Time Nick Message 01:43 VanessaE I'd like to propose a guideline for the Mod Releases section: Any mod that is hosted on some kind of git repo should include a "Browse the code" link that points to the repo itself, in addition to any links that point to the tar/zipball download. 06:26 kahrl of the 4 commits in omnicleanup, the farmesh removal is pretty much standalone 06:26 kahrl so my plan is to merge that to upstream right now (if anyone approves) 06:26 kahrl https://gist.github.com/kahrl/6218360 06:27 kahrl then I'll squash the 3 other commits and write up a summary 07:12 sfan5 kahrl: seems fine 08:42 iqualfragile ow noez, farmesh! did it died? 08:42 VanessaE yup 08:42 VanessaE evidently so 08:43 iqualfragile hmm… was broken anyways 08:44 iqualfragile but it would be nice to have if somebody would add that idea of sea of meme: reducing the voxel-resolution when the voxels are far away 08:45 VanessaE someone did point to a blog post a while back about that 08:45 VanessaE wish I could remember who and when (and what post :P ) 08:45 iqualfragile moment 08:47 VanessaE oh hell, it probably was that very one :P 08:49 iqualfragile multiple ones: http://sea-of-memes.com/LetsCode64/LetsCode64.html http://sea-of-memes.com/LetsCode67/LetsCode67.html http://sea-of-memes.com/LetsCode71/LetsCode71.html http://sea-of-memes.com/LetsCode72/LetsCode72.html 08:49 VanessaE '72' was the one 08:56 VanessaE interesting that the guy who did this seems enamored with Larry Niven.. :) 08:57 iqualfragile this blog is quite interesting 11:09 celeron55 did anything come out of that project? 11:09 celeron55 i was left with the impression that it basically died once the author played without with his rendering tricks 11:09 celeron55 s/without/enough/ 11:28 proller sdl 2.0 ! 11:59 psedlak hi, today i pulled after few days, tried to play with the freetype option ... but i'm unable to compile ... "error: cannot allocate an object of abstract type ‘irr::gui::intlGUIEditBox’" 12:00 psedlak full error part and cmake opts are at http://www.fpaste.org/31755/63950861/ 12:01 psedlak i didn't pulled/updated irrlicht for a long time, so i tried now but it didn't helped :/ 12:01 psedlak so can anyone help/point me to some direction? thanks :] 12:12 celeron55 what version of irrlicht did you have, and what version of irrlicht do you have now? 12:12 celeron55 and did you get the same error on both 12:14 psedlak celeron55: not sure about previous version ... master github branch from around Mar 11 12:14 psedlak celeron55: current version i've just cloned master from github 12:14 psedlak from git://github.com/zaki/irrlicht.git 12:15 psedlak celeron55: and yes it was same error 12:17 psedlak heh is it possible that it somehow does not uses my specified irrlich version but the one from system? even when i specified those "IRRLICHT_*DIR" ...? 12:20 psedlak in system i have 1.7.3 ... should that be ok or ...? 12:21 celeron55 well minetest doesn't support unreleased versions of irrlicht 12:21 celeron55 you probably had something newer than 1.8 before and something newer than 1.8 now 12:21 celeron55 both of which are unreleased 12:21 celeron55 1.7.3 will work fine 12:24 psedlak celeron55: ok, thanks will try that :) 16:13 rubenwardy I think server and client should be merged and called "multiplayer", and should display the client tab with a button called "start server" which opens up the server window. 16:27 iqualfragile sfan5: und? schöne beschreibung ausgedacht? 