Minetest logo

IRC log for #minetest-dev, 2015-07-31

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

All times shown according to UTC.

Time Nick Message
00:17 paramat joined #minetest-dev
00:22 crazyR joined #minetest-dev
00:24 crazyR_ joined #minetest-dev
01:15 paramat left #minetest-dev
01:30 BrandonReese joined #minetest-dev
01:31 crazyR joined #minetest-dev
01:33 paramat joined #minetest-dev
01:34 paramat might push later #2976
01:34 ShadowBot https://github.com/minetest/minetest/issues/2976 -- Dungeon generation: Fix code style issues in dungeongen.cpp by paramat
01:49 paramat hi VanessaE please can you check this PR game#598 i suspect it is wrong
01:49 ShadowBot https://github.com/minetes​t/minetest_game/issues/598 -- Fix mese block alias by CraigyDavi
02:55 est31 hmmmm, can you check https://github.com/est31/minetest/commit/​7217fd32eb8af01d3012a2ed883598c6a9fbf78e
02:55 est31 !title
02:55 ShadowBot Fix 0.4.11 githash for android · est31/minetest@7217fd3 · GitHub
03:01 hmmmm errm sure, looks good
03:06 OldCoder joined #minetest-dev
03:09 Player_2 joined #minetest-dev
03:32 zat joined #minetest-dev
04:31 Taoki[mobile] joined #minetest-dev
04:55 paramat joined #minetest-dev
05:01 paramat hi hmmmm #2976 is possibly ready to go if you need to review, perhaps i should reduce line lengths to well below 90, some lines are at 90
05:01 ShadowBot https://github.com/minetest/minetest/issues/2976 -- Dungeon generation: Fix code style issues in dungeongen.cpp by paramat
05:01 est31 paramat, fyi the code style guidelines say "lines with 90 characters or above wont be tolerated"
05:01 paramat yeah i just realised that =)
05:01 est31 :)
05:02 paramat i was working to a limit of 90, when it should be a few over 80
05:03 est31 well, 89 is ok
05:03 paramat lol
05:03 est31 but yea 80 is the soft limit.
05:04 est31 but i'm not one of those who think that a line with 86 chars needs a break, imo we dont need a soft limit at all.
05:22 leat1 joined #minetest-dev
05:29 paramat ugh i find <90 is the point where code get chopped up to an extreme and ugly degree, but will go ahead =/
05:32 leat1 joined #minetest-dev
05:36 crazyR joined #minetest-dev
05:36 crazyR joined #minetest-dev
05:42 leat1 joined #minetest-dev
06:13 paramat 86 columns, will push it tomorrow
06:14 Hunterz joined #minetest-dev
06:33 nore joined #minetest-dev
07:10 Darcidride joined #minetest-dev
07:13 Darcidride_ joined #minetest-dev
07:18 nrzkt joined #minetest-dev
07:21 Krock joined #minetest-dev
07:35 est31 ohhhhh
07:35 est31 dammit
07:35 est31 if what i think is right, its really bad
07:35 est31 lemme see whether I can confirm
07:36 est31 dammit
07:36 est31 i am right
07:37 est31 well, i am proud to announce that I've found a way to speed up the android re-build!
07:37 est31 very good for developers
07:39 nrzkt est31 => android fan :p
07:42 Calinou joined #minetest-dev
07:57 nrzkt but if you found a way this is good. I hope at a moment we could switch to cmake
08:01 est31 yea this would unify things
08:01 est31 e.g. look at commit at master
08:01 Yepoleb_ joined #minetest-dev
08:01 est31 only needed because we have separate logic for android and rest of build system
08:01 est31 perhaps one day I'll write cmake based compile
08:03 nrzkt :)
08:10 est31 nrzkt, can you have a look at https://github.com/est31/minetest/commit/​649e351e2a464f92a7fc26f5ee761cb87a9a1a48
08:11 nrzkt +1
08:12 err404 joined #minetest-dev
08:12 est31 thanks, pushed.
08:13 swaaws joined #minetest-dev
08:26 sfan5 left #minetest-dev
08:26 sfan5 joined #minetest-dev
09:09 bluegreen joined #minetest-dev
09:29 Niebieski joined #minetest-dev
09:29 Niebieski I'm finding it hard to find a server with a particular mod on the servers list.
09:29 Niebieski is there something we can do about it ?
09:31 Niebieski Probably client sided.
