Minetest logo

IRC log for #minetest, 2015-03-07

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

All times shown according to UTC.

Time Nick Message
00:00 est31 what is the problem
00:04 Xenoth joined #minetest
00:06 WarrTab It just reorders the snapshots, not the actual changes, from the look of it
00:06 WarrTab I'm trying cherry-pick now
00:06 WarrTab If I can just exclude a few commits as not being the cause of the problem, the diff to examine gets a lot shorter
00:07 est31 yup
00:10 Warr1024 takes forever to build
00:10 Warr1024 maybe because I always build from absolutely clean
00:10 Warr1024 maybe I'll try preserving the deps cache
00:11 meldrian joined #minetest
00:24 domtron joined #minetest
00:33 LazyJ joined #minetest
01:02 Viper168 joined #minetest
01:19 domtron joined #minetest
01:20 paramat joined #minetest
01:24 jojoa1997 joined #minetest
01:35 domtron joined #minetest
01:35 Zeno` joined #minetest
01:37 Wayward_One hello Zeno` :)
01:37 Zeno` hiya :)
01:37 Zeno` How's things?
01:39 * est31 is trying to add travis builds for android, but had to find out that build proces is highly non-deterministic with result: once it fails once it doesn't
01:40 Wayward_One things're good here, how about you?
01:40 * Zeno` has gotten depressed over the android issues :/
01:41 est31 Warr1024 is trying to bisect further for the text issue ... inside non-building terrain
01:44 Zeno` I'm 99% sure it's 9e2a9b55e185b92074b32d3df336920f33b29e5a
01:45 Warr1024 don't wait up for me, this could take weeks :-/
01:46 chchjesus_ joined #minetest
01:46 Zeno` The thing is I don't even really know why that commit was really necessary
01:47 Zeno` It didn't fix anything particularly necessary
01:48 Warr1024 seemed to be a performance thing
01:48 Warr1024 it looked like the idea was to stop redundantly converting the same strings over and over
01:49 Zeno` nah, it was more of a "to avoid future minor memory leaks in the future" by the looks of it
01:49 Warr1024 except that I'm not sure what the full implications of const declarations are
01:49 Zeno` there are instances now where string are converted from mbs to wide and back to mbs
01:49 Zeno` the const declarations are not a concern...
01:50 Warr1024 seemed odd the way it used to be converting, then using, then freeing, but then that got changed to converting, then using, no freeing...
01:50 _schism_ joined #minetest
01:50 MinetestBot [git] paramat -> minetest/minetest: Dungeongen: Optionally set ignore to be untouchable to disable floating dungeons ffdf8de http://git.io/pUW1 (2015-03-07T01:41:53Z)
01:50 est31 ok now its deterministically *failing*
01:51 est31 some weird issue with openal
01:51 est31 it doesnt do anything thats the problem
01:51 Warr1024 It's tough for me to read because I work a lot in C and C#, and this involves a lot of the parts of C++ that are like neither.
01:51 Zeno` Warr1024 it's still doing that but it's in a "wrapper" (the allocate and free)
01:51 est31 you can sit there for 10 mins, and its completely silen
01:51 est31 t
01:51 est31 no cpu usage, nothing
01:51 Zeno` est31, it's all that downloading that android build does probably
01:52 Zeno` e.g. it downloads irrlicht and then patches it
01:52 est31 no network activity too
01:52 est31 it did already all those downloads
01:52 Zeno` oh, sorry. I thought you were on Travis
01:53 est31 first step is to get it work locally
01:53 Warr1024 est31: biggest cause of non-det build results I usually see is not starting from a fully clean tree.
01:53 est31 where I don't have to wait 10 minutes in the queue
01:53 est31 I'm doing an rm -rf and fresh clone every time
01:53 Warr1024 also common is parallelism
01:54 est31 that can be the issue
01:54 est31 parallelism is somehow taken wrongly here
01:54 est31 because submakes got -jN option
01:54 est31 no good
01:54 Warr1024 it's set hard in the makefile to auto-detect, I had to patch it out myself
01:56 * est31 fixes the now two lines spanning command that pulls, copies config, and does make
01:56 est31 so that it also patches that out
01:58 Zeno` I have non-det builds for Android even without Travis :D
02:02 est31 still hanging
02:02 est31 hanging at: [armeabi] Compile thumb  : freetype2-static <= autofit.c
02:03 est31 perhaps I just have no free ram
02:03 est31 600 MB free
02:03 est31 rebooting...
02:06 Zeno` Another "problem" is that 9e2a9b55e185b92074b32d3df336920f33b29e5a changes more files than it needed to :/
02:08 twoelk|2 joined #minetest
02:09 est31 joined #minetest
02:13 Warr1024 kinda makes me wonder if it's possible to revert that problem commit against master and restore the android build
02:18 Warr1024 yeah, bisecting isn't working so well
02:19 Warr1024 straight cherry-picks won't build
02:19 Warr1024 and trying to fix each bosection by hand is probsbly beyond my c++ ability
02:20 Miner_48er joined #minetest
02:20 domtron_ joined #minetest
02:21 est31 it really seems to build
02:22 est31 I think I can start creating first travis scripts
02:26 est31 wow successful
02:26 roboman2444 joined #minetest
02:31 Zeno` I'm going to try and partially revert it
02:32 Zeno` I'm sick of trying to work out what the problem is and if we have a "stable" (haha) base to start with I might have more success
02:32 Zeno` s/stable/something-that-works
02:43 Warr1024 Zeno`: you been working on this problem long?
02:44 Zeno` well about 10 days but 4 or 5 was getting Android to work
02:44 Zeno` then there was another day or 2 when I reinstalled OS
02:45 Zeno` it's just so slow
02:45 Zeno` build/test-on-emulator/build/etc
02:47 Warr1024 it's faster to test on a physical device, but the emulator might offer the option of automating regresion testing
02:48 Zeno` maybe
02:48 Warr1024 which would be awesome considering how much of a pain this one regression has been.
02:48 Zeno` are you ok with applying git patches temporarily? Warr1024
02:48 Warr1024 yes
02:48 Warr1024 if you're asking me if I can test a patch, I can probably handle it
02:49 Zeno` https://github.com/Zeno-/minetest/commit/06e​774b100bc6d093700e7fa6001e72f39dd4093.patch
02:49 Warr1024 I'd be testing on a samsung gt-p5113
02:49 Zeno` I only just did that... still compiling myself
02:49 Zeno` it'll leak memory I'm just curious to see if the @#&*$(@*$( bug goes away
02:53 est31 are you sure its not a ##*&\\? bug?
02:54 Warr1024 more like a \a\a\a\a\a bug
02:54 Zeno` haha :P
02:54 Zeno` getting a sandwich :) brb
02:55 Warr1024 apparently my git-fu is weak when it comes to applying patches
02:56 Warr1024 but patch -p1 seems to have done it
02:56 Zeno` make a branch; switch to branch; and git am blah.patch is easiest
02:57 Warr1024 building now
02:57 Warr1024 I tried git am, but I must have got my repo in a weird state
02:57 Warr1024 maybe I left a pending bisect or rebase or smth
02:58 Warr1024 would be nice if git-status told you about ALL types of state
02:58 Warr1024 eh, I'll eventually probably blow it away and reclone
02:59 Zeno` see, this is the problem I have. My build just failed for no @*($#(@ reason
02:59 * Zeno` starts it again
02:59 Zeno` I think I need to set up a VM for Android builds
03:00 Warr1024 Zeno`: edit android Makefile and make both PARALLEL := lines set to 1
03:00 Zeno` ok
03:00 Zeno` so that'll stop -j8 ?
03:00 Warr1024 if it's running parallel builds, one thread can fail, but it lets the other keep going until it returms to make
03:00 Zeno` if it fixes the problem of builds I don't care if it takes longer... /me tries
03:00 Warr1024 whst that means is that the actual error ,essage gets buried deep
03:01 Warr1024 it may not fix, but it'll make it easier to see problems
03:01 Warr1024 -j8 huh?  must be nice
03:02 Zeno` it won't fix... my problem is that sometimes the sdk links properly and sometimes it doesn't. Maybe a bug in fedora, dunno
03:02 Zeno` model name      : Intel(R) Core(TM) i7-4770K CPU @ 3.50GHz
03:03 Warr1024 fancy
03:03 Zeno` fancy smancy lol
03:03 Warr1024 my top machines are a core 2 duo and an athlon 64 2-core
03:03 Warr1024 rest are all unicore, some HT some not
03:04 Warr1024 not including my tablet which I think is 2-core, but not much for compiling
03:04 Zeno` if they do the job, they do the job *shrug*. I got this machine when I was working and it was a tax deduction (I needed the CPU for GIS anaylsis stuff for work and the faster the better)
03:05 Zeno` if ubuntu wasn't so ugly I might be tempted to ditch fedora
03:05 Zeno` although my server uses fedora as well and I've used fedora forever so I probably still wouldn't :(
03:06 * Zeno` gets talkative when he's eating
03:06 Zeno` heh
03:06 Zeno` must be the adrenaline :-o
03:07 Warr1024 I'm mainly an openbsd junkie, but tried ubuntu to see what all this maimstream "linux" is all about
03:07 Warr1024 ended up on debian stable
03:07 Warr1024 which just turned your patch into an apk
03:08 Warr1024 time to test
03:09 domtron_ joined #minetest
03:10 Warr1024 shit, it works
03:10 Zeno` I know that
03:10 Zeno` but need(ed) confirmation :)
03:10 Zeno` ok, now time to clean it up
03:10 Warr1024 awesome
03:10 Zeno` thank you for that
03:11 Warr1024 maybe I should play it a little and give it a more thorough test
03:12 Zeno` yeah that'd be good. I'll start making the code look pretty and fix potential leaks
03:14 Warr1024 I logged into a server, putzed around, accessed inventory, and saw nothing amiss
03:16 Warr1024 I wonder if this will be sufficient to get an official android build up on minetest.net again
03:16 est31 now lets wait until we get to the front position in the queue: https://travis-ci.org/est31/minetest/jobs/53427091
03:16 Zeno` actually the whole reason ShadowNinja probably did this was to avoid having to free the string returned by getttext... but there are places where it's still deleted[] so I have no idea why there aren't memory errors anyway :-o
03:16 est31 and then hope it works :D
03:16 clymancer joined #minetest
03:16 Warr1024 though I suppose that may depend a lot on someone being willing to maintain it
03:16 Zeno` returned by wgettext*
03:17 Warr1024 Zeno`: yeah, I saw the inconsistency in freeing/not-freeing the strings, and was thoroughly confused.