16:28 sfan5 iqualfragile: falscher channel; ja 16:29 Jordach ^ wrong channel 16:29 PilzAdam Jordach, thats what "falscher channel" means 16:32 * celeron55 imagines pronouncing that word with some kind of a completely ridiculous sound at the middle 16:35 Jordach fuh-al-shh-uh <- is falscher 16:42 Akien /falʃə/ :) 17:07 sfan5 Jordach: sounds like "fall shoe" to me 17:10 * kahrl wrote a novel... https://github.com/minetest/minetest/pull/867 17:12 Jordach > If you plan to: what is wrong with you? 17:12 Jordach i'm reading that diff 17:13 Jordach holy... 17:19 celeron55 kahrl: lol 17:20 celeron55 "nobody can review thousands of lines of code" "okay i'll write thousands of lines of explanation instead" 17:21 Exio4 haha 17:32 celeron55 kahrl: i couldn't find anything that i would be against 17:33 celeron55 let's wait some more eyeballs and put it in soon 17:51 nore_ did any of you had that mainmenu bug where clicking on a world would start it, and when exiting, the world that was previously selected was selected again? 17:52 nore_ I am planning to fix that, and I will make a pull request soon 17:53 nore_ what I intebd to do is to remove that doubleclick event, and when a click is made on a world already selected, that world will be started 17:55 nore_ I will also make a pull to fix the formspec escapes, and one to add a readonly textarea 17:56 nore_ what do you think of those? 18:17 kahrl "when a click is made on a world already selected, that world will be started" 18:17 kahrl nore_: isn't that how it is done currently? 18:18 kahrl (with a timer of a maximum of 500ms between selections) 18:18 kahrl the problem being that when the formspec is created, it calls setSelected and that counts as a "click" 18:24 celeron55 so the problem is that irrlicht's GUI system sucks by any standards 18:25 kahrl you, sir, are correct 18:25 kahrl ;) 18:25 celeron55 i'd rather code for windows 2.0 18:27 kahrl I also don't think a readonly textarea is possible without some hacks, or a completely custom editbox 18:27 kahrl that being said, we already have a custom editbox: intlGUIEditBox 18:27 celeron55 i hate it that i didn't make a custom GUI system right from the start 18:27 kahrl it's not used in GUIFormSpecMenu though 18:27 celeron55 now i don't dare to, because it'd need unicode support from the ground up and all that 18:28 celeron55 (it would've been fine if it would have grown over time) 18:28 kahrl right, and all that delicate formspec stuff you'd have to try not to break 18:29 celeron55 if someone exposed some 2D and text drawing and mouse/keyboard input primitives to main menu lua, maybe we could have one that way? 8) 18:31 kahrl sapier would be sad that another thing of his would be partially reverted 18:31 celeron55 can't care about that 18:32 celeron55 also, it's iteration 18:33 celeron55 sapier is the pioneer of menu improvements no matter whatever comes out of this in the end 8) 18:35 PilzAdam I was bored so I made this: https://github.com/PilzAdam/minetest/commit/6188fd8b073c0603265ef7f1f6637e00fa0cadf7 18:35 PilzAdam its this helpful or complete nonsense? 18:35 PilzAdam *si 18:35 PilzAdam *is 18:36 celeron55 that's not how it should be done 18:36 celeron55 you need to move it over the old file with a filesystem operation 18:37 kahrl the temp file should probably be in the same directory as the final file 18:37 kahrl (rename() is only atomic when renaming on a single file system) 18:37 celeron55 yes 18:38 PilzAdam I previously used file+"~" 18:38 PilzAdam is that safe? 18:38 celeron55 maybe something like +"~mt" 18:39 kahrl put a "." in front as well, maybe? 18:39 celeron55 some text editors use ~ for their temporary files and overriding them could be nasty 18:39 kahrl "."+file+".~mt" 18:40 iqualfragile kahrl: wow, its realy a novel rest: sorry for the wrong channel :) 18:42 celeron55 PilzAdam: implementing the filesystem level moving of files appreas to be fairly simple; on posix it's just rename(from, to) 18:42 celeron55 windows tends to have something similarly simple 18:42 kahrl fun fact: try checking out my omnicleanup branch and run "git diff HEAD^ HEAD|wc -l" 18:44 celeron55 actually... is that a C standard library function 18:45 celeron55 looks like it is 18:45 kahrl celeron55: even if it exists on windows I wouldn't trust that whatever C library we're compiling with does it in an atomic way 18:45 PilzAdam for some reason it fails when I add a "." in front of it 18:47 PilzAdam well, "." + file + ".