 2293670... or -2293670 09:09 nerzhul celeron55, do you know who decided to do that with floats ? 09:12 nerzhul which platform does shit with float like this ? Pentium 1 ? 09:12 nerzhul i never see a such bad and problematic implementation 09:35 DS-minetest does minetest_game still use the old setting functions?! 09:38 DS-minetest at least the creative mod does 09:38 DS-minetest it bad 09:41 nerzhul !tell paramat for release can you fix the settings interfaces ? 09:41 ShadowBot nerzhul: O.K. 10:05 DS-minetest 13 times are the old setting functions used in MTG 11:20 Fixer since tabs were merged I need to translate few things, but I can't find some words in weblate, not updated? 11:21 Fixer like "Local Game" "Play Online" etc 11:24 nerzhul weblate is updated 1 time per day i think 11:25 red-hunter-002 nerzhul, doesn't a script need to be run on the repo? 11:25 red-hunter-002 https://github.com/minetest/minetest/commits/master/po/minetest.pot 11:26 red-hunter-002 https://github.com/minetest/minetest/blob/master/util/updatepo.sh ? 11:29 nerzhul there is a script to run, but i don't remember if weblate do it automaticly or not to have proper diff, because updatepo.sh change the whole files... each time 11:46 celeron55 nerzhul: decided? it's really just a random hack from the days of minetest was just a... random test 11:47 celeron55 s/of/when/ 11:47 nerzhul oh and do you remember why it was introduced ? 11:47 nerzhul because nowadays it seems it's just a cause of many bugs reported in our serialization on large float values 11:47 celeron55 well, i guess maybe for player positions 11:47 celeron55 players were drawn as trollfaces back then 11:50 nerzhul very strange... i should re-test a hacked version without this "fix" 11:51 celeron55 fixing float serialization might be a good candidate for a 0.5 release 11:51 celeron55 should probably have fixed it for like 0.2 but eh, it worked, why bother 11:53 nerzhul yeah, i hope it's not used for map coordinates... 11:57 nerzhul okay it's hopefully only used over network 11:57 nerzhul just removing the calcul or divide by 1.0f should help 11:58 nerzhul celeron55, is this due to a cross platform problem ? 12:10 nerzhul nice Krock 5747 remaining and we will have an interesting CSM preview for this release 12:12 Krock 50 messages in the queue? that seems huge 12:13 Zeno` maybe we should serialize using fixed point 12:14 Zeno` 0.56 --> serialize --> de-serialise -> == 0.56 really is flaky 12:15 nerzhul Zeno`, what do you mean ? in fact i don't know what is the fixed point, is this the C representation of a float ? here fixed point is value * FIXEDPOINT_FACTOR which is 1000.0f 12:15 Zeno` nerzhul, the situation is the same as comparing two floats for equality 12:16 nerzhul can you be more precise ? 12:16 Zeno` good joke 12:16 nerzhul lol 12:16 Zeno` but if it's *1000f and then /1000f when deserialising then everything should be fine. Just don't expect the floats to *always* be exactly the same 12:17 Krock review done 12:18 Krock (5747) 12:18 nerzhul ty 12:19 Krock np 12:23 celeron55 nerzhul: it just was a surefire way to make float serialization work for literally anything 12:26 Krock apparently it doesn't work for "literally" everything :P 12:26 nerzhul okay, maybe using 1.0f can be nice to prevent problems, or i should look at portable network serialization for float to have the real float, not a fake value :) 12:27 nerzhul merging #5732 in ~10mins 12:27 ShadowBot https://github.com/minetest/minetest/issues/5732 -- [CSM] Add function to get the definition of nodes and items by bigfoot547 13:05 celeron55 well, for every platform, not for every value 13:06 celeron55 i guess it'd be just fine to grab the ieee 754 value directly from memory 13:06 celeron55 every architecture uses that anyway 13:07 nerzhul 144 pr opened it's nice, it should be stable before release now :) 13:56 Fixer never forget blinking https://i.imgur.com/pckqtLq.png 14:21 Fixer