09:31 nrzkt http://servers.minetest.net
09:31 est31 Niebieski, also you can get the raw json
09:31 est31 and parse it yourself
09:31 est31 ebieski@197.38.23.183) has joined
09:31 est31 err
09:31 est31 http://servers.minetest.net/list
09:32 Niebieski Nice, thanks.
09:34 Niebieski Wow, didn't know that my IP is just showing up for everyone. xD
09:36 bluegreen are you running a server?
09:37 Niebieski No...
09:37 est31 Niebieski, ask on #freenode on how to get a mask
09:38 Niebieski Alright.
09:46 Niebieski joined #minetest-dev
09:54 sfan5 Niebieski: you need a freenode account with NickServ to get a mask
09:55 H-H-H joined #minetest-dev
09:55 Niebieski Hmmm. I see.... I know this is akward but anyone tried to visit my IP ? anything showed up ? xD
09:55 sfan5 connection refused
09:57 Niebieski Alright thanks.
10:04 FR^2 joined #minetest-dev
10:12 Niebieski joined #minetest-dev
10:23 bluegreen joined #minetest-dev
10:26 kilbith joined #minetest-dev
11:04 bluegreen joined #minetest-dev
11:10 proller joined #minetest-dev
11:12 crazyR joined #minetest-dev
11:12 crazyR joined #minetest-dev
11:12 Calinou http://phoronix.com/scan.php?page=news​_item&amp;px=SQLite-Release-Micro-Opts
11:12 Calinou will we update the bundled SQLite?
11:16 kilbith sqlite ain't bundled anymore afaik
11:21 proller joined #minetest-dev
11:21 be3weefderfg Is luajit going to be bundled ?
11:26 crazyR_ joined #minetest-dev
11:26 crazyR_ joined #minetest-dev
11:32 bluegreen_ I think I fixed the bug that happens when water meet an unloaded block
11:32 bluegreen_ Should I open a pr?
11:33 nore bluegreen_: yes
11:38 err404 joined #minetest-dev
11:55 kilbith seems like a bug with the forum's signature : https://lut.im/FljoUsEi/rtNVgW9y
11:55 kilbith ^ c55
11:56 bluegreen_ also the reset password doesn't send an email
12:17 err404 joined #minetest-dev
12:20 crazyR joined #minetest-dev
12:27 alket joined #minetest-dev
12:55 crazyR_ joined #minetest-dev
12:55 crazyR_ joined #minetest-dev
12:56 Megaf joined #minetest-dev
13:00 alket joined #minetest-dev
13:04 EvergreenTree joined #minetest-dev
13:07 crazyR joined #minetest-dev
13:07 crazyR joined #minetest-dev
13:10 Wayward_One celeron55: it seems the FAQ on forum.minetest.net returns a 403 error
13:14 H-H-H joined #minetest-dev
13:42 Warr1024 joined #minetest-dev
13:48 Zeitgeist_ joined #minetest-dev
14:08 bluegreen_ joined #minetest-dev
14:13 WSDguy2014 joined #minetest-dev
14:17 H-H-H joined #minetest-dev
14:27 bluegreen joined #minetest-dev
14:28 bluegreen #2977
14:28 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
14:30 Niebieski joined #minetest-dev
14:34 alket_ joined #minetest-dev
14:50 TBC_x http://sprunge.us/ijDK list of [OS: OpenGL version; graphic renderer] from http://feedback.wildfiregames.com/report/opengl/ filtered, removed entries which had only one user or less
14:50 est31 joined #minetest-dev
14:50 est31 nrzkt, next android build system fix: https://github.com/est31/minetest/commit/​302109f4fb303179d1bcee2581c41bb8f18be664
14:50 est31 can you have a look at it?
14:51 est31 new version here: https://github.com/est31/minetest/commit/​975685db351d2f06ee054d084232af4aec93c3f7
14:52 hmmmm joined #minetest-dev
14:53 est31 hmmmm, case you don't read logs, can you check https://github.com/est31/minetest/commit/​975685db351d2f06ee054d084232af4aec93c3f7
14:54 nrzkt est31: please use awk instead of grep + 2 pipes !
14:54 est31 nrzkt, can you give a good awk alternative?
14:55 est31 I accept it if its shorter and also general
14:55 hmmmm just make sure that $${VERSION_FILE_NEW} can't possibly be / or /usr or something really bad
14:55 hmmmm whenever I see "rm ${something}" I get nervous
14:56 blaze joined #minetest-dev
14:57 Niebieski joined #minetest-dev
14:57 blaze joined #minetest-dev
14:59 bluegreen #2977
14:59 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
15:01 est31 hmmmm, well, the only "generated" value in $${VERSION_FILE_NEW}
15:01 est31 is ROOT
15:01 est31 and that's the output of `shell pwd`
15:02 est31 I'm pretty sure that one escapes spaces in file names, if needed.