~mt" doesnt work since file is the full path... 18:48 celeron55 on windows there's MoveFileEx with MOVEFILE_REPLACE_EXISTING 18:48 kahrl celeron55: I don't think that's atomic, either 18:48 celeron55 but that appears to be atomic only with good luck 18:48 celeron55 8D 18:48 kahrl MoveFileTransacted? 18:49 kahrl there's also ReplaceFile which does some funky backup stuff, probably not atomic 18:49 PilzAdam "If newname names an existing file, the function may either fail or override the existing file, depending on the specific system and library implementation." 18:49 PilzAdam for rename(old, newname), that is 18:50 celeron55 kahrl: MoveFileTransacted doesn't work on XP 18:50 celeron55 also that seems somehow specific to NTFS, or is it? 18:50 kahrl call GetProcAddress and if it fails, fall back to MoveFileEx? 18:52 celeron55 PilzAdam: on posix rename() will override it always when it has permissions and other required things 18:52 celeron55 ...why must the C standard library of windows be such a toy?= 18:53 celeron55 it's like they intentionally made it completely useless 19:28 PilzAdam this should work for now: https://github.com/PilzAdam/minetest/commit/85e0187d730397a5c3b3be5b36c796ef7f68ec73 19:30 sapier I won't be sad about mainmenu rewrite but be sure I will be as pedantic as all others have been ;-P 19:32 sapier I know irrlichts gui is quite limited so if anyone volonteers to write a better gui toolkit I'd be glad to help by telling what is still not as "good" as irrlicht was ... but be sure this is not a 1-2 month task as formspec mainmenu has been rather a 3-6 months job ;-) 19:33 sapier full time ... aka 8h a day 19:34 sapier btw I just claim online modstore ;-) that's been the reason for mainmenu rewrite everything else was unwanted but neccessary 19:35 celeron55 PilzAdam: call it "safeWriteToFile()" and you're done 19:37 sapier wait ... whats line 705??? 19:37 kahrl don't the stringstreams need std::ios::binary? 19:37 kahrl and why stringstream instead of ostringstream? 19:38 celeron55 sapier: look up how windows rename() works 19:39 sapier maybe we should use move rename delete in this case? 19:39 kahrl i'd make the parameter const std::string &content 19:39 sapier yes it's paranoid ... 19:39 celeron55 sapier: we aren't making a nuclear power plant controller 19:40 celeron55 all of kahrl's points are good 19:40 celeron55 also const std::string &file, and actually const std::string &filename (because file sounds like it's some kind of a file handle) 19:40 sapier I guess we support winxp? 19:40 celeron55 i've preferred "path" though 19:40 celeron55 sapier: of course 19:41 kahrl unlike MS, we do 19:41 sapier my company doesn't support winxp any longer and we support very very old stuff 19:41 sapier so it's not "of course" ;-) 19:41 sapier as of win2008 there seems to me a more suitable rename 19:42 celeron55 sapier: please don't repeat the full irc log from two hours ago 19:42 celeron55 actually one hour ago 19:42 sapier sorry didn't realize this already was discussed 19:43 sapier at least it's way better than before :) 19:45 nore_ sapier, why is formspec escape still bugged? 19:45 sapier because noone reported additional errors 19:46 nore_ I can't make a field with ; even with a \ before it 19:46 sapier because ; never will be an allowed character 19:46 nore_ and I reported errors, but it looks like it was not seen 19:47 nore_ why not allowed? if it is escaped, it should be ok? 19:48 nore_ if it is not allowed, could you make a read-only textarea with the same properties as normal textareas? 