Friday, 2012-12-14

*** shmerl_ has quit IRC00:07
*** chebastian has quit IRC00:11
ka6soxyou are assuming we are sober00:13
*** jstaniek has quit IRC00:18
*** faenil has joined #mer00:23
*** M4rtinK has joined #mer00:26
*** phaeron has quit IRC00:33
*** phaeron has joined #mer00:33
*** M4rtinK has quit IRC00:40
*** arcean has quit IRC00:40
*** jayrulez has joined #mer00:41
*** hasselmm__ has quit IRC00:51
*** sonach has joined #mer00:55
*** himamura has joined #mer00:55
*** tanty has joined #mer00:58
*** tanty has quit IRC01:18
*** reels_ has joined #mer01:35
*** reels has quit IRC01:39
*** unifly has left #mer01:42
*** merder2 has quit IRC01:45
*** davidqi has joined #mer01:50
davidqimorning all01:51
*** zengweitotty has joined #mer01:51
*** jooncheol has joined #mer02:14
*** MrBlueSky125 has quit IRC02:14
*** VDVsx has joined #mer02:15
*** Shaan7 has quit IRC02:48
*** Shaan7 has joined #mer02:54
*** VDVsx has quit IRC03:07
*** zengweitotty has left #mer03:08
*** KaiRo_Mozilla has joined #mer03:16
*** Milhouse has quit IRC03:19
*** KaIRC has quit IRC03:20
*** Jade has quit IRC03:24
*** Shaan7 has quit IRC03:33
*** Shaan7 has joined #mer03:33
*** bugzy has quit IRC03:34
*** bugzy has joined #mer03:41
*** davidqi has quit IRC04:00
*** furikku has joined #mer04:01
*** Jade has joined #mer04:01
*** Jade has quit IRC04:01
*** Jade has joined #mer04:01
*** DocScrutinizer05 has quit IRC04:04
*** DocScrutinizer06 has joined #mer04:04
*** tilgovi has quit IRC04:29
*** yashshah has joined #mer04:31
*** yashshah_ has joined #mer04:53
*** yashshah has quit IRC04:54
*** faenil has quit IRC04:55
*** KaiRo_Mozilla has quit IRC05:30
situMorning everyone05:44
*** Shaan7 has quit IRC05:47
*** Beineri_ has quit IRC05:48
*** Beineri_ has joined #mer05:49
*** Shaan7 has joined #mer05:53
*** phinaliumz has joined #mer05:56
*** Figure_ is now known as Figure05:56
*** beford has quit IRC06:04
*** beford has joined #mer06:05
*** yashshah_ has quit IRC06:16
*** spiiroin has joined #mer06:21
*** melonipoika has joined #mer06:26
*** rdqfdx has joined #mer06:28
*** jjarven has quit IRC06:28
*** zalan has joined #mer06:32
*** zchydem_ is now known as zchydem06:32
*** phinaliumz has quit IRC06:34
lpotterzzzz... autotest06:39
Stskeepsmorn lpotter06:39
Bostikwhat do you call it when everything imaginable and then a few things beyond imagination go wrong?06:42
StskeepsBostik: monday?06:42
lpottermurphys law06:42
BostikI doubt even mr. Murphy could have foreseen this one06:42
Bostikbut yes, eternal monday sounds pretty accurate06:43
Bostiklpotter: I *did* record this in the project progress  - "Mr. Murphy, I submit to your tenacity."06:47
ka6soxI'd call that a normal day...06:55
*** imunsie has quit IRC07:00
*** e8johan has joined #mer07:07
*** plfiorini has quit IRC07:15
*** edgar2 has joined #mer07:18
*** veskuh has joined #mer07:19
*** SfietKonstantin has joined #mer07:19
*** Adityab has joined #mer07:22
*** himamura_ has joined #mer07:23
*** himamura has quit IRC07:23
*** kallela has joined #mer07:27
*** niqt has joined #mer07:28
*** pirut has joined #mer07:28
*** e8johan has quit IRC07:29
*** davidqi has joined #mer07:37
davidqihi , I am now gonna run mer on qemu, the filesystem seems mounted, but met following error07:39
davidqiStarting UX launcher; Started UX launcher; Stopping UX launcher... ; Stopped UX launcher.07:40
davidqiStarting UX launcher; Started UX launcher; Stopping UX launcher... ; Stopped UX launcher.07:40
davidqiagain and again07:40
Stskeepsdavidqi: what qemu? arm? x86?07:40
davidqithen: Reached target Graphical Interface.07:41
Stskeepsok, show me the .ks you're using?07:42
davidqiwait a second07:43
davidqimine is mer-core-armv6l-xorg-basic-qmlviewer.ks07:44
davidqithe standard one07:44
Stskeepsokay, log in as root and systemctl stop uxlaunch.service07:45
Stskeepsthen look at journalctl07:45
Stskeepsand what kernel?07:45
Stskeepsis in use07:45
davidqi3.0.4, I can not log in07:45
Stskeepstry mer/rootme07:45
Stskeepsand then su to root07:46
davidqiI can not reach the bash shell07:46
davidqithe system hangs at the "Reached target Graphical Interface"07:46
*** lamikr has joined #mer07:47
davidqiI used following commands:07:49
davidqisudo qemu-system-arm -M versatilepb -cpu arm1176 -m 256 -hdc rootfs1.ext2 -kernel zImage.3.0.4.kun -append "root=0800 rw init=/sbin/init" -serial stdio07:50
davidqito start the qemu07:50
Stskeepsok, you need to make sure that your zImage has the right kernel settings like
Stskeepsthe script tests your .config07:51
veskuhIs it possible that enabling HW virtualization actually makes my I/O performance in VirtualBox actually slower?07:53
Stskeepsveskuh: it might be, but the allocated disk space if it's dynamic may also be it07:53
davidqiyes, I have verified my config with mer_verify_kernel_config07:54
veskuhStskeeps, yeah, don't know about that. Using ready made SDK image. In anycase I/O perf is absolutely horrible07:54