15:04 WSDguy2014 left #minetest-dev
15:05 est31 hmmmm, seems you are right
15:05 est31 pwd doesnt escape
15:05 est31 the rm needs ""
15:07 nrzkt est31: example => awk '/^set\(VERSION_MAJOR .*\)/{print sub("/)","",$2)}' CMakeLists.txt
15:08 nrzkt 4 commands merged into 1 :)
15:11 est31 its shorter, will integrate it.
15:11 est31 also same generality
15:11 est31 except one space
15:11 est31 will add it
15:11 nrzkt with awk you can parse a whole file, here it's a "trivial" usage, but you can script all the awk string to do things :)
15:12 est31 e.g. if given set(VERSION_MAJOR 31)#test your code will print 31#test
15:12 est31 yea
15:12 Hunterz joined #minetest-dev
15:12 est31 old behaviour replaced the ) with " "
15:12 nrzkt you are right, the idea is to get .* value but i can't remember how to do it, my awk is a little bit old :p
15:12 est31 but your script is easy to adjust
15:13 est31 nrzkt, doesnt work bte
15:13 est31 btw*
15:13 nrzkt i need to go, man powa
15:14 est31 awk: line 1: regular expression compile failed (missing '(')
15:14 bluegreen can someone please take a look at  #2977
15:14 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
15:17 est31 The code style needs adjustment
15:17 est31 braces {} on same line, except for methods
15:17 est31 so write } else { etc
15:17 est31 also space between ) and {
15:18 est31 otherwise, I don't know liquid code.
15:18 est31 perhaps the mapgen guys paramat or hmmmm can have a look at it.
15:20 bluegreen well the code works
15:20 hmmmm looks good from a functional standpoint, but what's the point of NodeNeighbor ignore?
15:21 bluegreen none really I geuss
15:21 bluegreen I will make it a bool
15:25 enesbil joined #minetest-dev
15:27 TBC_x @P=split//,".URRUU\c8R";@d=split//,"\nrekcah xinU / lreP rehtona tsuJ";sub p{ @p{"r$p","u$p"}=(P,P);pipe"r$p","u$p";+​+$p;($q*=2)+=$f=!fork;map{$P=$P[$f^ord ($p{$_})&6];$p{$_}=/ ^$P/ix?$P:close$_}keys%p}p;p​;p;p;p;map{$p{$_}=~/^[P.]/&& close$_}%p;wait until$?;map{/^r/&&<$_>}%p;$_=$d[$q];sleep rand(2)if/\S/;print works too
15:27 TBC_x but can you tell what does it do when you read it ?
15:28 hmmmm just another perl?
15:28 TBC_x yes
15:28 hmmmm looks like an overcomplicated fork bomb
15:28 bluegreen what is that?
15:28 TBC_x a code that works
15:29 est31 hehe
15:29 bluegreen wow
15:29 est31 yea it works alot
15:30 TBC_x that's why we have standards for coding style
15:31 bluegreen Is it only me or is compiling minetest on linux faster?
15:32 TBC_x it's not just compiling
15:32 bluegreen what is it also?
15:33 TBC_x destroying your hard drive for example
15:33 est31 TBC_x, bash is an incredibly powerful language. Ever heard of the donald e knuth pascal story?
15:33 est31 or sh
15:34 * TBC_x was playing with plan9 lately
15:34 TBC_x that thing started in two seconds
15:35 TBC_x on raspberry pi
15:36 kilbith left #minetest-dev
15:37 bluegreen I updated #2977
15:37 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
15:37 est31 well, perhaps send some prayers to systemd and our freedesktop.org priests will deliver us with a quickly hacked and feature incomplete reimplementation of plan9, make their whole setup depend on it by hard, ready to be deployed on linux distros within the next 2 weeks.
15:37 TBC_x what's with his pascal story?
15:37 kilbith joined #minetest-dev
15:40 est31 http://www.leancrew.com/all-thi​s/2011/12/more-shell-less-egg/
15:44 nrzkt joined #minetest-dev
15:47 est31 "starts in 2 seconds" reminds me of one of the reasons I had liked systemd once.