19:48 sapier formspec uses different separators evaluated at different stages 19:48 nore_ like in main menu 19:48 nore_ and? 19:49 sapier ";" wasn't ever supported and I didn't add support for it as priority was to support old features 19:49 nore_ if the code splits without removing the escapes, and removes those right before using them, it could work 19:50 nore_ I mean, after all splits are done 19:50 sapier let me have a quick look at ";" separation code to not tell anything wrong 19:51 nore_ I was able to fix that changing the code of split so that it always looks for escapes 19:52 sapier it's a parameter you should use it 19:52 nore_ it did not look like it produced bugs 19:52 sapier did you test all 100000 cases? 19:52 nore_ so I dont know why there is still a parameter 19:53 nore_ no, of course 19:53 sapier so you are just guessing it doesn't cause bugs 19:53 nore_ but I did not see formsoec corruption 19:53 nore_ in those I looked at 19:54 sapier exactly what I told you don't know ... you may be right but you may be wrong too 19:54 nore_ i may be 19:54 nore_ but what about those readonly textareas? 19:54 nore_ could it be done, like in mainmenu? 19:56 sapier correct place to add your changes would be the parse functions ... but this will fail for a size element ... but there's no reson to escape ";" in a size element 19:56 sapier size element doesn't use split 19:56 sapier what do you mean with readonly textareas? 19:57 sapier everything supported in mainmenu will work ingame too 19:57 sapier except the enginge.get..index 19:58 nore_ yes, but textareas are readonly in aminmenun and not rezdonly in game 19:58 nore_ but I have a request 19:59 sapier I'm surprised it's exactly same code why should this behave different? 19:59 nore_ make it configurable wheter return will or not close a formspec 19:59 nore_ well, i think it is that, but i should check first 20:00 sapier I guess you don't use exactly same formspec textareas are drawn different dependent on what fields are set 20:00 sapier yes it's dirty I wouldn't have written it that way but formspec was there before I did a lot of cleanup for mainmenu 20:01 nore_ no, it is same fields filledd 20:01 sapier looking at code you may be right that the split parameter isn't really neccessary 20:01 nore_ just checked in in Tp code 20:01 PilzAdam celeron55, kahrl, https://github.com/PilzAdam/minetest/commit/c51af61f5cac21f7fc9766ce36192b06f6518f62 20:01 sapier guess it's purpose vanished due to other cleanups 20:02 sapier can you post your non read only formspec? 20:03 kahrl PilzAdam: does RollbackManager::flush still append? 20:04 PilzAdam oh 20:04 nore_ no, i understood... in TP code I use the label field 20:04 celeron55 lol, good catch 20:04 celeron55 that shouldn't be used at all in rollback 20:04 PilzAdam Ill write a safeAppendToFile 20:04 celeron55 don't 20:04 sapier textarea[X,Y;W,H;;;] will result in a read only textarea 20:05 celeron55 the rollback file can be enormous, what do you think you would do there? 20:05 celeron55 you can't copy it by any expectation 20:05 nore_ but can $ be used in that too? 20:05 nore_ or formspec escape 20:05 celeron55 just don't change it at all 20:05 nore_ because I need to put text with ; 20:06 PilzAdam ok 20:06 PilzAdam anything else? 20:06 sapier if you need ";" you first need to add ";" escape support remove the parameter make escaping default and do some testing if it really behaves correct 20:07 sapier don't forget to fix the lua function removing ";" 20:07 nore_ nopd, formspec_escapes replace ; with \; 20:07 nore_ so it is already good 20:08 kahrl PilzAdam: rest looks good 20:08 sapier iqualfragile_ do you remember making moddb not sending empty version elements to json clients? 