Stskeepsveskuh: is the sdk image made with VBoxManage or qemu-img?07:55
Stskeepsjust troubleshootin07:55
veskuhStskeeps, I think VBoxManage.07:55
*** pvanhoof has joined #mer07:56
*** SfietKonstantin has quit IRC07:56
veskuhStskeeps, with make -j8, I'm able to only use 25% of CPU07:56
Stskeepsdo you allocate multiple CPUs to the VM?07:56
Stskeepsthere's a setting for that07:56
veskuhYes, 4 and HW virtualization enabled07:56
Stskeepsis this with shared home folder?07:56
veskuhStskeeps, yes, that is the setup we have for VM SDK07:57
Stskeepshow much ram?07:57
Stskeepsand does gcc use -pipe ?07:57
*** Beineri_ has quit IRC07:58
*** Beineri_ has joined #mer07:58
phaeronyou could run a simple dd benchmark just to localize the issue07:58
*** SfietKonstantin has joined #mer07:58
*** Beineri_ has quit IRC07:58
*** rcg-work has joined #mer07:58
*** Beineri_ has joined #mer07:58
*** veskuh has quit IRC07:59
*** mvogt has quit IRC08:00
*** veskuh has joined #mer08:00
Stskeepsand does gcc use -pipe ?08:01
*** pvilja has quit IRC08:01
veskuhI've given 4 GB of RAM to VM but top only shows 800MB for some reason (well uses 200MB)08:02
StskeepsCONFIG_HIGHMEM=y might be useful in the kernel08:04
*** pvilja has joined #mer08:04
*** gabriel9|work has joined #mer08:05
veskuhlbt, ^ for the SDK image08:05
veskuhStskeeps, and yes gcc uses -pipe08:05
Stskeepsother ideas: ext2 file system in vm08:07
Stskeepsisn't there like a VM I/O monitor in virtualbox/s08:07
Stskeepsso you can see where the traffic goes08:08
Stskeepsalso, it might entirely be that you don't have cache for the shared home folder stuff08:08
*** CosmoHill has joined #mer08:12
*** calvaris has joined #mer08:14
*** denexter has quit IRC08:15
*** DocScrutinizer06 is now known as DocScrutinizer0508:18
*** calvaris has quit IRC08:18
*** slaine has joined #mer08:18
*** calvaris has joined #mer08:18
*** hasselmm has joined #mer08:20
*** tanty has joined #mer08:22
veskuhWith single core for the VM the I/O seems better at least compile feels faster08:22
*** tanty has quit IRC08:22
veskuhand the CPU is not the bottleneck08:22
*** jpetersen has joined #mer08:24
veskuhyep, the compile time was something like half with just single core enabled for the VM08:25
phaeronveskuh: are you sure the kernel has proper multi core support ?08:26
phaeronI think x86 kernel was made for netbooks and tablets08:27
*** blam_ has quit IRC08:27
phaeronthat's why it doesn't have highmem support08:27
Stskeepswell, even netbooks have hyperthreading08:27
Stskeepshighmem makes sense due to performance08:27
Stskeepsbut most today have >= 1gb08:27
veskuhphaeron, no idea I'm just using the SDK VM image. At least /dev/cpu has the cores I defined08:28
*** melonipoika has quit IRC08:34
*** melonipoika has joined #mer08:34
*** Shaan7 has quit IRC08:35
*** denexter has joined #mer08:35
*** Shaan7 has joined #mer08:38
CosmoHillyou know qt puts the tests in /usr/tests08:38
CosmoHillis there a way to get it to install them somewhere else?08:38
Stskeepsqt4 or 5?08:39
Bostiksounds like qt5, I'd say08:39
Stskeepsprobably a bad ./configure thing08:40
BostikCosmoHill: I'm betting on a missing -testdir option for ./configure08:40
Bostikor was it -testsdir (note the extra 's')08:41
Bostikeither way... some variable used for install paths is unset and does not get any sensible default (remains empty) -> installation path of FOO goes wrong08:42
CosmoHillmy friend says there's no testdir or testsdir option in the configure08:43
CosmoHillhe says the path seems hardcoded08:43
Bostikif [ -z "$QT_INSTALL_TESTS" ]; then #default08:44
Bostik    QT_INSTALL_TESTS="$QT_INSTALL_PREFIX/tests" #fallback08:44
Bostikokay, from qt5 but looks exactly the same08:45
CosmoHillwhat file is that from?08:46
Bostikso it's probably not hardcoded but it IS set at *Qt build time08:46
CosmoHillso he'd  set QT_INSTALL_TESTS at configure08:47
*** jstaniek_QFridge has joined #mer08:48
*** melonipoika has quit IRC08:48
*** SfietKonstantin has quit IRC08:50
*** SfietKonstantin has joined #mer08:50
CosmoHillthank you :)08:51
CosmoHillah, turns out it's not there in his qt408:52
Stskeepsseen this before?08:53
*** plundstr has left #mer08:54
*** Martix has joined #mer08:55
BostikStskeeps: yes, I have and haven't found a reason for it; it appeared at -RC1 for modules that did not have the -rc1 tag, and it's probably there still08:55
Stskeepswell, this is from tarball08:56
BostikStskeeps: haven't had opportunity to sync with rc2 yet, I'll probably hit it too08:56
*** davidqi has quit IRC08:56
*** yunta has joined #mer08:59
*** sonach has left #mer09:05
*** stefan_schmidt_w has joined #mer09:07
*** pvilja has quit IRC09:15
*** tilgovi has joined #mer09:17
*** pvilja has joined #mer09:18
*** reels_ is now known as reels09:19
*** niqt has quit IRC09:21
*** plundstr has joined #mer09:23
*** e8johan has joined #mer09:24
*** fk_lx has joined #mer09:24
*** rabit has joined #mer09:29
* lbt afk all morning ... bbl8r09:30