15:47 est31 ok updated pr here https://github.com/est31/minetest/commit/​76ff6ebd04bd98e663ac5ea077b326aa2e5f9186
15:48 TBC_x hehe, me too
15:49 bluegreen installing linux on a hard drive grained a whole new meaning http://spritesmods.com/?art=hddhack&amp;page=7
15:49 sfan5 maybe someone should run minetest on their hdd
15:50 bluegreen an create a bug report if i crashes
15:51 nrzkt est31: systemd boot is user experience, but for sysadmin experience, managing and creating service is proper, it's a definition file and not a ugly bloated shell
15:52 nrzkt i like tomcat systemd file
15:52 nrzkt i want to kill the tomcat init.d script creator
15:52 sfan5 lol
15:52 sfan5 are we arguing about systemd in here now too=?
15:53 est31 nrzkt, btw the error is awk: line 1: regular expression compile failed (missing '(')
15:53 bluegreen we should argue about in in ##windows
15:53 nrzkt where is lenart ?
15:53 nrzkt with my command ?
15:53 TBC_x hdd clone protection?
15:53 est31 there is an actual freenode channel for windows?
15:53 est31 lol
15:54 Calinou yes, unofficial of course
15:54 Calinou you won't see Satya_Nadella there 8)
15:54 est31 are they accustomed to type the "forward" slash all the time?
15:54 Calinou (heh, imagine if Windows was a regular FLOSS project)
15:54 est31 /join #windows etc
15:54 Calinou no, they type \join ##windows :P
15:54 est31 ok enough trolling :)
15:54 bluegreen reactos.org
15:54 est31 nrzkt, yes with your command
15:54 nrzkt can you paste it, i change my computer :)
15:55 TBC_x wouldn't that hdd clone protection break cd burner software?
15:55 est31 nrzkt> est31: example => awk '/^set\(VERSION_MAJOR .*\)/{print sub("/)","",$2)}' CMakeLists.txt
15:55 nrzkt thanks
15:55 nrzkt on which system are you running it ?
15:56 est31 ubuntu
15:56 bluegreen TBC_x why?
15:56 bluegreen whould it breck cd burner solfware
15:57 TBC_x isn't *.iso supposed to be read sequentally?
15:57 bluegreen I don't think an iso would take up a whole sector of a hard disk
15:58 est31 sectors are 4kb, no?
15:58 est31 or 8 k
15:58 est31 depending on disk
15:58 TBC_x depends
15:58 bluegreen your right
15:58 nrzkt 4k is the maximum atm
15:59 nrzkt both in SATA, SAS and SCSI
15:59 TBC_x are you cloning your blu-ray with backups?
16:01 TBC_x also, those HDD backdoors are security concern
16:01 est31 but very interesting article bluegreen
16:01 nrzkt est31, on archlinux this is working but not on FreeBSD. Here is the portable way: awk '/^set\(VERSION_MAJOR .*\)/{print sub("/[)]/","",$2)}' CMakeLists.txt
16:01 est31 works
16:03 est31 nrzkt, is there a pipe mode of that command too?
16:03 est31 I want to get the CMakeLists.txt onto a separate line
16:04 nrzkt just do " \"
16:04 habs joined #minetest-dev
16:04 nrzkt and call CMakeLists.txt on next line
16:05 habs How can I tell objectively / programatically if minetest is being accelerated by my graphics card or not?
16:06 est31 I want to integrate it with "export VERSION_MINOR=$$("
16:06 est31 so therefore I need to add something before
16:06 est31 not after
16:07 nrzkt your problem is line length, right ?
16:07 est31 yes
16:07 nrzkt in the brackets '{blah}' you cna do lines, it's a script
16:07 est31 well no
16:07 est31 makefile doesnt allow it
16:08 est31 and its not beautiful
16:08 est31 I'll use pipe mode then
16:08 nrzkt ➜  minetest git:(master) ✗ awk '/^set\(VERSION_MAJOR .*\)/{print sub("/[)]/"\
16:08 nrzkt quote> ,"",$2)}' CMakeLists.txt
16:08 nrzkt :(
16:08 nrzkt not possible with makefile ?
16:08 bluegreen how can I add a whole block to  m_transforming_liquid
16:08 TBC_x are you trying to shorten makefile lines?
16:09 nrzkt remove many pipes to read stdout or file only 1 time not 4 :p
16:10 TBC_x nrzkt: how much data is the shell script processing?
16:10 est31 in fact its longer
16:10 nrzkt TBC_x it's not a problem of shell data processing, it's a processing optimisation by using textproc commands right :)
16:11 nrzkt it's longer in shell, but in exec it's better because doesn't read 4 times stdout
16:11 est31 TBC_x, its done ~10 times every compile, on a file with 237 files
16:11 est31 err lines*
16:11 TBC_x that optimization makes sense then...