20:08 nore_ but without parameter, I had good results with fields 20:09 sapier nore_ if you do some testing just issue a pull request ... of course not ignore the parameter but really remove it 20:09 nore_ yes, of course 20:09 nore_ I will do it soon 20:10 nore_ and about world selection, if I change it so that CHG and DCL events are handled the same way 20:10 sapier what do you mean with same way? 20:11 nore_ but if the world being clicked on is the same as previously selected, then start the game 20:11 nore_ it could fix a few bugs 20:11 nore_ like starting the world you click on with a single click 20:11 sapier you need to save the previously saved world localy and honor tabs too 20:12 sapier -saved + clicked 20:12 nore_ and not chabging selected world when youbexit 20:12 nore_ no need to save it, setting_get 20:12 nore_ if =, then start, else setting_set 20:13 sapier be sure that won't break anything there have been a lot of issues with current world not beeing saved correctly 20:13 sapier btw I don't like the idea a selected world is beeing started by a single click 20:14 sapier but of course thats a personal preference 20:14 nore_ I am doing that tomorrow, right now I am on a tablet, so no git :( 20:14 nore_ sapier, why? 20:14 sapier if you click a selected world minutes after it was selected first it still starts the world 20:15 nore_ if world is already selected, and you click again, it is that you want to play... 20:15 nore_ and you can use the configure button 20:15 sapier i know this is smartphone default behaviour but not what older ppl aka 30+ are used to 20:16 sapier we are used to starting doubleclick selecting singleclick (thanks to microsoft ;-)) 20:16 PilzAdam celeron55, now you saved 20$ :-) 20:16 nore_ well, I never use doubleclick on my computer... 20:17 sapier ask some other ppl about their opinion first maybe it's fine the way you suggest 20:18 sapier completely different topic does anyone have a suggestion how server and client could be merged without adding tab in tab? 21:04 sapier << asking again for 418 (640) and 774 ? 21:08 PilzAdam https://github.com/PilzAdam/minetest/commit/7e401fcd7588ccebe706029c4ca28aac68a566c5 21:08 PilzAdam this makes env damage (like falling or lava) visible to others 21:09 sapier is 1735 compatible to old servers? 21:09 PilzAdam of course 21:10 PilzAdam it was always read, just the variable was commented out 21:11 sapier oh yes true 21:11 sapier why? 21:11 PilzAdam someone probably got annyoed by the "unused variable" warning 21:12 sapier I wonder how things like that could slip in with everyone beeing that pedantic ? :-) 21:12 sapier but -1 for merge for completely unrelated reasons ;-P 21:13 sapier it's a cosmetic only fix while other really usefull fixes are stalled 21:14 PilzAdam its labeled as "bug": https://github.com/minetest/minetest/issues/303 21:14 sapier blablabla ... sorry thats bullshit you can mark every cosmetic issue as bug 21:18 sapier I'm really pissed noone tells whats wrong with 418 640 and 774 for months now but their not merged either ... I really suspect this is due to not merging them will give simplemobs an advantage as you merge anything instantly that simplemobs benefits of 21:19 PilzAdam wut? I already said yesterday that 418 and 640 are good 21:19 PilzAdam and we had a discussion about 774 21:19 sapier that has been said weeks ago 21:19 sapier even not counting 774 21:20 PilzAdam one core dev agreeing is not enough to merge, though 21:20 sapier no other core dev agreed to stepheight fix 21:20 sapier so remove it instantly or behave consistent 21:21 sapier the only reason you added it was you assumed it'd fix your simplemobs jump issues 21:23 sapier sometimes I really understand why others left community .... 21:24 sapier celeron55 kahrl thexyz darkrose RBA proller sfan5 any second oppinion to 418? 21:25 PilzAdam oh, there was indeed nobody else agreeing on stepheight 21:25 PilzAdam (just looking at the logs) 21:26 sapier I know I was surprised you instantly added it until you said you need it for simplemobs 21:26 PilzAdam have I brought it up earlier? 21:27 PilzAdam dont want to search the logs now 21:27 sapier imho your behaviour is not correct about those things adding patches should be done by a at least somhow neurtal point of view .. I miss this when you add things 21:28 sapier maybe I'm biased but that's my opinion 21:29 proller i'm +1 for 418 21:29 PilzAdam I guess I brought the stepheight pull request up earlier, and nobody said anything about it 21:29 proller it can be used with carts? 