*** alterego_ is now known as alterego09:32
*** trbs has joined #mer09:34
*** CosmoHill has quit IRC09:36
*** e8johan has quit IRC09:36
*** notmart has joined #mer09:40
*** notmart has joined #mer09:40
*** oahong has quit IRC09:48
*** blam_ has joined #mer09:55
*** pocek_ is now known as pocek09:55
*** oahong has joined #mer09:55
*** thetet has joined #mer09:56
*** himamura has joined #mer09:59
*** otep has quit IRC09:59
*** MrBlueSky125 has joined #mer10:00
Stskeepsphaeron: merged gzip10:00
*** himamura has quit IRC10:01
*** CosmoHill has joined #mer10:01
phaeronStskeeps: saw thanks10:01
*** himamura_ has quit IRC10:01
*** otep has joined #mer10:02
*** auri__ has joined #mer10:03
*** arcean has joined #mer10:04
sledgesgood -Friday- morning!10:09
CosmoHillhi sledges10:10
*** fk_lx has quit IRC10:14
*** faenil has joined #mer10:19
rabitgood morning!10:20
rabiti'd like to create a newbies guide into mer/nemomobile/sailfish development, based the information I gratefully received from some of you through chats in the past week. the idea is to make it as easy as possible for developers with app development experience to get started writing their first mobile app using qt/qml/qt quick.10:22
leinirsounds like a great idea to me! :)10:23
rabiti've been looking into tizen before, and got attracted to mer by the jolla demo of sailfish10:24
*** cxl000 has joined #mer10:24
rabitbut at the moment it's relatively hard to go through various websites and wikis to get the relevant information to get started with mer/nemomobile or future sailfish development.10:25
phaeronStskeeps: I want to server git master, stable (latest - 1), latest , next from one mds. not sure if that's possible right now10:25
rabitthe document/page should contain installation instructions for the sdk/tools/qt creator and libraries needed for nemomobile and sailfish.10:25
Stskeepsphaeron: :nod:10:25
rabitincluding a list of hardware which has been used (existing smartphones or developer boards).10:26
Stskeeps is a prototype i worked on ,         - is the http server, it uses gitmds2 module as 'backend'10:26
rabitwhere should such a page/document/pdf be hosted?10:26
Stskeepsfor mer/nemo, on wiki.merproject.org10:26
phaeronStskeeps: so not doable with mds1 ?10:27
Stskeepsphaeron: the principle is to not have individual core checkouts, but just the git repo of core10:27
Stskeepsphaeron: we might as well fix the problem at it's root10:27
*** lizardo has joined #mer10:27
rabitdoes the wiki support generation of PDF out of a page?10:27
Stskeepsphaeron: so when you do Core:master:i486 it looks at Core's git repo, finds master branch, and i486 dir in it10:28
*** gabrbedd has quit IRC10:29
Stskeepsor tag, or what ever git reference instead of master10:29
phaeronI see10:29
rabitAnd the Jolla folks sent me here with another questions (from Twitter). which development board would you recommend for mer/nemomobile/sailfish development? Is this a good board?
*** fk_lx has joined #mer10:31
Stskeepsphaeron: so that may make a good basis - what do you think?10:31
phaerontrying to wrap my head around it10:31
rabitthe newbie guide should contain the instructions to get started developing using a freely available hardware, which one would you recommend (besides phones like N9 or 900/950?)10:32
phaeronok so first : this is mds1 split into two modules. one for handling git backend and one is fake obs interface10:32
*** himamura has joined #mer10:33
Stskeepsmds1 is split up similarly, for good measure - and gitmer.py10:33
Stskeepsand yes, you're right10:33
phaeronthe git handling is as you said , doing fetches based on ref instead of just current head10:33
phaeronwhich means make file approach doesn't need to do git checkout whatever10:34
Stskeepsyes, just rsync down10:34
phaeronStskeeps: what about the obs-repos part ; the binary repos10:35
Stskeepswell, you'd look those up like previously, rsync obs-repos/Core:version:i486 down, serve files10:35
phaeronand normal symlinks where :version: is == refname10:36
Stskeepsfor example, yeah - right now :latest points at the right version (in obs-repos/ ) with a symlink10:37
*** Martix has quit IRC10:39
phaeronso what's missing / where do I start poking10:39
*** chebastian has joined #mer10:40
rabitis there a list of development boards supporting all mer/nemomobile features right now?10:40
Stskeepsi think there's some obs api things that are missing, but i would start with a rsync of project-core and play around with curl towards the api and get a feel for it10:40
*** Jay_BEE has joined #mer10:40
Stskeepsand see what needs to be done10:41
Stskeepsi'd start from mds2 as above as it's a little more documented and cleaned up10:41
Stskeepsand throwing things at me when stuff is obviously broken10:41