03:18 est31 sudo make me a sandwitch
03:18 Warr1024 Password:
03:18 Warr1024 :-)
03:20 Warr1024 actually, maybe that should be "make: *** No rule to make target `me'.  Stop."
03:21 est31 oh cool
03:21 est31 sudo rm -rf /
03:21 Warr1024 I always wanted to get someone with a url like http://host/path?auth=58749305&amp;​/sbin/reboot&amp;key=4736383946565
03:22 Warr1024 if you happen to use a command line downloader like curl or wget
03:22 Warr1024 and paste that into your prompt naively...
03:23 Warr1024 I don't think it'll work on any realistically setup system though
03:23 devmarth joined #minetest
03:26 * Zeno` clicks the link
03:27 domtron_ joined #minetest
03:27 * Zeno` *** No route to host
03:27 Warr1024 yeah, it doesn't do anything interesting if you click
03:28 Warr1024 hey, I noticed in narrow_to_wide_c
03:28 Warr1024 this alwaus bugged me
03:28 devmarth heh
03:28 Warr1024 if char is 32..127, convert it
03:28 Warr1024 else if \n, convert that
03:28 Zeno` yeah it's dumb
03:28 Warr1024 else... what?
03:28 Zeno` else skip the character
03:29 devmarth whats this?
03:29 Warr1024 I guess it just squashes them into zero-width?
03:29 Zeno` nobody knows what that function is really supposed to do
03:29 Warr1024 ha
03:29 Zeno` it's one of those mysteries of minetest
03:29 Zeno` I know what it *does* do, but that's about it
03:30 Warr1024 the nested loops for reverse conversion also irk me in principle, though realistically they're not all that bad, since the inner loop is short
03:31 Warr1024 tempting to do a LUT or something, but I'm not too confident in the trade-off
03:32 Warr1024 maybe the best long-term fix would be to get gettext ported to the NDK or something and re-unify text handling code for android and desktop
03:53 WarrTab Android is also loading a ton faster since 0.4.10
03:55 ShadowNinja Zeno`: It's a replacement for mbstowcs because that function's broken on Android..
03:55 ShadowNinja -.
03:55 Zeno` ShadowNinja, I know that much
03:55 Zeno` :p
03:55 Zeno` strangely enough that's not what the issue seemed to be
03:56 Zeno` what I meant by I dunno what it does, was "I dunno if it does it properly" :)
04:03 est31 just so lol: something is rm'ing inside travis build
04:03 est31 here its missing make: https://travis-ci.org/est31/​minetest/jobs/53429047#L2618
04:04 est31 and here you can see its there (output of ls): https://travis-ci.org/est31/​minetest/jobs/53429047#L150
04:04 est31 black magic
04:25 est31 DAMN
04:25 est31 you should just damn google the error message
04:25 est31 http://askubuntu.com/questions/167530/w​hy-cant-i-run-any-android-ndk-commands
04:25 est31 *facepalm*
04:30 frecel joined #minetest
04:34 Moyst_ joined #minetest
05:00 devmarth Zeno how do I upload a directory to git?
05:01 devmarth * Zeno`
05:03 devmarth eh nvm
05:03 Zeno` worked it out?
05:03 Zeno` the github instructions are pretty useful as well
05:07 Zeno` Maybe he did it wrong
05:09 crazyR joined #minetest
05:09 Vazon joined #minetest
05:12 hmmmm joined #minetest
05:15 devmarth joined #minetest
05:16 ruskie joined #minetest
05:19 paramat left #minetest
05:23 Viper168 joined #minetest
05:26 Zeno` WarrTab, you still here?
05:31 heno72 joined #minetest
05:44 Wuzzy joined #minetest
05:45 est31 joined #minetest
06:24 * VanessaE pokes Zeno` ... just because.
06:24 Zeno` ouch :(
06:27 VanessaE aw
06:27 VanessaE it was just a friendly, gentle poke :)
06:27 Zeno` no. it hurt!
06:29 Zeno` hurt worse than a paper cut
06:30 VanessaE D:
06:31 * VanessaE hides
06:32 Zeno` blood everywhere!
06:51 VanessaE bbl, bed
06:59 devmarth_ joined #minetest
07:00 devmarth_ hey, anybody (preferably zeno, youve been a great help!) know why i get this error when compiling this: http://github.com/blokel (my compile script is in the README)
07:00 devmarth_ Makefile:133: recipe for target 'all' failed make: *** [all] Error 2
07:04 Zeno` there should be more errors than just that
07:07 devmarth_ i might have fixed it when i just tried something, but here, you try compiling with http://pastebin.com/j6tmWKNg that script, and tell me if u get errors
07:07 devmarth_ if youre not busy
07:08 est31_ joined #minetest
07:08 fusion44 joined #minetest
07:08 est31_ joined #minetest
07:09 devmarth_ im trying myself too, but if there is still errors, hopefully you can see too (if you dont want to compile, i can pastebin everything if you want)
07:10 est31_ pastebin the errors yes
07:10 devmarth_ compiling is at 36%. also i can compile minetest itself fine, its an error with the fork im working on..
07:13 est31 joined #minetest
07:14 Zeno` that's a freeminer fork though
07:14 devmarth_ No.
07:14 Zeno` oh
07:14 devmarth_ It used to be, but I deleted it and forked it directly from Minetest, as you can see at http://github.com/blokel/blokel
07:14 Zeno` it was the other day when I looked :)_
07:15 Zeno` hehe serves me right for not checking again today
07:15 devmarth_ You're right, that was an outdated source I had on GitHub and never got around to updating it, so it's not your fault :P
07:15 Zeno` I have a memory like a sieve, that's why I remembered it
07:15 devmarth_ Sorry for being so annoying today, trying to get some stuff done that the dev wiki seems to not cover :P
07:15 Zeno` something like that anyway
07:16 Zeno` the dev wiki doesn't cover much tbh
07:16 devmarth_ Yeah. I've already read the entire thing.
07:16 Zeno` I can't use that script, I'm sorry (I don't use a debian based distro)
07:16 Zeno` but it looks ok
07:17 Zeno` I would do "out of source builds though
07:17 Zeno` e.g.  mdkir minetest-build; cd minetest-build; cmake .. -DENABLE_GETTEXT=1 -DENABLE_FREETYPE=1 -DENABLE_LEVELDB=1 -DENABLE_REDIS=1; make -j$(grep -c processor /proc/cpuinfo)
07:17 devmarth_ http://pastebin.com/gtkTHdUE
07:18 devmarth_ thats my entire output from running the script
07:18 devmarth_ and of course, you know where the source is :P
07:18 Zeno` you're missing a closing ) at /home/devmarth/blokel/src/guiEngine.cpp:174:61
07:18 Zeno` i.e. line 142 of guiEngine.cpp
07:18 Zeno` err 174
07:19 devmarth_ oh it was a typo
07:19 devmarth_ any other errors (probably typos)
07:19 devmarth_ ?
07:20 Zeno` not in what you pasted
07:21 devmarth_ also before i go and compile it, can you tell me how to rename the output "minetest"? in the /bin folder by default xD I looked for hours in the dev wiki and it is nowhere to be found
07:21 devmarth_ It's quite annoying for people starting out with core development (I am pretty fluent in the lua part of it)
07:21 devmarth_ i want it to be "blokel" so i can type "./blokel" and itll run
07:22 devmarth_ currently its "./minetest"
07:22 Zeno` e.g. corrected:     mdkir minetest-build; cd minetest-build; cmake ../minetest -DENABLE_GETTEXT=1 -DENABLE_FREETYPE=1 -DENABLE_LEVELDB=1 -DENABLE_REDIS=1; make -j$(grep -c processor /proc/cpuinfo)
07:23 devmarth_ Do i need to use that? Isn't my current compile script fine?