nerzhul: you need to run that script to update it it seems, if est31 understanded my correctly 14:22 nerzhul yeah i think i should do it 14:22 nerzhul i will do it now then 14:22 nerzhul hmm no 14:22 nerzhul before i should import new translations, run po and export 14:24 Fixer nerzhul: i can't do new translation, because some words are not in that po database 14:24 Fixer related to tab merge 14:25 nerzhul yes i understand but i'm doing other stuff atm 14:25 Fixer ok 14:30 Fixer no worries 14:46 rubenwardy you should wait until just before the release to merge translations 14:47 Krock but some kind of update is required, no? Otherwise the guys on weblate don't know what to translate 14:47 rubenwardy oh, the other way 15:06 nerzhul there is not problem with translations 15:29 nerzhul pushing a lint fix and i will look at merging translations from weblate after that 15:32 Krock nerzhul, https://github.com/minetest/minetest/pull/5747/files#r117631282 15:33 nerzhul hmmm can't remember but there is a thing with the timer and allowance, it's mathematicly in the calcul for the message rate server side 15:34 Krock that's why comments are helpful sometimes. Prevent cargocult :P 15:38 nerzhul yeah i tend to comment, but it seems this line is not commented :p 15:38 nerzhul https://github.com/minetest/minetest/blob/master/src/remoteplayer.cpp#L199 15:38 nerzhul translations pushed 15:38 nerzhul looking for this updatepo in our log 15:39 nerzhul it seems it's run just after updating translations by est, found that on 14 dec 15:39 nerzhul and 30 Aug 15:40 nerzhul i'm pushing updatepo to permit translators to help us with last translations 15:41 Fixer nice, ty 15:42 nerzhul i know understand how est31 increase so high his commit lines xD 15:42 red-hunter-002 lol 15:44 VanessaE we have some kind of network problem with current HEAD, users timing out (even me) -- AFTER connecting 15:45 VanessaE like connect....receive media...nodes.... then time out before it's done 15:45 VanessaE rather than connect...wait...dead, like would normally be the case. 15:46 rubenwardy are you able to bisect?# 15:46 VanessaE I can but this is a production server. however, I did update last night, from 9b5effff to 60baf812 15:47 VanessaE and this morning is when I was informed that it was brokenb 15:47 VanessaE -b 15:47 VanessaE the servers are now at dfa0c15c 15:47 VanessaE and it's still acting the same 15:48 VanessaE and these timeouts are from users whom I know for a fact can reliably play on my servers 15:48 Fixer could be this? https://github.com/minetest/minetest/commit/ce9802266ef1def339ec2e119c59090d0fd07c90 15:48 Fixer Various code cleanup & little performance improvement on HTTP download 15:48 VanessaE you know, that might be. 15:48 VanessaE I *do* use HTTP and it DOES time out during the media transfer 15:49 nerzhul Fixer, i just changed map to unordered_map 15:50 nerzhul i tested on VanessaE server this pr and that works, strange, we should verify if map should be ordered (strange) 15:50 VanessaE like I said, it times out during the "Media" step or during "initializing nodes", NOT during the "connecting" step 15:51 VanessaE and it seems to do it when there are two or more users connecting from the same IP 15:51 VanessaE almost as if the engine gets confused as to who gets which packets? 15:52 VanessaE once the user is in, it seems to work fine, btw 15:53 VanessaE so something broke between 9b5effff and 60baf812 15:53 VanessaE or at least that's the commit range I have 15:59 VanessaE THIS is why I said there needs to be a much longer feature freeze, btw. 15:59 VanessaE invariably some bug gets into the engine just before release, and not enough time to work it out 15:59 * VanessaE rolls her server back to 9b5effff 15:59 rubenwardy we have two weeks of freeze 16:00 rubenwardy if there are unfixed blockers, then we extend it 16:00 rubenwardy I think that a better solution would be to start doing patches with