phaeronok .. now the punchline. is this doable so that I can switch to using it today / tomorrow at most ?10:42
Stskeepsif you say two-three days instead, we'd have a solution that'd both speed up mer and any use of mds, as well give us much needed flexibility in future :P10:43
*** calvaris has quit IRC10:43
Stskeepsas well as less moving parts/makefile insanity10:44
phaeronand that includes the other stuff we talked about ? :D10:45
*** calvaris has joined #mer10:45
phaeronnah let's not change too many things at once10:45
Stskeepsok, but we'll have to do this at some point anyway10:47
Stskeepsso, you can do individual git checkouts of Core (with mds1) with the above and point to them in mappings.xml10:48
Stskeepsyou'll want to do it in such a way that upstream == packages-git/mer/project-core, and you pull and check out from that10:48
phaeronnow you lost me10:49
Stskeepswell, you said we're not going to make mds2, soi'm returning to your orgiinal question10:49
Stskeeps[11:25] <phaeron> Stskeeps: I want to server git master, stable (latest - 1), latest , next from one mds. not sure if that's possible right now~10:49
Stskeepsthe way to do this is to make individual git checkouts of packages-git/mer/project-core for those branches and point to them in mappings.xml of mds110:51
phaeronmy comment was about the change of not having tarballs in git10:51
Stskeepsthat isn't part of this at all :)10:51
* w00t hides10:51
phaeronbut I want to use mds2 if I can get it to shape in ~2 days10:52
Stskeepsok, how about we do it this way: you make a test deployment of mds2, i help and if it doesn't look sane in 4 hours, we drop the case?10:53
*** ScriptRipper has quit IRC10:55
Stskeepsmake a checkout of mer/release-tools and drop the two files above into tools/ and remove the .txt extension10:55
Stskeepsdon't try to 'make' or anything like that :)10:55
Stskeepsif you have a copy of a full packages-git/ rsync, that'd be good to put in too10:56
*** SfietKonstantin has quit IRC10:58
Stskeepswe'll first get sources working and then we'll get binary repos working10:58
Stskeepsie, success criteria is that we can check out Core:0.20121205.0.2:i486 acl package10:59
Stskeepswithout having individually checked out Core:0.20121205.0.210:59
*** ScriptRipper has joined #mer11:00
*** VDVsx has joined #mer11:04
Stskeepsand we need a mappings.xml like:11:06
Stskeeps<mapping project="Core" path="packages-git/mer/project-core"/>11:06
*** Jade has quit IRC11:09
*** Jade has joined #mer11:09
*** Jade has joined #mer11:09
*** thopiekar has joined #mer11:09
phaeronStskeeps: still setting up11:10
phaeronwill ping you when ready11:10
*** fk_lx has quit IRC11:20
*** thopiekar has quit IRC11:24
*** VDVsx has quit IRC11:25
*** VDVsx has joined #mer11:25
*** niqt has joined #mer11:36
*** niqt has quit IRC11:36
*** VDVsx_ has joined #mer11:38
*** VDVsx has quit IRC11:40
*** icota has joined #mer11:43
*** rabit1 has joined #mer11:46
*** rabit has quit IRC11:48
*** faenil has quit IRC11:51
*** trip0 has quit IRC11:57
*** trip0 has joined #mer11:58
*** zhxt has joined #mer11:59
*** Shaan7 has quit IRC12:00
*** NIN101 has joined #mer12:03
*** aportale has joined #mer12:07
*** Shaan7 has joined #mer12:08
BostikStskeeps: with RC1->RC2 fixes to qtbase, qtdeclarative builds with RC1 spec; no modifications needed12:11
*** tilgovi has quit IRC12:11
Stskeepsqtquick1 was the one12:11
BostikI looked at the monster during a longer build...12:11
Bostikthere are others, believe me...12:11
Stskeepsi'm still pondering a strategy for those that have no tags and no release12:12
*** blam_ has quit IRC12:12
*** ridikulus_rat has joined #mer12:20
*** VDVsx_ is now known as VDVsx12:21
*** niqt has joined #mer12:22
*** ridikulus_rat has left #mer12:22
situStskeeps: Heard about Yota ?12:24
situand they're focussing on  just e readers ?12:26
*** Jay_BEE has quit IRC12:26
situI mean smartphones with focus on e readers.12:27
*** VDVsx_ has joined #mer12:30
*** VDVsx has quit IRC12:32
*** rabit1 has left #mer12:33
*** niqt has quit IRC12:35
*** merder2 has joined #mer12:37
*** dmch has joined #mer12:43
sledgeswow, Yota quite original12:44
*** jjarven has joined #mer12:47
*** VDVsx_ has quit IRC12:49
*** chriadam is now known as chriadam|away12:50
*** thopiekar has joined #mer12:51
BostikStskeeps: I have a very vague hunch on what *might* cause the ".qmake.conf" error, but this needs some testing still12:51