07:23 devmarth_ the one here: http://pastebin.com/j6tmWKNg
07:23 est31 devmarth_, there is currently an open pull request that does some fork-friendly changes
07:23 devmarth_ what do you mean by fork friendly? est31
07:23 Zeno` devmarth_, it's fine... it's just leaves all the crap in the actual source directory (I personally don't like that but a lot of people don't mind)
07:24 devmarth_ Zeno`, do you know how to make it build blokelserver and blokel instead of minetest and minetestserver? in the /bin?
07:24 est31 devmarth_,  https://github.com/minetest/minetest/pull/2402
07:25 est31 its still hardcoded
07:25 Zeno` personally I'd wait until the pull request shown by est31 is merged, but...
07:26 Zeno` or don't wait
07:26 est31 PR doesnt help in that issue
07:27 Zeno` oh it's still hardcoded?
07:27 est31 as I said its still hardcoded there
07:27 devmarth_ i can wait i guess, it doesnt really have anything i desire immediately
07:27 Zeno` well once ShadowNinja fixes that you can patch the PR to your own repository without waiting
07:28 est31 I dont think whether it should be fixed
07:28 est31 don't know*
07:28 devmarth_ and then it seems all the edits like versions and naming the game seems to be in CMakeLists
07:28 devmarth_ so if he fixes it, ill do that for easing the process
07:29 est31 because source filenames can't be set by a variable
07:29 devmarth_ oh duh
07:29 est31 btw whats the motivation for the fork?
07:29 est31 :)
07:30 devmarth_ just improving upon minetest, making it better and adding features i like, like rpg features, and the use of mapgenv7
07:31 devmarth_ cuz mapgenv7 is stone right now unless you have a lua mapgen with compatibility i believe
07:31 devmarth_ not too sure
07:31 est31 dunno paramat is mapgen pro
07:32 devmarth_ hell yeah
07:32 Zeno` well my android version works now
07:32 devmarth_ mine wont be anything like his
07:32 Zeno` but I'm paranoid after the last few days :/
07:32 devmarth_ i got an android port to work once but the GUI was tiny so it took fiddling to make it "usable"
07:33 est31 I agree current android gui wastes space especially in main menu
07:33 Zeno` anyway once somebody confirms I'll push the change
07:33 est31 especially on phone
07:33 devmarth_ nice
07:33 est31 there should be some way to do if phone then no-space else do-space endif
07:34 devmarth_ thatll be released under 0.5.dev branch right if its pushed?
07:34 est31 I lost oversight over that 0=0
07:34 devmarth_ what?
07:35 est31 about branching behaviour
07:35 devmarth_ never forget about that
07:35 est31 I guess the situation is that all breaking stuff gets to that branch else master
07:36 est31 but it seems it isnt even sure it gets merged
07:36 est31 I dunno
07:36 devmarth_ im kinda suprised with the fast updates recently. i remember 0.4.9 took forever to get to 0.4.10
07:36 est31 0.4.12 was a bugfix release
07:36 Zeno` yes, a release to add bugs
07:36 Zeno` I mean, fix
07:36 devmarth_ lmao
07:36 devmarth_ yeah, i saw the changelog
07:37 Zeno` nah, I mean add
07:37 Zeno` we had a secret name for it
07:37 devmarth_ hold on, got another error, lemme pastebin it quickly, and can u take a look?
07:37 Zeno` it's hidden in the binary
07:37 Zeno` "the buggiest release yet!"
07:37 est31 strings?
07:38 Zeno` yeah use strings
07:38 * Zeno` runs away
07:38 devmarth_ [sudo] password for prestotron:  Reading package lists... Done Building dependency tree        Reading state information... Done build-essential is already the newest version. cmake is already the newest version. libbz2-dev is already the newest version. libgettextpo-dev is already the newest version. libjpeg8-dev is already the newest version. libogg-dev is already the newest version. libpng12-dev is already the newest version.
07:38 devmarth_ oh shit
07:38 devmarth_ i thought i copied the link
07:39 devmarth_ http://pastebin.com/Tr4uWxm6
07:39 devmarth_ im such an idiot wtf
07:39 devmarth_ anyways, there it is, probably more typos because its 2am and im tired
07:40 est31 YAAAAY HOORRRAAAAYYYY
07:40 est31 https://travis-ci.org/est31/minetest/jobs/53436997
07:40 est31 8 hours of android building finally result in a working build
07:40 est31 yesssss
07:40 * est31 does the happy dance
07:40 devmarth_ does it look nice?
07:41 est31 depends on taste
07:41 devmarth_ Zeno`, if I give you contributor access to the fork, can you go in and patch it up? Or just tell me? :P http://pastebin.com/Tr4uWxm6
07:42 est31 https://github.com/minetest/minetest/issues/2287
07:42 est31 devmarth_, you just need a ) before the ;
07:42 devmarth_ Where?
07:43 devmarth_ what file i mean, and line #
07:43 est31 guiEngine.cpp line 572
07:43 devmarth_ same file, different line ugh
07:43 devmarth_ i knew it was typos
07:43 est31 " std::wstring toset = narrow_to_wide( std::string("Blokel 0.5.0")"
07:43 est31 thats what you can / for
07:45 devmarth_ here goes another 15 minute compile .-.
07:45 est31 just do make
07:46 devmarth_ http://pastebin.com/j6tmWKNg
07:46 devmarth_ thats my current compile script i wrote
07:46 est31 modern build systems have the feature of only updating changed files
07:46 devmarth_ it uses make
07:47 devmarth_ Oh, i'm just compiling from source every time, instead of doing it locally
07:47 est31 perhaps its most cleanest way but I dont recommend that for actual development. you have to compile very often, and waiting for compiler can slow you down
07:47 est31 or rather will
07:48 devmarth_ definately
07:49 devmarth_ 53% :/
07:50 Zeno` it the est31 phone working yet?
07:50 est31 no but the build
07:50 est31 get ready for android travis PR
07:50 devmarth_ heh
07:50 Zeno` the build should be fine
07:51 devmarth_ btw zeno, what distro are you running, jw since its not debian you said
07:52 Zeno` Fedora
07:52 devmarth_ I love fedora.
07:52 devmarth_ 21?
07:53 Zeno` yep 21 (since about 3 days ago)
07:53 Zeno` before that I was still on 20
07:53 devmarth_ btw im pretty sure you could just change "apt-get" to "yum" and remove the dependencies (since u already installed it) and it would compile
07:53 devmarth_ fedora was my first distro i ever used. i still use it to this day
07:53 Zeno` well I could, but my user isn't in the sudoer list and I won't add myself
07:54 Zeno` I don't particularly like sudo
07:54 devmarth_ ah alright
07:54 devmarth_ what do you use then?
07:54 Zeno` su -
07:54 devmarth_ ah root
07:54 Zeno` yes
07:54 devmarth_ i see
07:54 Zeno` If I am root I know I'm root
07:54 devmarth_ makes sense
07:54 Zeno` I don't get complacent with sudo blah
07:55 Zeno` I'd rather switch to root, do my stuff, get out :)
07:55 devmarth_ yeah, being pestered with password entries every command :P
07:55 devmarth_ 99% lets goo
07:55 est31 no, sudo has a cache
07:55 est31 I have no root passwod
07:55 devmarth_ wtf
07:55 est31 I do sudo bash
07:56 devmarth_ I opened the game up and it just closes immediately
07:56 devmarth_ it had no errors
07:56 devmarth_ Can someone check this out?
07:56 Zeno` even when I use ubuntu the first thing I do is set a password for root hehe
07:56 devmarth_ wait nvm
07:57 devmarth_ i see errors in lua
07:57 Zeno` if you disable external password logins there isn't any real reason not to
07:57 Zeno` (i.e. external logins using RSA only)
07:57 blaze joined #minetest
07:57 est31 yea I guess its a taste question
07:57 Zeno` or whatever crypto.. just not passwords
07:57 Zeno` yeah it is just a matter of taste
07:57 est31 rot27
07:58 Zeno` I don't think login supports rot27 yet
07:58 Zeno` they haven't even got around to implementing rot13
07:58 devmarth_ in the tab_credits.lua, i have " return "label[0.5,3.2;Blokel 0.5.0"]" "
07:58 devmarth_ it says it has an unecpected end
07:59 devmarth_ wait
07:59 devmarth_ i got it
08:00 devmarth_ no i dont got it
08:00 devmarth_ https://github.com/blokel/blokel/blob/m​aster/builtin/mainmenu/tab_credits.lua
08:01 devmarth_ Line 25, whats wrong?
08:01 devmarth_ Looks good to me
08:05 devmarth_ yo Zeno` est321, see any probs?
08:06 devmarth_ woah nvm
08:06 devmarth_ im dumb
08:21 Dartmouth joined #minetest
08:24 Zeno` bbl
08:25 Zeno` time for me to cook din din
08:25 Zeno` I'll probably have gourmet noodles tonight
08:25 Zeno` lol.
08:25 Zeno` (cbf cooking for once)
08:30 cheapie joined #minetest
08:30 Krock joined #minetest
08:30 Krock meow
08:36 someguy_irc joined #minetest
08:37 decimalguy joined #minetest
08:38 CWz joined #minetest
08:38 CWz For some reason placing a tree block doesn't seem to  appear in console
08:42 Tux[Qyou] joined #minetest
08:44 cheapie joined #minetest
08:47 Krock CWz, log level = 0?