16:11 nrzkt it's not a 1M lines files, i know, but i like to have good textproc parsing. Pipes are useful, but if you can optimize and use good optimizations practices
16:11 TBC_x btw, what does it do?
16:11 TBC_x and why?
16:12 est31 it extracts the version numbers from the cmake file
16:12 est31 so that we can use it for the android compilation
16:12 nrzkt it read file, find a line then do something and cut
16:12 est31 in c code
16:12 nrzkt with the awk i mentionned it do all things with one tool :p
16:12 est31 we need it because for android we dont use cmake
16:13 TBC_x does it need to process 237 files to get a version number?
16:13 est31 no lines
16:13 est31 there is no noticeable speed benefit at all
16:14 TBC_x so you're optimizing a script, because why not?
16:15 nrzkt i only review est31 pr and see a little shell optimization to do to optimize textproc parsing.
16:15 nrzkt there is no other question to say :p it's only a shell textproc optimization, nothing more.
16:16 Darcidride joined #minetest-dev
16:17 TBC_x I wouldn't mind if the script had to parse all files in src/
16:18 TBC_x but I think this "shell optimization" got a little too far
16:18 blaze joined #minetest-dev
16:18 blaze joined #minetest-dev
16:19 TBC_x and awk is a little cryptic for me
16:19 nrzkt héhé
16:19 Darcidride joined #minetest-dev
16:19 nrzkt good sysadmins look at crazy awk examples :p
16:19 nrzkt my example is only a basic example, i have a friend which parse BNP Parisbas logservers with thousands lines of awk
16:20 TBC_x oh man... that must even have routines for printing detailed statistics and making a coffee
16:20 nrzkt ofc, but they win some time by not using pipes :p
16:21 nrzkt awk is run on some IBM AIX servers , it's a very good machine :p
16:21 est31 making coffe without pipes? how does the milk flow into it?
16:21 TBC_x through variable containers ofc
16:21 est31 :)
16:21 nrzkt xD
16:22 nrzkt awks runs* sorry for the mistacke
16:22 nrzkt mistake*
16:22 nrzkt tired :(
16:25 TBC_x oh cmon, those piston guys require OpenGL 3.2 for everything
16:25 est31 well, RBA will like it
16:26 nrzkt which piston ?
16:26 est31 its a rust rendering engine
16:26 nrzkt OpenGL 3.2 is quite old now
16:26 est31 some people build a minecraft clone on top of it
16:26 nrzkt OpenGL 4.0 could be good
16:26 est31 but its using the minecraft protocol
16:26 est31 very bad choice they did
16:27 est31 even minecraft for windows 10 doesnt use that protocol
16:27 TBC_x I don't care as far as it is FOSS
16:27 TBC_x I was playing minecraft a lot
16:28 Calinou sadly, we carry with us a lot of legacy
16:28 Calinou requiring anything higher than OpenGL 3.0 for us is not sane for our project
16:28 est31 well the fact that they use that broken protocol for example limits the world height to 256
16:28 est31 or is it 512??
16:28 est31 dunno
16:28 Calinou but at least we're not OpenArena, we don't stay with OpenGL 1.x as minimum :)
16:28 Calinou 256 est31
16:28 est31 also they seem to require clients to install mods too
16:29 est31 and limited block id range
16:29 est31 well ours is limited too
16:29 est31 but much much larger
16:29 TBC_x the worst thing on the servo requiring 3.2 is that It crashes with "Unable to create GL context"
16:29 TBC_x err
16:29 TBC_x piston
16:29 TBC_x i keep confusing servo with piston, idk why
16:29 est31 hehe its no rendering engine yet :)
16:29 est31 err
16:29 est31 it is
16:30 est31 damn, both are
16:30 est31 ok one is 3d and for games
16:30 est31 and the other for html mostly 2d
16:31 TBC_x the best thing is that it is enough to strip the code from requiring OpenGL 3.2 context and it runs
16:32 Calinou make a pull request then? that requires only 3.0 or 3.1
16:32 TBC_x my best OpenGL is 2.1
16:33 est31 well, its also part of a guarantee
16:33 TBC_x 00:02.0 VGA compatible controller: Intel Corporation Mobile GM965/GL960 Integrated Graphics Controller
16:33 est31 perhaps piston will work, but sb can make a game that will break?
16:33 TBC_x sb?