only bug fixes 16:00 rubenwardy but: overhead 16:01 VanessaE the problem, rubenwardy, is that "blockers" sometimes get un-marked as such 16:01 VanessaE and milestones get skipped/delayed 16:01 VanessaE because OH G*D WE MUST RELEASE NAOW! 16:01 VanessaE :P 16:08 VanessaE ok 9b5effff is not it, unless it's client-side too 16:08 VanessaE I am reliably timing out on my server 16:09 * VanessaE goes back to the last pull prior to that one 16:09 Fixer VanessaE: try compiling 3526740 16:09 VanessaE why? 16:09 Fixer VanessaE: one before Various code cleanup & little performance improvement on HTTP download ( 16:09 VanessaE 57eaf62c6 is my next one actually 16:10 VanessaE and if not that one, 73de17afa 16:10 VanessaE those are the last two unique ones in `git log -g` 16:14 VanessaE great 16:14 VanessaE you guys fucking broke it and now I can't roll back to a known good commit 16:14 VanessaE 57eaf62c6 16:14 VanessaE oops 16:14 VanessaE 2017-05-21 18:13:26: ERROR[Main]: ServerError: AsyncErr: environment_Step: Runtime error from mod 'mesecons' in callback environment_Step(): ERROR: MapBlock format not supported 16:14 VanessaE that. 16:14 VanessaE (that is, however, the commit that error occurred at) 16:16 * VanessaE is majorly pissed off 16:16 sfan5 you can't go backwards over mapblock version upgrade "boundaries" 16:16 sfan5 this is not new 16:16 VanessaE yes I know that 16:16 sfan5 just something unfortunate when testing or bisecting 16:16 VanessaE but a bump occurred without adequate testing of the networking code 16:17 VanessaE so now my servers all time out users who I know for a FACT can reliably play 16:17 sfan5 in which commit? 16:17 VanessaE and I can't roll back to a non-breaking commit 16:17 VanessaE see above, I already said which is which, or at least what looks right 16:17 VanessaE I can't tell anymore 16:18 VanessaE I fear this release is going to be a clusterfuck just like the last one 16:24 VanessaE and...timed out again. 16:24 VanessaE (I had to put my servers back to HEAD, just so the map and mods work...and I haven't updated the mods in over a month) 16:26 red-hunter-002 couldn't you try and revent to last good comit and then reapply the mapblock version changing commit? 16:27 Fixer i don't think there were mapblock changes recently... last one was in Feb2017 iirc 16:28 VanessaE red, I tried rolling back to what I thought was the last good commit, but apparently it wasn't. 16:29 ShadowNinja nerzhul: You broke weblate :-| http://sprunge.us/IWWD 16:29 VanessaE so I went back to the build prior to that one. nope, mapblock version mismatch 16:29 VanessaE ShadowNinja: offtopic. 16:29 VanessaE this is 1000000x more important than some stupid translations 16:29 VanessaE (sorry) 16:30 ShadowNinja Eh, you can continue with whatever you're discussing. 16:34 nerzhul ShadowNinja, it's repaired 16:34 nerzhul i inverted two steps :) 16:35 nerzhul VanessaE, you failed to connect to which server please ? 16:35 nerzhul i need to test it myself 16:35 VanessaE nerzhul: ve-survival 16:35 VanessaE daconcepts.com 30001 16:36 VanessaE connecting with just one client usually works, but two or more from the same IP (I have users whose brother/sister also gets on at the same time) and one of them will likely time out 16:36 VanessaE during the "initializing nodes" stage 16:36 nerzhul i don't see why it can be related to the http std::unordered_map as it's TCP not udp there 16:36 * VanessaE shrugs 16:37 nerzhul i'm connecting two clients to test 16:37 VanessaE all I know is what my users tell me, and that I could reproduce it myself (my machine can handle at least 5 clients at a time) 16:37 nerzhul maybe i should drop my media cache 16:37 nerzhul i have 2 clients connected atm it works fine 16:39 VanessaE I suggest you just keep signing on/off until it barfs 16:40 nerzhul wow, gray world 16:40 VanessaE ther eit is 16:40 VanessaE there it is. 16:40 nerzhul with a connection timeout 16:40 VanessaE you're going to time out shortly 16:40 VanessaE yep 16:40 nerzhul but media are downloading, we are stuck after loading 16:40 VanessaE mmhmm 16:41 nerzhul i'm stuck when entering world 16:41 nerzhul don't known why only server logs can show that 16:41 VanessaE there's nothing in the server logs about it 16:42 VanessaE (but they're not running in 'verbose' mode) 16:42 VanessaE (because the damned logs are already bloated enough in the normal setting) 16:42 VanessaE nerzhul: you want an account on daconcepts.com so you can run a test server maybe? 