21:29 sapier it can be used with any entity 21:29 proller or boat 21:30 PilzAdam proller, cant be used with boats or Simple Mobs, they make it exactly the other way round 21:30 PilzAdam they apply velocity based on yaw 21:30 PilzAdam and carts already do it, they dont need an engine feature for that 21:30 sapier carts do it manualy? 21:31 sapier primary target entities are accelerated ones 21:31 PilzAdam they only have 4 directions, so I can just call set_yaw() when that changes 21:31 proller btw i know how to kill server with carts 8) 21:31 sapier try monorail proller ;-) 21:32 PilzAdam sapier, how does the yaw behave when I call set_yaw() in Lua, is it posible that it changes while the Lua code runs? 21:34 sapier if I remember correctly its a client only thing 21:34 PilzAdam it isnt 21:34 PilzAdam its done in content_sao too 21:34 proller .. and with simplemobs + flying top of them - 2 crushes 21:35 proller not crush, but 100% cpu use 21:35 sapier oh true ... some time has passed since i did that ... no it's singlethreaded 21:35 sapier so no conflict between lua api and this one 21:35 PilzAdam good 21:36 PilzAdam (then I can set that flag in simple mobs too, so I merge it now :-p) 21:36 sapier but enabling automatic and manually setting same time will not give the result you may have wanted of course 21:37 sapier autoupdate is done first so if you change velocity in lua this will take effect next step 21:43 proller sapier, monorail is not on github 8( 21:43 sapier of course not as it's not called carts noone uses it ;-P 21:46 sapier I don't even know if it's compatible to current mt have been busy with mainmenu for some weeks 21:46 PilzAdam wait, wasnt there something with M_PI and MSVC? 21:46 sapier windows devs? 21:47 proller for using M_PI need to make include 21:47 sapier which one? 21:47 PilzAdam #include "util/mathconstants.h" 21:48 PilzAdam lets rewrite history 21:52 PilzAdam sapier, about 640: http://irc.minetest.ru/minetest-dev/2013-08-12#i_3258596 21:53 sapier I'm sure i fixed the serchup I check the other thing 21:54 PilzAdam there are 2 other things 21:55 PilzAdam the function name and whether it should be part of ServerEnvironment 21:56 sapier I won't merge all thre of them where else? map? 21:56 sapier I don't care about that 21:58 proller why to rewrite history? 21:58 proller fix and make next commit? 21:58 PilzAdam meh, that just floods the commit log with useless stuff 21:59 proller who reads logs? 21:59 proller you make game or logs? 21:59 PilzAdam ehm, who doesnt read the logs? 22:01 proller adding 100+ commits maybe bad, but what problem with one? 22:01 sapier pilzadam it's in serverenvironment for historic reasons as it's lua function was in environment too 22:04 PilzAdam hmm... it runs from the bottom to the top, so it will returns caves instead of the actual surface 22:04 sapier a cave has a surface too 22:05 sapier it will find the first surface within search area 22:05 PilzAdam I would expect it to return the surface of the landscape 22:05 sapier that's not the usecase for this function 22:06 sapier there seems to be another way to get that information either 22:06 sapier I don't exactly know how someone mentioned it ... but as it's not the intended function of get_surface I don't remember it 22:08 sapier getSurface to map? 22:11 PilzAdam btw, lua-api.txt should say minetest.get_surface() 22:11 sapier ok so map class or environment? 22:12 PilzAdam dunno 22:12 sapier I don't care any other opinions? 22:17 PilzAdam sapier, whats the status on translation for the menu? 22:17 sapier nothing changed 22:17 kahrl if basepos.Y + searchup > 0x7fff and none of the mapnodes in the column are loaded (more generally, there is no surface), get_surface is an infinite loop 22:18 sapier true 22:20 sapier is there a max s16 macro available at posix and windows? 