08:47 CWz no
08:48 CWz other blocks do trigger a notice when placed
08:48 CWz fences also seem effected by the bug
08:49 CWz Test also show that tree block will not obey protection when placed
08:50 devmarth_ joined #minetest
08:51 devmarth_ Hey, I just accidently changed my GUI scaling and for some reason, I am unable to change it back. I even compiled a fresh copy of it and nothing worked. which is weird
09:04 Zeno` joined #minetest
09:04 Yepoleb joined #minetest
09:10 blaze joined #minetest
09:10 crack_ left #minetest
09:34 Krock !tell devmarth_ use a text editor to modify the minetest.conf scale setting
09:34 MinetestBot Krock: yeah, yeah
09:36 DarkNekros joined #minetest
10:15 ImQ009 joined #minetest
10:16 est31 joined #minetest
10:31 Krock blokel developers are better than freeminer and freeminer is better than minetest. (looking at the commit history)
10:31 SylvieLorxu joined #minetest
10:31 Haudegen joined #minetest
10:32 CWz In what way is freeminer devs are better then minetest?
10:38 Krock I'm speaking of the commit frequency
10:38 Krock not the content of it
10:39 Dartmouth joined #minetest
10:40 SopaXorzTaker joined #minetest
10:43 CoconutOfDeath joined #minetest
10:49 Tg1 joined #minetest
10:52 MinetestBot [git] Zeno- -> minetest/minetest: Fix Android text bug (no text displaying) 5698e2b http://git.io/pTQr (2015-03-07T20:51:07+10:00)
11:24 Viper168_ joined #minetest
11:37 FR^2 joined #minetest
11:59 Krock joined #minetest
12:02 ImQ009 joined #minetest
12:04 Haudegen joined #minetest
12:14 Krock boredom exists and I can't even play Minetest on my own server without risking to crash the Wireless network :(
12:17 nrzkt joined #minetest
12:38 The_Loko joined #minetest
12:44 MinetestBot [git] Zeno- -> minetest/minetest: For usages of assert() that are meant to persist in Release builds (when NDEBUG is defined), replace those usages with persistent alternatives ced6d20 http://git.io/pkcy (2015-03-07T22:41:47+10:00)
12:45 Tg1 joined #minetest
12:54 Haudegen joined #minetest
12:56 ImQ009 joined #minetest
13:07 roniz joined #minetest
13:10 arsdragonfly joined #minetest
13:12 HoloIRCUser1 joined #minetest
13:14 lemonade joined #minetest
13:20 nore joined #minetest
13:20 ubuntu joined #minetest
13:22 MinetestForFun joined #minetest
13:22 Jordach joined #minetest
13:38 lemonade joined #minetest
13:39 Krock !c 44 * 8
13:39 MinetestBot 352
13:41 Drangue joined #minetest
13:42 Jordach BFD: Minetest's most stable subgame: <JD-BFD> Server has been running for 188:10:17
13:44 Jordach Krock, still can't crash my own subgame :D
13:45 * Krock gives Jordach some old cookie he found under a table
13:49 exio4 Krock: frequency ain't quality
13:49 Krock I totally agree
13:49 * Jordach feels like making a subgame known as Voxel
13:50 Tuxedo[Qyou] joined #minetest
13:54 Jordach 2 legit 2 quit: http://a.pomf.se/wgokdk.png
13:59 meldrian joined #minetest
14:00 domtron joined #minetest
14:19 Zeno` joined #minetest
14:21 Tux[Qyou] joined #minetest
14:23 ElectronLibre joined #minetest
14:31 Jordach joined #minetest
14:38 Zeno` who wants a nice piece of salmon grilled on the bbq with some subtle spices and served with a fresh salad?
14:39 Tuxedo[Qyou] joined #minetest
14:39 Zeno` I think I will use australian spices
14:39 Zeno` lemon myrtle, tasmanian pepper, and australian salt bush
14:41 CWz um i think zeno broke minetest
14:42 CWz after git pull minetest passout after being started
14:43 * ElectronLibre checks if it compiles..
14:43 roniz joined #minetest
14:43 Tux[Qyou] joined #minetest
14:43 shadowzone joined #minetest
14:46 Jordach Krock, pls
14:46 Krock wat
14:46 Krock U want more cookies?
14:47 Jordach yes :e
14:47 Jordach :3
14:48 ImQ009 joined #minetest
14:50 ElectronLibre ( CWz, I have no problem. But I'm using a GNU/Linux self-compiled version, and from what I can guess from #minetest-dev there is something about M$ Windows' sizing.)
14:50 Zeno` CWz, it's fixed now
14:51 Zeno` and yes, a change as big as that so that *minetest handles errors properly instead of using shitty code* was bound to cause hiccups :)
14:52 fusion44 joined #minetest
14:53 Jordach Minetest handles errors correctly?
14:53 Jordach that's a first
14:55 Zeno` well, not it doesn't
14:56 Jordach and can we please have a HTML like formspec system
14:56 Zeno` I pushed a change that aims towards that though
14:56 Jordach (eg, have fluid positioning based on percentages instead of fixed size)
14:56 Zeno` and it cause a minor issue lol
14:57 Jordach i like Taoki's pull to be merged so my mesh chests get an upgraade
14:58 Jordach also client sided animations would save me a lot of work too
14:59 Jordach so it predicts how the entity will animate instead of waiting for a byte flag from the server
15:00 shadowzone joined #minetest
15:01 ecutruin joined #minetest
15:11 * Jordach meows at sfan5
15:20 ElectronLibre joined #minetest
15:26 Tuxedo[Qyou] joined #minetest
15:27 WarrTab Zeno, thanks for fixing android
15:29 WarrTab You fixed it all in one commit
15:30 WarrTab I'd have expected a commit that half fixed it, then another that fixed half of what was left, ad infinitum, never quite fixing it completely...
15:32 Zeno` WarrTab, you're welcome. I think others are happy as well
15:32 Zeno` I know I'm happy. It was starting to annoy me
15:36 WarrTab Please tell me I'm not the first one to make that joke ;-)
15:36 Zeno` Oh, there are very few "firsts" in this world lol
15:36 Zeno` thank *you* btw, for helping earlier
15:36 WarrTab There are still a couple of annoyances left in android btw
15:37 Zeno` couple? or 90?
15:37 WarrTab I have that free-touch option enabled, but it won't let me target stuff on like the top third of my screen
15:37 Zeno` list them somewhere and I'll try and address them. After today I've decided to have tomorrow off
15:38 WarrTab Also I don't know why I can't get v_range above 50 even though I have plenty want_fps to spare
15:38 Zeno` WarrTab, that's an issue with Linux as well
15:38 WarrTab Are you the android maintainer now?
15:38 Zeno` dunno about windows
15:38 Zeno` no I'm not!
15:38 WarrTab :-)
15:39 WarrTab I wonder if I grep -r 50 the src if I'll find the cause...
15:39 Zeno` But I bought an android to play with and dev for so I will fix issues when/if I can
15:39 WarrTab Awesome
15:40 Zeno` right now... 1:40AM
15:40 * Zeno` idles :)
15:40 exio4 :o it's party time
15:43 Zeno` I will party in my bed alone?
15:43 Zeno` lmao
15:43 WarrTab Ha, also forgot about the gui controls needing a shadow so as not to get lost in the clouds, like we had to do for text.
15:43 WarrTab I'll have to spend some time searching the existing bug db
15:50 Zeno` shouldn't take more than a day
15:53 turtleman_ joined #minetest
15:54 shadowzone joined #minetest
16:01 Warr1024 joined #minetest
16:02 Warr1024 it seems that I can't add labels to my own issues?
16:02 Warr1024 (or maybe I'm just not seeing how in the UI)
16:04 gravgun joined #minetest
16:06 Warr1024 whoops, figured out why v_range wouldn't go above 50
16:07 Warr1024 I guess the default setting for viewing_range_nodes_max is 50, at least on android.
16:07 Warr1024 I set mine in minetest.conf via a text editor and it worked.
16:09 ElectronLibre If I remember correctly, only contributors with write access can add labels.
16:10 The_Loko joined #minetest
16:11 Warr1024 ok, that's fine.  I just figured that if I knew enough to tag my own issues, it could save them some work.
16:11 Warr1024 yeah, the built-in default viewing_range_nodes_max for desktop is 240, while for android it's 50.
16:11 Warr1024 not unreasonable for a phone, but a bit limiting for a 10-inch tablet.
16:15 Warr1024 now I'm having a hard time reproducing my issue with free-touch being finicky on android.
16:15 Warr1024 maybe it's only if I'm doing certain things...?
16:23 ElectronLibre joined #minetest
16:29 Jordach_ joined #minetest
16:38 Vazon joined #minetest
16:43 theTroy joined #minetest
16:45 domtron joined #minetest
16:47 est31 joined #minetest
17:04 jojoa1997 joined #minetest
17:04 T4im joined #minetest
17:05 Jordach joined #minetest
17:17 Warr1024 When I'm logging into a server, what's going on during the "Done!" step on the progress bar?
17:18 Warr1024 On fast enough devices/networks, it seems to take up a notable portion of the loading time, so if it's not doing something actually useful beyond a "get ready to dive into the action" delay, it would be nice to be able to skip it with a click/tap or something.