16:43 VanessaE or I guess that won't matter 16:43 kilbith is has to do with mapblocks? 16:43 kilbith *it 16:43 VanessaE dunno 16:43 kilbith try reverting 071e114ffa maybe 16:44 VanessaE what's that one? 16:44 kilbith this one brought troubles with old worlds 16:44 VanessaE private node metadata? 16:45 kilbith yes 16:45 VanessaE that's the one that brought the map version bump with it isn't it? 16:45 kilbith I don't know but I could join my old world anymore with that 16:45 kilbith *couldn't 16:46 VanessaE sfan5: ^^^ 16:49 kilbith http://irc.minetest.ru/minetest-dev/2017-05-13#i_4919002 16:55 VanessaE ^^^ so it also stops new clients from loading old worlds? 16:55 sfan5 this doesn't make sense 16:55 VanessaE which part, sfan5 ? 16:56 sfan5 there part where that commit breaks anything 16:56 sfan5 at all 16:56 sfan5 the* 16:56 * VanessaE shrugs 16:56 VanessaE try connecting to my server 16:56 VanessaE see what it does for you 16:57 sfan5 which in in particulatr 16:57 VanessaE daconcepts.com 30001 16:57 VanessaE try with at least two clients at a time 16:58 VanessaE keep disconnecting and re-connecting one or both and it will eventually start timing you out for no reason 16:59 VanessaE from the point of view of others in-game, you never actually make it "in" if you "time out" while it's in the initializing nodes stage 16:59 sfan5 that doesnt sound like it would be related to private nodemeta 16:59 Fixer prior = page up ? 17:00 VanessaE Fixer: I think so. 17:01 VanessaE Fixer: yes. 17:01 VanessaE (so says xev(1) ) 17:01 Fixer ty 17:02 sfan5 kilbith: can't reproduce the failure to load old worlds 17:03 kilbith naturally, it's your feature and my error report doesn't make sense at all 17:05 sfan5 ? 17:16 Fixer how old should be the world? 17:56 Fixer i've tried to load kraydyfi world, it worked 17:56 Fixer maybe you mean when you go back and then try to load world played in newest engine 18:02 Fixer src/guiKeyChangeMenu.cpp - want to throw up (looks awful with long words) 18:10 VanessaE but have you tried connecting to mine? 18:11 VanessaE I mean, this is just vanilla engine, on good hardware, with a good network connection, and a mod collection that, while huge, doesn't drag the server down/ 18:18 Fixer nerzhul: btw, there is another encoding related problem, tooltip in cyrillic (Shift+M) looks like this: https://i.imgur.com/EUJ3vp4.png 18:18 Fixer $) 18:20 Fixer can confirm weblate was update and works with newest changes 18:20 Fixer was updated 20:16 Wuzzy Hi. I am currently translating Minetest. Am I correct here to ask string questions? 20:16 red-hunter-002 I guess 20:18 Wuzzy what exactly is the “needs_fallback_font” used for? 20:18 Wuzzy I know this is a special string but I forgot what it is used for 20:18 nerzhul will merge #5747 before feature freeze 20:18 ShadowBot https://github.com/minetest/minetest/issues/5747 -- [CSM] Add send_chat_message and run_server_chatcommand API functions by paly2 20:19 Wuzzy Too bad the .po file has NO source comment on this, this really should be fixed IMO 20:19 Wuzzy nerzhul: feature freeze not yet in place? yay 20:19 VanessaE nerzhul: fix the timeouts problem first! 20:19 VanessaE Wuzzy: tonight 11:59pm (not sure what time zone, UTC?) 20:19 Fixer Wuzzy: if you look into weblate UI - it has link to source file 20:20 red-hunter-002 Wuzzy, which file? 20:20 Fixer Wuzzy: look to the right 20:20 Wuzzy ... 