16:33 est31 e.g. usually you can specify your shaders pretty freely
16:34 est31 use a function thats only available in higher versions -- > compile fail
16:34 est31 and thanks to the shaders not being abled to precompile, nobody notices
16:34 est31 vulcan will do lotsa things right here
16:35 est31 yes Calinou I think these reduz arguments are convincing :p
16:35 Calinou I wonder if free/libre Vulkan graphics engines will pop up quickly or not
16:35 Calinou maybe OGRE will have a Vulkan back-end, I don't know
16:36 TBC_x the worst thing is that the OpenGL 3.2 is there explicitly... I mean, you can change it down to 2.0 but otherwise everything crashes mysteriously
16:36 est31 well, one of them is very eager to make a vulcan port...
16:36 est31 and we have seen somebody doing a voxel game with that engine, in pure c++
16:36 est31 ofc highly simple still
16:37 TBC_x OGRE is IMHO too heavy
16:37 Calinou yeah, it is heavy, but it is featureful.
16:37 TBC_x http://openmw.org
16:37 Calinou OpenSceneGraph wasn't made with games in mind, it's not as complete and the community is smaller.
16:38 TBC_x I prefer composition
16:38 nrzkt libstdc++ is also heavy, maybe we should switch.
16:38 est31 erm what
16:39 nrzkt ironic
16:39 est31 hehe
16:42 TBC_x perl wasn't probably made with CGI in mind aswell
16:43 est31 perl was made in mind i think with "make large programs without thinking about architecture"
16:43 est31 which, there is no other way, has to lead to spaghetti
16:43 bluegreen #2977
16:43 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
16:44 est31 nrzkt, is it possible to use other characters than )?
16:44 est31 that ) makes me trouble
16:44 est31 especially near the " ",$2)}
16:45 TBC_x I made a script that listens on :80 and parses requests in specific format and shoves the data into a SQL database
16:45 nrzkt the ) after $2 ?
16:46 est31 yes
16:48 kilbith bluegreen: don't need to spam the channel with your PR for the 6 times in 2 hours
16:51 habs How can I tell objectively / programatically if minetest is being accelerated by my graphics card or not?
16:51 MinetestForFun joined #minetest-dev
16:51 TBC_x habs: what's the problem?
16:52 Calinou habs, Minetest does not switch automatically to software rendering in case it can't find 3D acceleration; so if it runs, then the game is accelerated.
16:52 habs TBC_x: I think that 3D acceleration is working on my laptop, but I see no difference between chosing OpenGL and software rendering in minetest (both 8-10 FPS).
16:53 Calinou OpenGL lets you use shaders, unlike software rendering
16:53 TBC_x software rendering may get you sick
16:54 sfan5 Calinou: software rendering does not imply that you're not using OpenGL or are able to use shaders
16:54 TBC_x habs: do you use windows or linux?
16:55 habs TBC_x: I'm using an Ubuntu 14.04 deriv on a Thinpad X200
16:56 habs Calinou: If that is true then shouldn't I see a difference between chosing "OpenGL" and "software" in render settings? I get the same FPS for both which is what confuses me
16:57 TBC_x habs: open terminal and write glxinfo | grep rendering
16:58 habs TBC_x: "direct rendering: Yes"
16:58 TBC_x then you have hardware acceleration
16:59 habs TBC_x: Is the acceleration just faulty then because I see no difference compared to software rendering?
17:00 sfan5 fast software rendering just means that your computer is good
17:01 habs Software rendering is hardly fast though; for both I get 8-12FPS
17:01 TBC_x maybe Unity problem?
17:02 Calinou habs, latest Git doesn't have a GUI option to switch between renderers anymore :)
17:02 Calinou we don't expose it because it has too many bugs
17:02 kahrl joined #minetest-dev
17:02 Calinou and people change it without knowing what it actually does
17:04 kahrl habs: did you by any chance try the different driver settings in the main menu without restarting after switching?
17:04 habs kahrl: Yes, that is what I was trying
17:05 kahrl the setting only takes effect whenever the program starts
17:07 est31 ok nrzkt I'll make a pr with my changes, so that if you find a command that works, you can suggest it.
17:08 TBC_x #2977 should probably squash commits
17:08 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
17:09 habs TBC_x: I'm using Gnome 3 and not Unity as my DE
17:11 TBC_x do other games run normally?
17:13 zat joined #minetest-dev
17:13 habs kahrl: OK thanks, I tried with restarting, software rendering is like 20-25 FPS notably faster, but the graphics are *very* messed up that way; I took a screenshot
17:13 habs https://i.imgur.com/1ZCdwIt.png
17:13 TBC_x I'm no gonna look at it
17:13 kahrl yeah that's to be expected
17:14 habs TBC_x: I haven't tried many others; glxgears runs at 63 FPS fullscreen if it matters
17:14 est31 nrzkt, can you figure out the right command for #2978 ?