17:19 ElectronLibre I have no idea, but in my opinion it's not time to press the "esc" button. It must be starting to draw the nodes, meshes, and sending first informations to other clients.
17:19 est31 progress bar code is here https://github.com/minetest/minetes​t/blob/ced6d20295a8263757d57c02a07f​fcb66688a163/src/client.cpp#L1558
17:21 Warr1024 hm, so maybe it's the server that's working then, processing the "ready" signal from the client.
17:21 Warr1024 I was worried it was one of those questionable usability changes I sometimes see, and behind the scenes was a sleep or something.
17:22 Warr1024 My favorite questionable sleep call was on TurboTax web, where they had these "please wait while we maximize your tax refund" screens.
17:22 Warr1024 they were pointless flash animations, since those calculations were done continuously on every page load.
17:24 Warr1024 but I figure that anyone might be prone to adding something like that after they get like the 5th bug report from a user saying "it told me to wait, but then the game just started without telling me it was done"
17:35 Viper168 joined #minetest
17:38 Warr1024 I think I figured out how the "android is finicky about digging nodes above eye-level" thing works.
17:38 Warr1024 if your viewport is facing below the horizon, you can't dig nodes above eye-level, but if it's facing above the horizon, everything works as normal.
17:40 ElectronLibre joined #minetest
17:41 btoothpick joined #minetest
17:41 btoothpick I wish the Run Lua in worldedit would let me just past .we files.
17:41 btoothpick It's a pain in the neck to get them in the worlds/world/schems folder
17:41 Krock what do you mean?
17:42 ImQ009 joined #minetest
17:42 Krock I thought they're saved there automatically
17:43 barrydk joined #minetest
17:43 SopaXorzTaker joined #minetest
17:43 btoothpick Well if I find a .we that I want online I can't just put it in a server
17:44 btoothpick Without logging into the server and doing a wget or something similar
17:44 btoothpick I'd like to just be able to past the .we file in the RUN LUA section
17:44 T4im retrieving it remotly via WE opens a lot of possible security issues though
17:44 Krock exactly.
17:45 Sokomine hmm. the game still seems to hang for me for microscopic parts of a second. i didn't have this before
17:45 T4im having that extra step of having to download it makes sure for example that you have server access before possibly being granted server access ;)
17:45 * Sokomine looks puzzled
17:45 Sokomine happened with one of the recent changes...somehow
17:47 btoothpick Understand. I want to put this connect 4 game in https://raw.githubusercontent.com/co​rnernote/minetest-worlds/master/boar​d_games/schems/minigame_connect4.we
17:50 icetea joined #minetest
17:53 est31 joined #minetest
17:53 Sokomine connect 4? a game as a schematic? or is it only the board?
17:53 icetea_ joined #minetest
17:55 T4im btoothpick: the problem is, that there's always a chance that someone that has _no_ server access, only worldedit access, manages to supply an unechecked schematic that is capable of "breaking out" and executing custom lua code… to prevent that there would have to be a lot more checking on the schematic… maybe there is, probably not (I assume lazyness prevails)…
17:57 T4im that said, why don't you just write a quick minimod for your own use that does what you want? download a file and sort it into the right folder
17:57 T4im :)
17:57 est31 without mod security, this can quickly be upgraded even to shell access
17:58 est31 which is horrible
17:58 T4im well that's what lua code implies, yea :(
17:58 T4im os.execute() or what it was
17:58 est31 Not when this one gets merged: https://github.com/minetest/minetest/pull/1606
17:58 btoothpick Looks like I have some puzzles to solve.
17:59 MinetestForFun joined #minetest
17:59 icecream joined #minetest
17:59 Sokomine T4im: depends. we files are usually tables (at least they where the last time i checked), while .mts is a compressed, well defined format. i think .mts ought to be reasonably safe
17:59 T4im hmm that requires some mintest.mkdir() then ;)
18:01 Sokomine the other way around might be safer - i.e. offer a mail service for mailing the player an .mts file of the area he protected
18:01 T4im true
18:01 T4im well.. with some care :D
18:01 T4im spamrelay and that
18:02 Sokomine *nod*
18:02 Sokomine redcrab has this nice way of gaining interact...it's verified by email
18:03 T4im all requiring some form of require() or os.execute() though then :/
18:03 T4im meaning if that mod security thing comes it might become just as jailed in ;)
18:03 T4im but generally a good thing
18:04 DusXMT joined #minetest
18:07 est31 I'm personally a friend of world downloads
18:09 Jordach <3 isometric art
18:10 prozacgod joined #minetest
18:13 Jordach is this a vector? http://a.pomf.se/cvqzts.png
18:14 ElectronLibre To me it's a voice ("vox" in latin).
18:19 Megaf Whoah! Im alive!
18:19 Megaf Hello everyone!
18:19 ElectronLibre Hello Megaf.
18:19 Megaf uptime
18:19 Megaf 13:09:28 up 28 days, 21:52,  1 user,  load average: 0.00, 0.01, 0.06
18:19 Megaf :D
18:20 Megaf going to reboot now for reboots sake
18:20 Jordach Megaf, damn
18:20 Jordach i got 190+ hours
18:20 ElectronLibre Oh, 28 days, not bad.
18:20 Jordach [18:20:24] <JD-BFD> Server has been running for 192:48:47
18:21 Krock ShadowBot, give sfan5 https://gist.github.com/Smal​lJoker/cd4d2685a8d8ec9c3891
18:21 ShadowBot sfan5: <https://gist.github.com/Smal​lJoker/cd4d2685a8d8ec9c3891>
18:21 Megaf Thats actually the container where I run my server, the host itself is slitly higher uptime
18:21 Megaf # uptime
18:21 Megaf 18:21:05 up 29 days,  1:51,  1 user,  load average: 0.00, 0.05, 0.20
18:21 Megaf going to reboot it too
18:22 Megaf I think my best minetest uptime was 3 months
18:22 Krock Jordach, that's a very realistic... thing
18:23 est31 Megaf, I hope you still do backups using some db lock trick?
18:23 Megaf est31: I no longer do that uptime crazy stuff
18:23 Megaf now a days I stop the server and back it up almost every day
18:24 est31 no cronjob for that? quite ambitious :D
18:24 Megaf est31: no contrab, I do that MANualy
18:24 Megaf lol
18:28 Megaf the only automation I have is a script that set some iptables rules and a script that remove old build files and then run a new cmake
18:28 Megaf other stuff I like seeing whats going on
18:32 Megaf Sokomine: Im going to add some of your mods to my server soon
18:33 Endymion joined #minetest
18:33 Megaf Sokomine: Windmill and the large one, bridges and some others
18:33 Megaf anything I should know before adding them?
18:33 Megaf cheapie: is plasticbox stable too? I'm adding it too
18:34 Megaf and spawn :)
18:34 Megaf and infractruscture and streets
18:34 EndymionM Um, question - does homedecor include a book I can write in or paper I can write on?
18:34 est31 EndymionM, yes it does
18:34 est31 any colored book can be written into
18:35 EndymionM est31: Ah, that'd be the issue then, thanks.
18:35 est31 np
18:36 Megaf areas mod still borked, is there any mod compatbile with it?
18:36 est31 borked??
18:36 est31 which mod do you mean Megaf?
18:36 Megaf hm
18:36 Megaf advanced area protection or something liek that
18:36 est31 Shadowninjas port?
18:36 Megaf borked = not working properly
18:37 Megaf est31: yep
18:37 est31 s/port/version/
18:37 est31 why borked
18:37 est31 its working fine on trepca server, and on all of VanessaE's servers
18:37 Megaf and /listareas works?
18:38 est31 Its perhaps /list_areas
18:38 est31 what is the issue?
18:38 est31 for me it does
18:42 Jordach http://a.pomf.se/hoefbs.png ( ͡° ͜ʖ ͡°)
18:43 Megaf cheapie: your mod streets still broken
18:43 Megaf 13:43:01: ERROR[main]: /home/minetest/MegafServer/bin/../​mods/streets/streetsmod/init.lua:
18:43 Megaf 13:43:01: ERROR[main]: cannot open /home/minetest/MegafServer/bin/../mods/str​eets/streetsmod/libs/smartfs/smartfs.lua: No such file or directory
18:44 est31 bye
18:44 Megaf cheapie: https://github.com/webdesigner97/​streets/issues/42#issue-60218850
18:45 T4im Megaf: that's you not downloading the submodule
18:45 T4im ;)
18:46 Megaf T4im: I did a git clone on his repo
18:46 Megaf thats how it should work
18:46 Megaf T4im: how do I fix that?
18:46 T4im only if that clone has a --recurse-submodules
18:46 T4im git submodule update should fix it
18:47 T4im "git submodule update"
18:47 T4im :)
18:48 T4im he should probably write that down somewhere in the readme
18:49 Megaf T4im: submodule update didnt work but a fres git clone --recurse-submodules did it
18:49 Megaf Thanks
18:49 T4im ok :)
18:50 Megaf T4im: And by the way, it was me indeed there earlier on :P
18:50 T4im I know ;)
18:51 Megaf ok, I should have some new cool blocks now
18:51 Megaf cheapie: sorry, streets actually work
18:51 Megaf is there anything that can be done to make a git clone work with that mod without the need of --recurse-submodules ?