*** phinaliumz has joined #mer12:53
*** Jaded has joined #mer12:55
*** Jaded has quit IRC12:55
*** Jaded has joined #mer12:55
*** himamura has quit IRC12:56
*** Jade has quit IRC12:57
*** fk_lx has joined #mer12:58
*** KaIRC has joined #mer12:59
*** gabrbedd has joined #mer13:06
*** disco_stu_droid has joined #mer13:10
*** disco_stu has quit IRC13:13
*** disco_stu_droid is now known as disco_stu13:14
*** sp3000 has quit IRC13:18
*** sp3000 has joined #mer13:19
*** Jaded has quit IRC13:24
*** Jaded has joined #mer13:27
*** Jaded has quit IRC13:27
*** Jaded has joined #mer13:27
*** Jaded has quit IRC13:31
*** yashshah_ has joined #mer13:33
*** yashshah_ has quit IRC13:35
*** yashshah_ has joined #mer13:36
*** merder2 has quit IRC13:37
*** Guest80585 has joined #mer13:40
*** mpthx has joined #mer13:43
*** yashshah_ has quit IRC13:43
*** thetet has quit IRC13:45
*** niqt has joined #mer13:47
*** Jucato_ has joined #mer13:48
*** Jucato has quit IRC13:48
*** Jucato_ is now known as Jucato13:49
*** yashshah_ has joined #mer13:56
*** niqt has quit IRC13:56
*** yashshah_ has quit IRC14:04
*** denexter has quit IRC14:07
*** yashshah_ has joined #mer14:07
*** ian--- has joined #mer14:12
*** ian---- has quit IRC14:13
*** yashshah_ has quit IRC14:14
*** Guest80585 has quit IRC14:15
*** Jaded has joined #mer14:17
*** Jaded has quit IRC14:17
*** Jaded has joined #mer14:17
*** yashshah_ has joined #mer14:19
*** FSCV has joined #mer14:21
*** pvanhoof has quit IRC14:25
*** Jaded has quit IRC14:26
*** Jaded has joined #mer14:27
*** Jaded has quit IRC14:27
*** Jaded has joined #mer14:27
*** yashshah_ has quit IRC14:31
*** zenvoid has joined #mer14:31
*** yashshah_ has joined #mer14:31
*** yunta has quit IRC14:35
*** yashshah_ has quit IRC14:40
*** VDVsx has joined #mer14:42
*** plundstr has left #mer14:44
*** yashshah_ has joined #mer14:45
*** crevetor has quit IRC14:51
*** Superpelican has joined #mer14:53
*** spre_ has quit IRC14:55
*** mihero_ has quit IRC14:56
*** jayrulez has quit IRC14:59
*** yashshah- has joined #mer15:03
*** yashshah_ has quit IRC15:03
*** sp3000 has quit IRC15:04
*** kallela has quit IRC15:09
*** Superpelican has quit IRC15:12
*** Jaded has quit IRC15:15
*** Jade has joined #mer15:16
*** Jade has quit IRC15:16
*** Jade has joined #mer15:16
*** Jade has quit IRC15:19
*** CosmoHill has quit IRC15:20
*** icota has quit IRC15:21
*** icota has joined #mer15:21
*** CosmoHill has joined #mer15:21
*** Jade has joined #mer15:24
*** Jade has quit IRC15:24
*** Jade has joined #mer15:24
*** Superpelican has joined #mer15:25
*** jayrulez has joined #mer15:29
*** pvilja1 has joined #mer15:30
*** rcg-work has quit IRC15:30
Stskeepsphaeron: (context switched) - does Core:master:i486 acl work?15:31
Stskeepslatest isn't actually a git ref in mer, so15:31
*** yashshah- has quit IRC15:31
*** yashshah- has joined #mer15:32
*** pvilja has quit IRC15:32
phaeronStskeeps: no15:32
phaeronwill have to debug15:32
Stskeepsok, also, osc won't do - it uses / api, not /public15:33
Stskeeps(i think)15:33
phaeronmm possible15:33
Stskeepscurl http://localhost:8080/public/source/Core:master:i486/acl15:33
Stskeepser, 800015:34
phaeronosc -A api /public/source/Core:master:i486/acl15:34
phaeronKeyError: "Blob or Tree named 'i486' not found"15:34
*** himamura has joined #mer15:34
Stskeepsthat keyerror is from mds2 i presume?15:34
phaeron  File "/var/tmp/mer-release-tools/tools/", line 81, in get_project15:35
*** yashshah__ has joined #mer15:36
phaerongit contains source only so it won't have archs in it15:36
*** yashshah- has quit IRC15:36
*** yashshah__ is now known as yashshah15:37
Stskeepsno, this is actually supposed to be a dir :)15:37
Stskeepsjust wondering why it can't find that15:37
phaeronbecause I didn't download it15:38
phaeronthought it needed only mer-core git tree15:38
SuperpelicanThis may be a stupid question15:39
SuperpelicanBut does the N900 support UHS-1 sd cards?15:39
SuperpelicanI want to purchase a new micro SD card specially for running Nemo Mobile on my new N900 :)15:39
Stskeepsi don't actually know15:40
SuperpelicanUHS-1 is suppose to be backwards compatible, right?15:40
SuperpelicanI currently choosing between a Sandisk Class 4 8GB and a Sandisk Mobile Ultra UHS-1 8 GB15:41
Stskeepsjust get a class 10 and save the money15:41
SuperpelicanBoth seem to have good random io speeds15:41
SuperpelicanI read classes don't matter, when you're running OS of a SD card15:42
Stskeepsclass 10 does matter15:42
SuperpelicanAnd that random read and write are the things that matter15:42
*** yashshah has quit IRC15:42
SuperpelicanStskeeps:So you think the Mobile Ultra will work?15:43
Stskeepsi don't know15:43
phaeronStskeeps: packages-git/mer/project-core/ is a bare git repo, is that correct ?15:43
SuperpelicanI currently have a class 4 8gb Kingston sd card15:43
Stskeepsphaeron: correct15:43
SuperpelicanBut lots of Kingston cards seem to have pretty bad random io15:43
SuperpelicanSo I'm thinking of selling my current card together with my old Android phone15:44
Stskeepsphaeron: no need for checkout15:44
SuperpelicanWhich has become obsolete, thanks to my awesome N900 :P15:44
*** rubdos has joined #mer15:44
phaeronStskeeps: not working15:46
Stskeepsphaeron: okay, and you have mappings.xml that contains only what i had above?15:46
phaeronagain my bad15:46
* phaeron is a bit tired15:46
Stskeepssame here, no need to apologize :)15:47