22:20 kahrl maybe just replace it with while ((searchup--) && ...) { ... } 22:21 sapier I thought about s16 max = MYMIN(searchup + basepos.Y,0xFFFF); 22:21 sapier but I'd prefere that macro to 0xFFFF 22:21 kahrl max s16 is always 0x7fff 22:21 sapier thats the reason why I'd prefere the macro ;-) 22:22 sapier kahrl do you prefere server environment or map? 22:23 kahrl map 22:24 kahrl for me, conceptually, map is more about the nodes and environment is more about the players & objects 22:25 Exio4 PilzAdam: what do you think about the omnicleanup? 22:25 Exio4 same for sapier 22:25 sapier I don't care about concepts anymore they change more often than I adopt to the last one 22:25 kahrl we could #define MAX_REGISTERED_CONTENT 0x7fff, and then you could use that define :P 22:26 kahrl I don't think irrlicht defines any MAX_S16 or something like that 22:27 kahrl not sure if std::numeric_limits::min() works everywhere, but it's kind of ugly 22:28 sapier is there any way to access gamedef from within map? 22:28 kahrl oh 22:28 kahrl just pass the nodedef, that's how isOccluded does it 22:29 kahrl or actually no, there is m_gamedef 22:30 kahrl sorry, forgot that isOccluded is a static function 22:30 sapier sso I'd need to pass whole gamedef to a map function? I don't think this is right 22:31 kahrl no, just do m_gamedef->ndef() 22:31 sapier still not much better 22:31 kahrl seriously? 22:32 sapier wait 22:33 sapier forget about it gamedef is available in map ... thanks to minetests missing data encapsulation ... why not make everything global? 22:33 kahrl you just asked whether you can get the gamedef and now you're complaining about it? 22:34 sapier I didn't expect it there because it's not at all sane to be there ;-) 22:36 kahrl the map has to know the node defs to properly propagate light and stuff 22:36 sapier everyone needs to know about everything in minetest ;-) 22:36 kahrl gamedef is not everything... mostly just what used to be static tables in 0.3 22:39 sapier take "everyone" and "everything" as too many things need to know too much 22:47 Tesseract PilzAdam: BanManager::load(): Couldn't write file <-- Should be save()? 22:47 PilzAdam seems so 22:48 PilzAdam its not my fault, though ;-) 22:48 PilzAdam Ill fix an error in lua-api.txt too and push that 22:49 kahrl oh wow, was that one still there 22:49 kahrl I believe I mentioned it here a while before I left 22:52 sapier what are the new lag estimate functions for? 22:55 sapier https://github.com/minetest/minetest/pull/640 fixed and rebased 23:10 Tesseract I can confirm this: https://github.com/minetest/minetest/issues/853 23:11 kahrl i've seen it before, too 23:12 Tesseract sfan5: This can be closed, both me and kahrl confirm that it is not so: https://github.com/minetest/minetest/issues/726 23:12 PilzAdam it was totally wrong a while ago and celeron55 fixed it a bit 23:13 kahrl Tesseract: hm? I don't remember testing that 23:13 Tesseract kahrl: You said that it is done in such a way that the proper player is recorded. 23:14 Tesseract (I forget the terminology) 23:14 kahrl oh, well that was from reading the code 23:14 kahrl probably 23:14 kahrl it's easy to make mistakes 23:15 Tesseract Either way it is recorded, even if it uses the nearest player aproximation. 23:15 kahrl it should record the one who used the bucket, otherwise I'd say it is a bug 23:16 Tesseract Job for set_node_as(pos, node, name) 23:17 kahrl well no, the rollback system already does that, assuming it works 23:17 kahrl I haven't tested it though, is what I'm saying 23:24 PilzAdam is anyone against rasing liquid_loop_max to 10000 again? 23:26 Exio4 the thing should be; why was it reduced as first 23:27 Exio4 just applied to my local minetest https://github.com/WilliamBundy/minetest/commit/228ec338393d3312f12a94769b8cf6a5bb9e6568, took a whole minute to recompile 23:27 Exio4 amazing 23:30 PilzAdam Exio4, ah, now I see why you want that omnicleanup, you heard "better compile time" :-) 23:30 Exio4 :D 23:30 Exio4 unless you buy me a better cpu & decent cooling system, of course!