20:20 Wuzzy I just downloaded the file, no comment for needs_fallback_font there 20:20 Wuzzy German .po file 20:20 Wuzzy this can't be right, right? 20:21 Wuzzy Also, IIRC, Weblate shows source comments in web interface anyway 20:21 red-hunter-002 https://hosted.weblate.org/translate/minetest/minetest/de/?checksum=&offset=&q=needs_fallback_font&search=substring&source=on&target=on&type=all&ignored=False 20:22 red-hunter-002 Source string location: src/fontengine.cpp 20:22 Wuzzy ... 20:22 Wuzzy i was hoping for any explanation 20:23 Wuzzy what is the difference between fallback font and normal one? 20:23 Wuzzy which languages are supposeed to use the fallback font? 20:23 nerzhul VanessaE, feature freeze != bugfix freeze :p 20:23 red-hunter-002 https://github.com/minetest/minetest/blob/e10e5fd16c2bd55b9b106edc6cedd07530cf103d/src/fontengine.cpp#L70 20:23 red-hunter-002 ??? 20:23 red-hunter-002 wtf is that code doing 20:23 Wuzzy e.g. does Greek need a fallback font? 20:23 VanessaE nerzhul: features < bugfixes. :) 20:24 red-hunter-002 Wuzzy, just leave it 20:24 Wuzzy red-hunter-002: at least there should be a comment for translators, dont u think? 20:24 red-hunter-002 it's some sort of ugly hack 20:24 Wuzzy .po format supports comments 20:25 red-hunter-002 Wuzzy, sadly the person that added this didn't leave a comment 20:25 Wuzzy i know this field expects the value "yes" or "no" but translators probably dont know 20:25 Wuzzy i know this "needs_fallback_font" from other projects 20:25 Wuzzy well, i guess i'll open an issue about that one 20:26 red-hunter-002 Wuzzy, the code feels wrong 20:26 Wuzzy otherwise we will see literal translations pouring in which we dont want :D 20:26 Wuzzy e.g. "braucht_fallback_schrift" xD 20:26 red-hunter-002 lol 20:26 red-hunter-002 well that would be a no accoring to is_yes 20:27 red-hunter-002 according* 20:27 red-hunter-002 well this is a 3 year old mess 20:27 Wuzzy code feels wrong? You just invented EDD - Emotionally-Driven Development. :D 20:32 Fixer oh schrift is german word 20:33 Fixer sounds exactly like we say it too 20:33 Fixer VanessaE: please report that issue, it will be reviewed before release 20:34 VanessaE no it won't. 20:35 Fixer VanessaE: post an issue 20:37 Wuzzy What is “Terrain alt noise”. Or more specifically, why is it called that way? Why “alt”? Such a strange name. 20:38 Wuzzy Is “alt” a slang word? I don't really find it in my dictionary 20:38 VanessaE it means "alternate" 20:38 Wuzzy okay 20:38 Wuzzy Why not writing “alternate”, then (at least in the UI, I don't care about internal naming). 20:38 VanessaE (at least that's what it *usually* means, if not necessarily in this context) 20:39 Wuzzy Note: I refer to mgv7_np_terrain_alt 20:39 Wuzzy This is a setting described as “Y-level of lower terrain and lakebeds.” 20:39 Wuzzy still doesnt really explain the “alt” however 20:40 red-hunter-002 altitude 20:40 red-hunter-002 altitude 20:40 Wuzzy okay 20:41 Wuzzy but “alternate” would actually make sense in this context as well :& 20:41 Wuzzy I'm inclined to suggest a renaming 20:41 Wuzzy Just to be sure: Feature freeze != string freeze? 20:41 rubenwardy correct 20:41 Wuzzy phew 20:42 Wuzzy because there are a few strings i want to be rewritten :) (nothing serious, however) 21:08 nerzhul there are many new things to translate on this release, impressive 21:09 nerzhul it was a good idea to run updatepo.sh 21:16 nerzhul i added some Fr translations to weblate 21:16 nerzhul see you tomorrow feature freeze in less than 3 hours, i don't think there will be new feature added to minetest :) 23:33 Wuzzy hey all 23:33 Wuzzy anyone wanting to merge / review a last-minute PR? 23:33 Wuzzy :D 23:33 Wuzzy https://github.com/minetest/minetest/pull/5795 23:35 Wuzzy nerzhul: This may be interesting for you 23:35 Wuzzy above PR makes status messages (“Chat shown”, etc.) translatable 23:36 Wuzzy rubenwardy: interested in last minute PR review? :D