18:52 T4im smartfs is in there via submodule… and since it's an external git repo, that's not wrong per se
18:52 T4im "git submodule update" should have worked though.. its supposed to clone missing submodules :/
18:53 T4im its also used to update them to the ref that the git repo tells you to have
18:53 T4im submodules can make sure you have the "right" version with you
18:54 T4im though I wished there was a shorter parameter available :D
18:55 hmmmm joined #minetest
18:55 T4im oh pardon me… I just realized you'd have to do git submodule update --init the first time
18:55 T4im that's why it didn't work :D
18:56 T4im but the clone does that implicitly, so that's not needed anymore now
18:56 Megaf the bridges mod is quite nice actually
19:08 Megaf I think ifrastructure wasnt activated on my server
19:08 Megaf streets is active tho
19:12 Megaf T4im: How is the concrete mod called again?
19:12 T4im prefab
19:12 Megaf Thanks
19:13 T4im Megaf: https://github.com/VanessaE/prefab
19:14 Megaf Thanks a million
19:16 ElectronLibre joined #minetest
19:18 Megaf oh, I need technic =/
19:18 Megaf 14:18:06: ERROR[main]: mod "infrastructure" has unsatisfied dependencies:  "digilines" "prefab" "technic"
19:18 Megaf thats sad
19:19 devmarth joined #minetest
19:20 devmarth I set my GUI Scale Factor way too high and screwed up, now even when I reinstall the game or compile from scratch, it's still unusable ebcause the GUI scale is too high
19:20 devmarth Any suggestions?
19:20 Megaf yep
19:20 ElectronLibre Close minetest. Remove the GUI's scale value from your minetest.conf and restart minetest.
19:20 Megaf it depends on your OS
19:21 devmarth I just compiled it, so theres nothing but minetest.conf.example and I tried making one manually
19:21 devmarth No help.
19:21 devmarth I'm running Ubuntu.
19:21 T4im try ~/.minetest or so
19:21 devmarth I did.
19:22 T4im or in ~/.config/ maybe?
19:22 ElectronLibre Try in your source directory (where minetest is extracted/cloned and compiled).
19:22 devmarth Ah, yes. How did I forget about the hidden files in /home
19:22 devmarth I pressed Ctrl + H, and it let me goto the minetest dir
19:23 Megaf VanessaE: 14:22:29: ERROR[main]: /home/minetest/MegafServer/b​in/../mods/castles/init.lua:
19:23 Megaf 14:22:29: ERROR[main]: /home/minetest/MegafServer/bin/.​./builtin/game/register.lua:60: Name castle:pillars_stonewall_bottom does not follow naming conventions: "modname:" or ":" prefix required
19:23 T4im Megaf: make sure that mod also is named "castle" then
19:23 ElectronLibre The modname is castle, why is the directory called castles?
19:25 devmarth Yeah, notice "castle:pillars", which means the modname should be "castle" and not "castles" in extension to the above message
19:25 Megaf now, thats weird, I had a mod called castle and anotehr called acastles
19:26 Megaf I removed castle and castles and did a new git clone of castles
19:26 Megaf and now I get this, 14:24:52: ERROR[main]: /home/minetest/MegafServer/bi​n/../mods/castles/init.lua:1: attempt to concatenate a nil value
19:26 Megaf 14:24:52: ERROR[main]: stack traceback:
19:26 Megaf 14:24:52: ERROR[main]:  /home/minetest/MegafServer/bi​n/../mods/castles/init.lua:1: in main chunk
19:26 Megaf was kicked by ShadowBot: Paste flood detected. Use a pastebin like pastebin.ubuntu.com or gist.github.com.
19:26 Megaf joined #minetest
19:26 * Megaf kicks the stupid ShadowBot
19:27 dzho joined #minetest
19:27 devmarth Heh
19:27 devmarth ShadowBot has a pastebin crush
19:28 Megaf well, the mod castles get files from the folder castle
19:28 Megaf thats askward
19:28 Megaf akward
19:29 Megaf awkward
19:29 Megaf I give up
19:29 devmarth combine them into a modpack or something
19:30 Megaf renamed the mod castles to castle and now it works
19:30 Megaf I wonder if people even test their own mods...
19:31 Megaf they make just very basic naming mistakes...
19:32 devmarth heh
19:38 Calinou joined #minetest
19:43 ElectronLibre joined #minetest
19:47 Megaf Nice post https://forum.minetest.net/viewtopic.php?p=138999​&amp;sid=44bb04d838b49fdf778f4f277565f556#p138999
19:50 devmarth Does anyone know a way I can just like download someones pull request, like all the changed files?
19:51 Calinou devmarth, clone their fork!
19:51 Calinou it's that easy
19:51 devmarth Oh yes!
19:51 Calinou if you want all the changed files, you'll have to make a diff
19:51 devmarth Yeah, I just want the changed files
19:51 devmarth http://github.com/minetest/minetest/pull/2402
19:52 devmarth I want to download 45 of the changed files.
19:52 FreeFull joined #minetest
19:52 devmarth I know you can clone it, I just want just the 45 files, not everything else haha
19:52 Calinou you'll have to clone the whole repository :P
19:53 Calinou or…
19:53 Calinou perhaps you can use “git am”
19:53 devmarth I've never heard of that git command
19:53 devmarth What does it do?
19:53 Calinou it merges patches, IIRC
19:54 devmarth I dont want to merge, I just want 45 files, but I guess I'll clone it and manually get it myself
19:54 devmarth So the clone URL is https://github.com/ShadowNinja​/minetest/blob/build-cleanup/ correct?
19:55 ElectronLibre joined #minetest
19:55 devmarth Or I might have to do git:github.com:ShadowNinja/​minetest/blob/build-cleanup
19:55 Calinou no, it's a branch
19:55 devmarth I'm confused O.o
19:55 Calinou git clone https://github.com/ShadowNinja/minetest; cd minetest; git checkout build-cleanup
19:55 ShadowNinja devmarth: You use the regular clone URL, you just check out a different branch than master.
19:56 Calinou this command clones repository, goes to minetest directory, then checkouts the build-cleanup branch
19:56 devmarth Okay. Thanks!
19:56 Calinou ShadowNinja, by the way, branch names should use underscores
19:56 Calinou (this is Minetest convention)
19:56 ShadowNinja Calinou: dashes are my convention.
19:56 devmarth Heh
19:57 devmarth Hm, it did not seem to get the changes from that command
19:57 devmarth It obviously got Minetest but it didnt get the build-cleanup branch
19:58 Calinou did you type that command while in the minetest directory?
19:58 devmarth Yes
19:58 devmarth i typed the same thing you did
19:59 devmarth Oh wait, it did work, but it's the not the same changes as in the pull request he made
19:59 devmarth I checked github manually and under the build-cleanup branch, the changes in the pull aren't identical
20:00 devmarth Wait, I'm dumb, he just updated it
20:00 devmarth That's why
20:02 ShadowNinja I didn't just update it.
20:03 devmarth It's different then last night
20:03 ShadowNinja Oh, then maybe.
20:03 devmarth wait, no, im wrong
20:04 ShadowNinja git fetch ShadowNinja build-cleanup && git checkout build-cleanup && git reset --hard FETCH_HEAD
20:04 devmarth It's not the same changed files from the pull request
20:04 devmarth Okay, lemme try that.
20:05 devmarth woukdnt it be git fetch ShadowNinja minetest && git checkout build-cleanup %% git reset --hard FETCH_HEAD
20:05 devmarth wouldnt*
20:05 devmarth That doesnt work anyways, jesus
20:05 ShadowNinja Nope, git fetch takes a remote name and a branch name, not a GitHub user name and a repo name.
20:06 ShadowNinja If you've got my branch in a seperate folder the remote name will be `origin` instead of `ShadowNinja`.
20:06 devmarth Okay, so whats the command to get your repository and commits (45 file changes) from that pull request?
20:08 ShadowNinja devmarth: From a clone of the main minetest repo: git remote add ShadowNinja https://github.com/ShadowNinja/minetest.git && git pull ShadowNinja build-cleanup && git checkout build-cleanup
20:09 devmarth Okay, thanks
20:11 devmarth fatal: Not a git repository (or any of the parent directories): .git
20:11 devmarth I assume I should do "git init"?
20:11 ElectronLibre Just cd in the directory you cloned (minetest).
20:11 Inspire049 joined #minetest
20:12 devmarth oh nvm
20:12 Inspire049 hey!
20:12 Inspire049 is it just me, i cant get onto inchra.net
20:13 Megaf Inspire049: is down for me too
20:13 Megaf [20:13:14] * Connecting to isis.inchra.net:6667...
20:13 Megaf [20:13:15] * Connection failure: Host not found
20:13 ElectronLibre You can't connect irc.inchra.net port 6667? I'm on it.
20:13 Calinou I am on inchranet
20:13 ElectronLibre Isis is down.
20:13 Calinou isis.inchra.net is down
20:13 Calinou connect to irc.inchra.net
20:14 Calinou this is why you use round-robin addresses :P
20:15 Inspire049 oh so name change?
20:15 Calinou isis was shut down, replaced by osiris
20:16 ShadowNinja Inspire049: Don't use specific server addresses unless you *need* to connect to only that server.