*** thopiekar_ has joined #mer15:47
*** pirut has quit IRC15:47
phaeronnow works15:47
phaeronand with tag ref it works15:49
*** zhxt has quit IRC15:49
*** yashshah has joined #mer15:50
phaeronStskeeps: so it works and shows diffferent content for qt in master and qt in an old tag for example15:50
Stskeepsdo you have mappingscache.xml in packages-git/ btw?15:51
*** thopiekar has quit IRC15:51
Stskeepsthat's one thing i kinda wouldn't mind to pre-generate somehow, right nwo you have to re-generate it after each rsync of packages-git15:51
Stskeepsbut that's just a general comment15:51
*** rubdos has quit IRC15:52
phaeronI asked about that earlier and you said there were "issues" ? :D15:52
*** ced117 has joined #mer15:52
*** ced117 has joined #mer15:52
Stskeepscould you somehow time how quickly you can get the index?15:53
Stskeepsfor various packages15:53
*** rubdos has joined #mer15:53
phaeronwith or without mappings cache ?15:53
Stskeepswith, it doesn't operate without15:54
Stskeepstiming the /public/source/Core:blahblah:i486/acl stuff15:54
Stskeepsbecause it's one of the most common operations15:55
phaeronunder a scond15:56
phaeronindex of random tag of qt ~1s15:57
Stskeepscan we try with 'time'?15:57
phaeronthis is with time15:57
Stskeepsi'd really like to get that lower, with 300+ packages it's 300 seconds15:57
phaerontime curl http://localhost:8000/public/source/Core:0.20120816.0.1:i486/qt15:58
phaeronnext try 0m0.636s15:58
phaeronthird 0m0.577s15:58
Stskeepsis there such a thing as profiling python?15:59
Stskeepsto see where it spends time15:59
*** RedMage_ has joined #mer16:00
*** gabriel9|work has quit IRC16:00
Stskeepsjust thinking that if we have to do it right, we might as well do it fast, too :P16:00
phaeronso obs lists them sequentially ?16:01
Stskeepsyeah, it's stupid like that.. it goes and asks for every single package16:01
Stskeepsoften 3-4 times16:01
phaeronmaybe we should fix obs ?16:01
Stskeepsanyway, python -m cProfile tools/mds2.py16:02
Stskeepsand ctrl^c after a couple of tries16:02
Stskeepsbrb 40 min, have to eat16:04
*** LoCusF has quit IRC16:04
phaeronmost of the time is spent in sre_compile and sre_parse which afair are due to xml parsing16:06
phaeronprobably switching to lxml from standard xml would be better16:06
*** yashshah has quit IRC16:07
phaeronhmm or before that , reorganizing the code so it actually caches the parsed dom16:08
*** LoCusF has joined #mer16:10
phaeronit seems you already cache it16:11
Superpelicanstskeeps: "Class 10 = Fast transfer speeds and slower random I/O"16:14
SuperpelicanStskeeps: "Class 4-6 = Slower transfer speeds and faster random I/O"16:15
SuperpelicanLots of more people saying this on other websites16:15
SuperpelicanI'll be using my micro sd only for running Nemo Mobile16:16
SuperpelicanAs 32 GB MMC of N900 is plenty of storage for me16:16
SuperpelicanSo transfer speeds don't really matter16:16
*** e8johan has joined #mer16:16
phaeronStskeeps: I'll try with lxml16:17
*** jstaniek_QFridge has quit IRC16:17
*** mihero has joined #mer16:18
*** veskuh has quit IRC16:23
SuperpelicanStskeeps: "An SDHC UHS-I card will work in any SDHC compatible device at lower speeds, but to take advantage of the UHS-I speed, a UHS-I compatible device is essential."16:25
SuperpelicanStskeeps:From the Integral (a SD manufacturer) website16:26
SuperpelicanSo that Sandisk Mobile Ultra should work16:26
SuperpelicanAlthough I just read it's class 6, not class 1016:26
SuperpelicanBut that doesn't really matter16:26
SuperpelicanOn the RPi forums there were benchmarks of SD cards16:27
SuperpelicanAnd Sandisk Mobile Ultra was lightning fast there16:27
Superpelican(random io)16:27
*** jayrulez has quit IRC16:28
*** phdeswer has quit IRC16:29
*** stefan_schmidt_w has quit IRC16:29
SuperpelicanRPi benchmarks:
*** jayrulez has joined #mer16:32
*** auri__ has quit IRC16:35
*** pvilja1 has quit IRC16:36
*** schmooster has quit IRC16:42
*** yashshah has joined #mer16:44
*** chebastian has quit IRC16:47
*** e8johan has quit IRC16:52
*** jayrulez has quit IRC17:01
thauta_is there a list of supported bluetooth profiles for mer?17:03
*** faenil has joined #mer17:03
Stskeepssame as bluez?17:03
*** yashshah has quit IRC17:04
*** reels has quit IRC17:05
thauta_hmm, so mer doesnt have any specific configuration files/options installed?17:05
Stskeepscorrect, it's usually up to the vendor17:05
thauta_alrighty then17:05
faenilo/ people17:09
*** yashshah has joined #mer17:12
*** calvaris has quit IRC17:14
*** calvaris has joined #mer17:14
*** RedMage_ has quit IRC17:18
*** pvanhoof has joined #mer17:20
*** Jade has quit IRC17:21
*** _aegis__ has joined #mer17:25
*** _aegis__ has quit IRC17:28
*** AndreasWatch has joined #mer17:31
*** yunta has joined #mer17:32
*** alh has quit IRC17:33
*** yunta has quit IRC17:41
*** lpotter-home has joined #mer17:41
*** lpotter has quit IRC17:41
phaeronStskeeps: converted most of minidom usage  to etree (lxml), no 500ms or less17:42