17:14 ShadowBot https://github.com/minetest/minetest/issues/2978 -- Android: Add githash header to spare rebuilds after new commits by est31
17:14 est31 or +1 it :)
17:16 TBC_x habs: do you have shaders disabled?
17:16 err404 joined #minetest-dev
17:18 TBC_x and how long was the game even running?
17:20 habs TBC_x: Yes shaders are disabled, and I've played for ~30 minutes at a time with the same results. I still don't really understand why OpenGL would be slower than software rendering in this case
17:22 TBC_x running a debug build could also cause low FPS
17:22 habs TBC_x: This is a stable build, 0.4.12
17:33 TBC_x at this point, I would check other games
17:34 leat1 joined #minetest-dev
17:36 nrzkt est31: sset CMakeLists.txt filepath in a variable permit to have only 1 line
17:37 nrzkt and for path maintenance it's interesting to have only 1 var
17:37 est31 nrzkt, that didn't fix it
17:37 est31 also its a global variable, thats bad
17:37 nrzkt what is the problem ?
17:37 est31 the problem is it doesnt work
17:38 nrzkt can you paste me the result ?
17:39 est31 if you want you can try it yourself, change the vatiables, cat src/android_version_githash.h after a make.
17:39 est31 variables*
17:40 est31 the output is like "VERSION_GITHASH "}' /path/to/CMakeLists.txt).}' /path/to/CMakeLists.txt).}' /path/to/CMakeLists.txt)-845d0501-Android"
17:40 est31 so the actual variables are set to ""}' /path/to/CMakeLists.txt)"
17:41 nrzkt very strange
17:41 nrzkt i need to test it but i'm on another thing
17:42 est31 its due to how make does the shell command
17:42 proller joined #minetest-dev
17:42 est31 its VARIABLE := $(shell your-command-here)
17:42 est31 if your-command-here contains a ), you are in trouble
17:43 est31 but as it seems, the earlier occurences of ) didnt cause problems
17:43 nrzkt backslashing doesn't solve the problem ?
17:43 est31 so very weird...
17:43 est31 nope
17:43 nrzkt :(
17:43 est31 well, look at it if you have time
17:43 TBC_x you want to keep only single line off a multiline?
17:48 est31 ?
17:50 TBC_x that looks horrible
17:54 TBC_x can't you just create directory buildscripts/ and move that ugliness there?
17:56 TBC_x because this looks like Java inside a .c file
17:56 est31 it would still involve lots of repetition
17:58 TBC_x Does running this involve using android SDK?
17:58 OldCoder joined #minetest-dev
17:58 est31 yes, sdk and ndk
17:59 TBC_x then I probably won't help you
17:59 est31 lol why
17:59 TBC_x don't want to mess with that thing
18:00 est31 heh
18:00 TBC_x with android sdk
18:01 TBC_x why cmake doesn't work?
18:02 est31 it could work
18:02 TBC_x but?
18:02 est31 but nobody has coded it yet
18:03 TBC_x https://github.com/taka-no-me/android-cmake
18:03 est31 do you trust it
18:04 est31 since I've seen horrible things happening with libiconv, I'm very distanced from "buildsystem" repos for android on github.
18:05 TBC_x you can give it a try
18:07 TBC_x have you ever looked at GNU Hurd?
18:09 est31 its a nice research project
18:10 TBC_x they imho went too far with multiple UIDs
18:13 TBC_x though, I like that it is recursive
18:18 H-H-H joined #minetest-dev
18:34 bluegreen joined #minetest-dev
18:38 kilbith https://github.com/freeminer/freeminer/comm​it/7ba086c25509c740c48fd3a9c4a16d90b460decb
18:38 kilbith relevant ?
18:39 kilbith (recent RBA code btw)
18:39 est31 and horrible style
18:39 est31 using auto and a cast
18:39 nrzkt auto isn't forbidden, it exists then it can be used :)
18:39 est31 should be relevant.