20:32 T4im hm Sokomine, now after updating toa more recent git commit I know what you meant with those fraction of a second freezes :/
20:32 devmarth Okay, so I want to clone ANY git repository, I understand how to do that, but how do I ONLY clone a specific branch?
20:32 devmarth And no, this repository is not mine, I just want to get this specific branch
20:33 T4im git clone -b branch repository
20:33 T4im however, it will still download it all
20:33 T4im well
20:33 T4im not all
20:33 T4im much :)
20:34 T4im there are ways to make a shallow copy to skip the history, but updating from that is messy
20:34 devmarth I tried that
20:34 devmarth git clone -b build-cleanup https://github.com/ShadowNinja/minetest
20:35 devmarth I tried that, but I seem to be getting the master branch
20:35 devmarth Not the build-cleanup
20:35 T4im well after you cloned it, you can use checkout to switch to it
20:35 devmarth git checkout build-cleanup?
20:35 T4im yup
20:35 T4im you can't clone an already cloned repository :)
20:35 devmarth do i have to cd to the clone first
20:35 T4im yea
20:36 T4im all git operations from within the git repository
20:36 T4im otherwise you would have to tell him explicitly what repository you mean
20:37 devmarth i cd'd to the folder and typed "git fetch build-cleanup" and it says "Could not read from remote repository."
20:37 ElectronLibre T4im, you can clone a cloned repository, somewhere else from where you cloned it :).
20:37 T4im ElectronLibre: yes
20:37 devmarth So why is that happening?
20:38 T4im devmarth: you just told him to fetch from a remote called that way.. not from a branch
20:38 shadowzone joined #minetest
20:38 T4im I don't know how you named your remote of that.. if you did nothing its probably called "origin", origin is by default the one you cloned from
20:39 T4im alternativly, just do a git fetch --all ;)
20:39 T4im note that fetching does not updating your working directory
20:39 T4im update*
20:40 T4im it will just retrieve the files and update your git "database" if you will
20:40 ElectronLibre When you clone the remote is origin, so git fetch origin build-cleanup.
20:40 devmarth Okay lemme start from scratch
20:41 T4im no need :D
20:41 T4im you already cloned it…  didn't you?
20:41 devmarth I want to clone http://github.com/ShadowNinja/minetest but JUST the build-cleanup branch
20:41 T4im just do a git clone -b build-cleanup http://github.com/ShadowNinja/minetest then
20:41 T4im but only if you have no directory called minetest already
20:41 T4im ;)
20:41 T4im but no need to clone over and over again
20:42 T4im just use the clone you already have
20:43 devmarth Okay, so it seems I had it right the first time but
20:43 devmarth I want the changes from a pull request, so I assumed it was under this branch
20:43 T4im can I see that pr?
20:44 devmarth I want the 45 changed files from http://github.com/minetest/minetest/pull/2402/
20:44 T4im oh… you don't get just the changed files
20:44 T4im you get minetest with those changed files
20:44 devmarth Yes
20:44 devmarth I want that.
20:44 T4im alright
20:44 devmarth I want the minetest source with the changed files
20:45 T4im always recognize any errors, as they might give a clue where this can go wrong… just go into the directory and do a git checkout build-cleanup
20:45 devmarth so, git clone http://github.com/ShadowNinja/minetest
20:45 devmarth and then git checkout build-cleanup
20:45 devmarth ?
20:46 ElectronLibre Don't forget to cd in it.
20:46 T4im that would work too, or alternativly git clone -b build-cleanup http://github.com/ShadowNinja/minetest
20:46 devmarth Well, that wont work
20:46 T4im both do the same
20:46 devmarth I mean, it will
20:46 devmarth but
20:46 devmarth The changed files from the pul request arent in that branch
20:46 devmarth I tried cloning and checking out the branch, and even looking at it from github, it doesnt contain the changed files from the PR
20:46 ElectronLibre They aren't in master, but if you git checkout they will be placed.
20:47 T4im it does for me
20:47 ElectronLibre And when you clone, the default branch is master.
20:47 T4im ^^ unless you specify -b, then a checkout is done afterwards
20:47 devmarth Goto http://github.com/ShadowNinja​/minetest/tree/build-cleanup
20:47 devmarth It doesnt have the changed files
20:47 devmarth Just look at CMakeLists
20:47 devmarth It's not the same.
20:47 devmarth as from the PR
20:47 T4im "Clean up and tweak build system"
20:48 T4im that's the changed one
20:48 T4im so your problem is that you assume the wrong changes :)
20:48 ShadowNinja devmarth: Um, it does have the changes..
20:48 T4im you might have read that  pr wrong :D
20:48 devmarth thats weird
20:49 devmarth https://github.com/ShadowNinja/minete​st/blob/build-cleanup/CMakeLists.txt
20:49 devmarth THat is not the same as in the PR
20:49 Jordach https://forum.minetest.net/vi​ewtopic.php?p=172099#p172099
20:49 T4im it is de
20:49 T4im it is devmarth*
20:49 T4im you seem to not understand the diff
20:49 T4im the red lines are removed the green ones added
20:50 T4im all the green ones are in that file
20:50 T4im and the white ones
20:50 T4im github not just shows you the file, but the difference of the changes
20:50 devmarth http://github.com/minetest​/minetest/pull/2402/files
20:50 devmarth CHeck that, the files are NOT the same, at least for me
20:51 devmarth Maybe I should clear cache
20:51 Jordach Krock, in my game, everything from stairs to ingot blocks, are made from smaller or larger voxels
20:51 T4im do you have in line 10 string(TOLOWER… ?
20:52 T4im devmarth: https://github.com/ShadowNinja/minetest​/blob/build-cleanup/CMakeLists.txt#L10
20:52 T4im that line is ONLY in the changed file
20:52 T4im and it fits with https://github.com/minetest/minetest/pull/2402/​files#diff-af3b638bc2a3e6c650974192a53c7291R10
20:52 T4im that change
20:53 devmarth in the PR, theres a "+project(minetest)" in the changed fle
20:53 T4im no
20:53 T4im it's -project(minetest)  and +project(Minetest)
20:53 devmarth there is no + or - in https://github.com/ShadowNinja/minetest​/blob/build-cleanup/CMakeLists.txt#L10
20:54 ShadowNinja ..
20:54 devmarth theyre not identical
20:54 ShadowNinja devmarth: You fail to understand what a diff is.
20:54 devmarth I'm very confused
20:54 ShadowNinja devmarth: If you want the patch file it's at https://github.com/minetes​t/minetest/pull/2402.patch
20:54 devmarth theres this: project(Minetest) string(TOLOWER ${PROJECT_NAME} PROJECT_NAME_LOWER)
20:54 T4im yes, that's the changed file
20:54 devmarth and this: -project(minetest) +project(Minetest) +string(TOLOWER ${PROJECT_NAME} PROJECT_NAME_LOWER)
20:55 T4im is the diff
20:55 devmarth Thats the changed?
20:55 T4im no that's the change
20:55 T4im not the changed file
20:55 devmarth I'm a fucking idiot. I was looking at it as if the first was an addition
20:55 devmarth I see now.
20:55 ShadowNinja devmarth: The + and - tells you want's been added and removed.
20:55 T4im "-" tells you a removal of a line, "+" an addition of a line
20:55 devmarth I'm so dumb. I'm sorry.
20:55 T4im well now you got it, that's most important :)
21:01 SylvieLorxu joined #minetest
21:01 bleak_fire_ joined #minetest
21:14 Fritigern joined #minetest
21:36 Warr1024 does installing a texture pack cause a client not to download the overridden textures from the servr?
21:36 Calinou Warr1024, they are always downloaded, so no
21:36 Calinou (in case you stop using a texture pack)
21:36 Warr1024 aw
21:36 Warr1024 ah, I see
21:37 Warr1024 I thought you had to exit to the menu to switch texture packs, which would send you through the download process again...?
21:37 Calinou it would
21:37 Calinou but textures are always downloaded anyway, the server doesn't know you're using a texture pack
21:38 Warr1024 hm, seems like that would be a nice optimization, so long as either texture packs continue to be unchangeable while connected, or media download mid-game is added
21:38 Warr1024 I'm thinking because of the whole "cache zip" thing that's become common practice
21:39 hmmmm https://github.com/minetest/minetest/pull/2459
21:39 Warr1024 but I wanted to keep my base textures 16x16 for fast downloads and accessibility
21:39 Warr1024 but wanted to offer an upscaled texture pack as an alternative
21:39 Warr1024 e.g. so filters could be used without looking too crappy
21:40 Krock hmmmm, are you serious?
21:41 hmmmm Krock:  dead serious
21:43 Krock I don't see any significant problems due to those commits
21:43 VanessaE what the...?
21:43 * VanessaE looks at hmmmm
21:44 hmmmm Krock:  you don't, but others do...
21:44 * Krock is looking through all issues
21:44 hmmmm they introduced a stupendous amount of instability and performance regressions
21:45 hmmmm and seeing as how nerzhul's attitude on the problems is so shitty, this is the only answer
21:46 hmmmm gaining an outsider's perspective, I was able to see what the hell he was trying to do and it's so damn obvious by now
21:46 hmmmm this is absolutely harmful to minetest and it must be fixed
21:46 Krock somehow I notice a political-like left and right side
21:47 hmmmm yeah, there are lot of politics at play here.