phaeronwill profile more17:42
*** bigbluehat_ is now known as bigbluehat17:44
*** jwhite has quit IRC17:45
*** lamikr has quit IRC17:49
*** phinaliumz has quit IRC17:50
*** Aurium has joined #mer17:54
*** lpotter-home is now known as lpotter17:55
*** MrBlueSky125 has quit IRC17:56
*** plfiorini has joined #mer17:58
* Superpelican is transferring his music collection to his N90018:00
* Superpelican thinks openssh/scp and unix rock18:00
*** jpetersen has quit IRC18:01
*** fk_lx has left #mer18:02
*** yunta has joined #mer18:06
*** thopiekar_ has quit IRC18:07
*** icota has quit IRC18:10
*** AndreasWatch has left #mer18:15
*** VDVsx has quit IRC18:16
*** jayrulez has joined #mer18:17
*** VDVsx has joined #mer18:17
phaeronStskeeps: there ?18:19
*** ali1234 has joined #mer18:20
Stskeepsphaeron: wife keeping me from my computer18:20
phaeronStskeeps: most packages in about 50ms , qt in 400ms18:23
Stskeepswhat'd you do?18:23
phaeronreplace all minidom with etree18:23
phaerontime for i in `ls packages-git/mer-core/ -1 | xargs` ; do curl http://localhost:8000/public/source/Core:master:i486/$i ; done18:24
Stskeepsthat usually takes ages18:24
phaeronfull ls in 20 second18:24
Stskeepssee, i'm a lousy coder ;)18:25
Stskeepscan you see how it behaves with 6 at a time hitting it?18:25
phaeronhmm it is supposed to be threaded , right ?18:26
phaeronStskeeps: so I fork bomb my laptop :D18:26
phaeronwell at least I am not offline18:27
*** popey has quit IRC18:27
phaeronwent fine with almost all requests in parallel18:28
phaeronwell I ran out of sockets somewhere in the middle18:28
*** popey has joined #mer18:30
phaeronetree has much clearer syntax compared to minidom. most operations are one line vs. 3 or 4 originally18:30
*** aportale has quit IRC18:30
Stskeepsok, i was wondering on time of completion under stress18:32
phaeronI think it was roughly linear , as in didn't change18:32
phaeronbut this is an ssd , so no seek time18:32
*** Superpelican has quit IRC18:34
*** pvanhoof has quit IRC18:36
phaeronStskeeps: can push a new branch18:44
* phaeron mumbles something about gerrit18:44
phaeron ! [remote rejected] HEAD -> refs/for/mds2 (branch mds2 not found)18:46
*** two3four has joined #mer18:46
Stskeepsjust send it to master i guess, we can build on top18:48
phaeronI can modify that changeset , right ?18:49
Stskeepsyes, amend commit with changeid18:50
Stskeepscant see it right now, on n95018:50
*** faenil has quit IRC18:54
Stskeepsor just stack changes on top18:58
*** jpetersen has joined #mer19:04
*** slaine has quit IRC19:05
phaeronStskeeps: so what's next19:06
*** jayrulez_ has joined #mer19:07
*** jayrulez has quit IRC19:08
*** notmart has quit IRC19:18
*** Adityab has quit IRC19:18
*** alh has joined #mer19:21
*** ka6sox is now known as ka6sox-away19:23
Stskeepsphaeron: moment, will look at code19:25
Stskeepscan you access files within the package too? like /acl.spec ?19:26
phaeronStskeeps: yes19:28
SageStskeeps: ping SR#7490 in cobs19:28
StskeepsSage: ok19:29
Stskeepsi'll get to that at some other point than tonight, i have some debt from the qa tests atm too19:29
Stskeepsphaeron: ok, so far so good..19:29
*** mdfe has joined #mer19:30
*** Martix has joined #mer19:31
*** Jucato has quit IRC19:31
Stskeepsphaeron: ok, so, thinking a bit about intelligence.. should we have api /update that rsyncs packages-git and updates mappings cache?19:32
*** Jucato has joined #mer19:32
*** M4rtinK has joined #mer19:33
phaeronStskeeps: inside the python code ?19:33
Stskeepswell, instead of having to go and run mystical makefiles - the reason i'm bringing it up, is that it might also be sane to have logic on what releases to follow, or ability to grab a certain binary release, etc19:34
Stskeepsso it's a more intelligent solution19:35
*** faenil has joined #mer19:35
phaeronbut fakeobs api is facing to obs and that won't use such api19:35
phaeronor are you thinking on demand ?19:35
Stskeepswell, i was thinking more like for example a crontab kicking it, or you running a command to request an update19:36
Stskeepsso i thought a http command might make sense instead of a seperate command socket19:36
Stskeepslike curl http://fakeobs:8000/update19:36
*** icota has joined #mer19:36
Stskeeps(sorry if i'm not making sense, bit tired)19:38
phaeronwell .. ok . if it is preconfigured to follow certain repos and it won't need to be restarted19:38
phaeronold mds spun at 100% cpu when you updated stuff while it was running19:39
*** CosmoHill has quit IRC19:39
Stskeepsmind if we continue tomorrow when i can think a bit clearer? you have just improved source part of mds immensely, so that shows pretty good for the 4 hour thought19:41
Stskeepswe need to probably implement binary repo serving19:42
phaeronStskeeps: ok19:42