18:40 nrzkt but we should use scene::SMesh*
18:40 nrzkt here is a leak, he is right
18:40 TBC_x damn... I was looking at that code
18:40 nrzkt this fix is totally correct
18:40 TBC_x I couldn't decipher it
18:40 nrzkt the style is not the best for our rules, but the fix is correct
18:41 nrzkt if we replace mesh with new mesh and don't delete older memleak
18:41 hmmmm d'oh
18:41 hmmmm i missed another RBA leak
18:41 hmmmm sorry guys, i was the one supposed to review it
18:42 nrzkt no problem, we are human not bots
18:45 hmmmm i think we should make a new contribution rule
18:45 hmmmm if the committer's name matches "RealBadAngel", then current HEAD with your patch cannot have any new valgrind warnings
18:46 est31 lol
18:46 VanessaE heh
18:48 nrzkt xD
18:52 johnnyjoy I'm trying to build by buildwin64.sh on Windows 7, and I'm getting "-- Looking for include file endian.h - not found" Is this a known issue, or it my setup at fault?
18:54 bluegreen don't do that
18:54 hmmmm that's a bug.. USE_ENDIAN_H is supposed to be not set
18:55 johnnyjoy Thanks, hmmmm
18:55 hmmmm HAVE_ENDIAN_H rather
18:55 hmmmm you can work around this by manually unsetting it
18:55 johnnyjoy Will do.
18:56 TBC_x bluegreen: you should squash commits on #2977
18:56 ShadowBot https://github.com/minetest/minetest/issues/2977 -- Stop liquids from flowing over ignore. by red-001
18:57 bluegreen how do I do that?
18:57 bluegreen i'm not good with git
18:57 TBC_x google for `git squash commits`
18:59 sloantothebone joined #minetest-dev
19:00 sloantothebone joined #minetest-dev
19:12 est31 joined #minetest-dev
19:12 est31 hmmmm, what exactly are the blocker issues for 0.4.13 again? We have a tracker in github but your list is I think not included.
19:20 hmmmm i don't think i added it
19:20 hmmmm it's in the minetest-dev logs
19:20 hmmmm i'm sorta really busy right now but i'd otherwise love to help
19:21 VanessaE http://irc.minetest.ru/minet​est-dev/2015-07-28#i_4342517
19:28 OldCoder joined #minetest-dev
19:44 nrzkt oO github has changed its mails notification for commiters
19:44 nrzkt he gives me PR patched and diff links for your PR by mail est31
19:46 hmmmm add to that list in the logs the bugs that vanessae posted backtraces for
19:47 hmmmm they seem pretty easy to track down.  just need to make time for them
19:47 hmmmm it's the weekend so i should be able to get something done hehe
19:58 bluegreen joined #minetest-dev
20:00 blaze joined #minetest-dev
20:08 kilbith joined #minetest-dev
20:19 MinetestForFun joined #minetest-dev
20:29 OldCoder joined #minetest-dev
21:31 hmmmm ha ha!  minetest time!
21:32 VanessaE uh oh.
21:32 VanessaE ;)
21:40 bluegreen Hi hmmmm
21:51 sloantothebone joined #minetest-dev
22:02 proller joined #minetest-dev
22:06 paramat joined #minetest-dev
22:17 proller joined #minetest-dev
22:21 proller joined #minetest-dev
22:35 err404 joined #minetest-dev
22:43 err404 joined #minetest-dev
22:51 bluegreen game#602
22:51 ShadowBot https://github.com/minetes​t/minetest_game/issues/602 -- Allowed players with server priv to access any locked chest. by red-001
22:59 paramat will push #2976 soon
22:59 ShadowBot https://github.com/minetest/minetest/issues/2976 -- Dungeon generation: Fix code style issues in dungeongen.cpp by paramat
23:17 blaise joined #minetest-dev
23:18 VanessaE two more backtraces added to #2913
23:18 ShadowBot https://github.com/minetest/minetest/issues/2913 -- Unexplained, random crashes (segfaults, aborts, OOM)
23:35 paramat now pushing 2976
23:44 paramat complete
23:53 paramat VanessaE and all see game#598 is there any point in aliasing 'default:mese_block' to 'default:mese'? everyone knows that ':mese' is a block as it has always been and ':mese' is used in the crafting recipies. can we just drop this ':mese_block' name?
23:53 ShadowBot https://github.com/minetes​t/minetest_game/issues/598 -- Fix mese block alias by CraigyDavi
23:54 VanessaE by now, it's probably safe to drop the alias
23:55 paramat ah i can see ':mese_block' was used for a while from late 2012
23:55 VanessaE yeah, it was just a transitional thing
23:56 bluegreen game#602  should I add a new privs for accessing locked chests?
23:56 ShadowBot https://github.com/minetes​t/minetest_game/issues/602 -- Allowed players with server priv to access any locked chest. by red-001
23:57 paramat yeah and ':mese' was used for the ore, eww =P what a mess
23:57 paramat okay i'll add to the PR

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