21:47 hmmmm i won't deny that.
21:47 hmmmm the one playing politics is a single person though
21:49 hmmmm leaving MT was such a huge mistake that i regret.  I didn't realize things would turn into such a mess
21:54 VanessaE glad you're back
21:54 turtleman_ joined #minetest
21:55 hmmmm no more being pressured into stupidly huge sweeping changes like this
21:55 VanessaE that PR aside, what's your plan?
21:56 hmmmm fix what got F'd
21:56 VanessaE fair enough :)
21:56 hmmmm add a network translation layer so changes like this can be made without being cemented into the revision history
21:58 hmmmm now keep in mind I support *a lot* of the ideas behind the changes made, but their implementation was horrible, the attitude behind the developer supposed to maintain it is horrible, the regressions are horrible
21:59 hmmmm maybe they can be done in a more sane, conservative manner backed by a good implementation made by a developer without hidden intentions
22:01 VanessaE I'll keep out of the politics-related side of things, but I can say for sure that I've seen the regressions...  it's like the 0.4.8->0.4.9 transition when sapier was rewriting parts of the network stack and he had to put all that backward compat code in
22:01 VanessaE at times it just...heh...didn't work, to put it mildly :)
22:02 domtron joined #minetest
22:02 VanessaE I've described same to nrz but I wasn't aware of any plans to properly correct the problems
22:03 hmmmm sapier's changes weren't purposely made difficult to remove since everything used roughly the same interfaces and maintenance was taken seriously
22:03 hmmmm nerzhul explicitly stated that he refuses to fix outright bugs because "they'll be fixed with the next network rewrite"
22:03 hmmmm that's a huuuge red flag right there
22:05 VanessaE yeah, that's a no-go, as I've said before I plan to keep my servers on the 0.4.x compatibility train for a good long while, but the plan apparently with the network stuff has been to push for 0.5.x-break-everything soonish
22:05 hmmmm i don't like that plan to break compatibility.
22:05 VanessaE (I don't fathom losing all my users)
22:06 hmmmm for one it's not necessary, two, there's the forceful attitude behind it
22:06 VanessaE all the work toward that end is supposed to be on a separate branch now, btw
22:06 hmmmm yes
22:06 hmmmm i saw that
22:06 hmmmm that's only after celeron stepped in and told him
22:06 VanessaE yeah
22:06 hmmmm I don't get what the fuck is his problem
22:07 hmmmm this is an open source project and we are very grateful to anybody who contributes, but we are not grateful to petty little dictators stepping in
22:07 VanessaE I think it's a matter of him being some sort of network professional but not having a good handle on the "legacy" needs of this project, maybe?
22:07 hmmmm no, we DON'T need to commit a huge, sweeping network breakage because you insist progress must be made
22:08 hmmmm VanessaE:  I'm a network professional too
22:08 hmmmm I'm not too big on labels
22:08 VanessaE oh, I didn't know that actually
22:09 VanessaE however, the difference is you already know minetest's history
22:09 VanessaE I'm just saying I don't think the guy had bad intentions - just not well-directed.
22:10 hmmmm here are my reasons why I believe that not to be the case:
22:10 hmmmm - he introduced extremely silly errors and suspect design choices
22:10 hmmmm - he insisted his work gets split up and some committed immediately with the said problems
22:11 EvergreenTree joined #minetest
22:11 hmmmm when the more controversial commits were "blocked" in a sense, he made them urgent by making the decision that no fixes will be added in the master branch, only the branch containing the more controversial commits
22:11 hmmmm the vast majority of his actual changes are superficial at best
22:12 hmmmm look at the commit logs - a minor bug fix here, minor bug fix there, file renames/reorganizations, changing instances of std::list to std::vector, etc.
22:12 Salbei joined #minetest
22:12 VanessaE yeah, saw that
22:12 hmmmm things that make a lot of visible change
22:13 hmmmm if you look on ohlol.com or whatever that open source tracker thing is, they rank based off of number of inserts/deletions and commits
22:13 hmmmm he ranks highly in that no doubt because of what I mentioned
22:13 VanessaE I dunno.  zeno and I tested some of the code and it worked -- at first -- and then the problems started showing up once we got enough people connecting and being willing to actually complaion
22:13 Calinou it's called Open Hub now :P
22:13 VanessaE (you know the old rule, 1 person complains for every 10-100 that have the problem)
22:13 Calinou because everyone called it ohlol instead of ohloh
22:14 hmmmm oh, lol.
22:15 hmmmm there was just so much of a push toward making huge sweeping changes because he wanted to make the project his own rather quickly
22:15 hmmmm as one other developer put it (in private to me):  "the entire landscape just keeps changing from day to day"
22:18 Tuxedo[Qyou] joined #minetest
22:19 HoloIRCUser3 joined #minetest
22:25 Menche joined #minetest
22:28 ElectronLibre left #minetest
22:32 DusXMT left #minetest
22:39 cheapie_ joined #minetest
22:42 cheapie joined #minetest
22:43 paramat joined #minetest
22:43 phantombeta joined #minetest
22:51 ShadowNinja I don't support breaking the protocol at this point.  There has to be some real benefit in doing so, and "it's cleaner" doesn't count.
22:52 ShadowNinja But if there is one breaking change that can't be done compatibly and has a real benefit then we might as well break everything and do it right while we're at it.
22:53 ShadowNinja We just have to make sure that the new way is really better and we're not going to want to change it later.
22:53 paramat hmmmm, you have my support for 2459, i trust you
22:54 ShadowNinja I like a lot of the recent network changes.  Of course there were bugs, but I didn't notice an obcene number of them, or a lot of complaints about bugs.
22:55 ShadowNinja Right now I don't know of any real problems with the new network things.
22:55 hmmmm the same changes can be made in a lower-impact manner
22:56 hmmmm performance regressions, instability, outright bugs, etc.
22:56 hmmmm look at this one for example https://github.com/minetest/minetest/issues/2439
22:57 hmmmm to be frank, I would support the changes a lot more if they didn't have an agenda behind them
22:58 hmmmm .. it's about time we take back control of minetest
23:00 ShadowNinja I haven't noticed any performance regressions.  Bugs definitely can't be brushed off like that though.
23:02 hmmmm people who run servers did
23:02 hmmmm those i've talked to counted at least 6 complaints
23:02 VanessaE I wonder though - aside from me and perhaps zeno, who else?
23:02 hmmmm he never told me who else
23:02 hmmmm gregorycu is another one
23:02 VanessaE hm, ok
23:04 hmmmm if I don't get this PR merged, at the bare minimum I would at least like nerzuel to fix his damn bugs and to stop "moving fast and breaking things" because of his flawed interpretation of Agile he's forcing onto the project
23:06 ShadowNinja I don't agree with reverting everything, but I agree that he should fix bugs and avoid breaking changes.
23:11 Hirato joined #minetest
23:11 hmmmm it's bad enough the changes were in right before 0.4.12
23:12 hmmmm celeron55 said: "this is going to be the most stable release of minetest ever"
23:12 hmmmm :(
23:12 VanessaE I have asked at least a couple of times for bugfixes in this realm and for a 0.4.13 release to be made.  my request was...denied.
23:12 iio7 joined #minetest
23:13 hmmmm ShadowNinja:  the problem is that the changes are really too big for any of *us* to solve
23:13 hmmmm this is basically a nerzhel-only job, and he seems dedicated to letting it go
23:13 hmmmm all to apply more force on the giant protocol breakage
23:14 iio7 I was trying to compile Minetest on Alpine Linux which runs with a PAX kernel. LuaJIT doesn't not work on that, I get "PANIC: unprotected error in call to Lua API (runtime code generation failed, restricted kernel?)".. If I compile with plain Lua will there be any performance loss?
23:14 hmmmm iio7:  yes, of course.  plain lua is on average twice as slow, but how much it actually affects you is dependent upon how heavy the mods are you're using
23:15 hmmmm on vanilla minetest you probably won't notice a difference
23:15 iio7 hmmmm, we have a lot of mods :)
23:15 hmmmm it also depends on the contents of those mods... mods that do lots of math and computations IN LUA are going to have a greater slowdown, obviously
23:16 iio7 The problem with LuaJIT on a PAX kernel, is that related only to Lua or has it got something to do with "unsafe" API calls from the Minetest code?
23:16 hmmmm PAX I assume offers protection against self-modifying code?
23:16 ShadowNinja iio7: Try my Lua5.2 branch maybe, I've heard it's faster than 5.1 (although it may or may not actually be so).
23:17 iio7 hmmmm, I believe so.
23:17 hmmmm iio7:  Yeah, that's clearly a problem for anything that does JIT compiling
23:17 hmmmm are you able to disable NX for a specific process?
23:18 iio7 hmmmm, I don't know.
23:18 everamzah joined #minetest
23:19 hmmmm hmmm, you should be able to.
23:20 hmmmm have no idea how to do this though, seems scary and complicated
23:20 * hmmmm runs away
23:22 iio7 hmmmm, yes indeed!
23:22 iio7 So either plain Lua or LuaJIT on another OS which is "less" protected. Perhaps a FreeBSD jail then.
23:46 Viper168 joined #minetest
23:56 Haudegen joined #minetest
23:59 * Sokomine sees hmmmm running away and tries to catch him!

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