Stskeepswhich is kinda trivial, just look at the old code19:42
phaeronyeah, code can be lifted from there I guess19:42
*** icota has quit IRC19:43
Stskeepsand then there's the issue of lastevents19:43
Stskeepswhich is showing what changed from one point to another19:43
*** CosmoHill has joined #mer19:44
*** Jucato has quit IRC19:45
Stskeepswhich we may want to indicate for branches but not for tags19:45
phaeronI think lastevents was broken in old mds19:46
*** Jucato has joined #mer19:46
Stskeepsit was definately not functioning properly19:46
Stskeepssufficiently, but broken by design19:46
*** jstaniek has joined #mer19:48
*** ka6sox-away is now known as ka6sox19:49
*** Jucato has quit IRC19:52
*** Jucato has joined #mer19:53
*** furikku has quit IRC19:54
*** Superpelican has joined #mer19:56
*** Jucato has quit IRC19:58
*** Jucato has joined #mer20:01
*** Superpelican has quit IRC20:01
*** pohly has joined #mer20:02
*** yashshah has quit IRC20:04
*** yashshah_ has joined #mer20:06
*** Milhouse has joined #mer20:18
Stskeepsevening vgrade20:20
*** pohly has quit IRC20:21
Stskeepsg/g vgrade20:21
*** tilgovi has joined #mer20:22
*** Jucato has quit IRC20:23
*** Jucato has joined #mer20:25
*** alien_ has joined #mer20:25
*** rubdos has quit IRC20:26
*** leinir has quit IRC20:28
*** leinir has joined #mer20:28
*** wmarone_ has joined #mer20:30
*** joe____ has joined #mer20:30
*** kerio has quit IRC20:30
*** disco_stu_droid has joined #mer20:31
*** tilgovi has quit IRC20:32
*** joe____ has quit IRC20:32
*** joenas has joined #mer20:33
*** cristi_ has joined #mer20:34
*** zumbi_ has joined #mer20:34
*** zumbi_ is now known as Guest2703520:35
*** disco_stu has quit IRC20:35
*** zumbi has quit IRC20:35
*** wmarone has quit IRC20:35
*** alien_ has quit IRC20:35
*** disco_stu_droid is now known as disco_stu20:36
*** Jucato has quit IRC20:37
BostikStskeeps: figured out how to fix qttools build for RC220:37
*** joenas has quit IRC20:37
Bostiknot qmake.conf, unfortunately20:38
*** Jucato has joined #mer20:38
Bostikthat was all there was, something gets set differently than the build system (and/or digia guys) assume20:39
Stskeepsone more year and we'd be a tier one and they'd know when they break stuff.. :P20:39
*** Sentient6 has joined #mer20:39
*** yashshah_ has quit IRC20:42
*** arcean_ has joined #mer20:46
*** arcean has quit IRC20:46
*** arcean_ is now known as arcean20:46
*** mdfe has quit IRC20:49
*** alh has quit IRC20:49
*** kdub has quit IRC20:49
*** peavey has quit IRC20:49
*** niqt has joined #mer20:54
Bostikyeah right20:54
*** lizardo has quit IRC20:57
*** SfietKonstantin has joined #mer21:02
*** cristi_ has quit IRC21:03
*** Cosmo[PB] has joined #mer21:05
*** Jucato has quit IRC21:06
*** Jucato has joined #mer21:07
*** CosmoHill has quit IRC21:07
*** Cosmo[PB] has quit IRC21:11
*** CosmoHill has joined #mer21:11
*** Sentient6 has quit IRC21:11
*** arcean has quit IRC21:15
*** Cosmo[PB] has joined #mer21:16
*** CosmoHill has quit IRC21:17
*** Cosmo[PB] is now known as CosmoHill21:17
*** plfiorini has quit IRC21:23
*** plfiorini has joined #mer21:23
*** arcean has joined #mer21:27
*** jayrulez_ has quit IRC21:27
*** Martix has quit IRC21:27
*** jpetersen_ has joined #mer21:28
*** Martix has joined #mer21:29
CosmoHillmy computer has been "logging off" for the last half hour21:29
*** jpetersen has quit IRC21:31
*** mdfe has joined #mer21:33
*** ftonello has quit IRC21:33
*** mdfe has quit IRC21:33
*** ftonello has joined #mer21:33
*** SfietKonstantin has quit IRC21:34
*** Martix has quit IRC21:34
*** Martix has joined #mer21:35
*** Martix has quit IRC21:36
*** Martix has joined #mer21:37
*** Martix has quit IRC21:38
*** ftonello has quit IRC21:41
*** blauzahl_ is now known as blauzahl21:41
*** blauzahl has joined #mer21:41
*** edgar2 has quit IRC21:46
*** alh has joined #mer21:50
*** kdub has joined #mer21:50
*** peavey has joined #mer21:50
*** Martix has joined #mer21:53
*** JLP has quit IRC21:54
ShadowJKCosmoHill, call your local sawmill and ask why the fuck they're logging in your computer21:54
*** Aristide has joined #mer21:55
*** disco_stu has quit IRC21:55
*** Martix has quit IRC21:56
*** Martix has joined #mer21:57
*** lpotter has quit IRC21:58
*** Aurium has quit IRC21:59
*** yunta has quit IRC22:04
*** ftonello has joined #mer22:10
*** faenil has quit IRC22:20
*** FSCV has quit IRC22:20
*** FSCV_ has joined #mer22:22
*** pvanhoof has joined #mer22:26
*** disco_stu has joined #mer22:32
*** zalan has quit IRC22:36
*** jayrulez has joined #mer22:37
*** CosmoHill has quit IRC22:39
*** NIN101 has quit IRC22:48
*** JLP has joined #mer23:00
*** JLP has joined #mer23:00
*** blaroche has quit IRC23:01
*** zenvoid has quit IRC23:02
*** yunta has joined #mer23:04
*** gabriel9 has joined #mer23:06
*** calvaris has quit IRC23:09
*** gabriel9 has quit IRC23:12
*** M4rtinK has quit IRC23:13
*** merder2 has joined #mer23:14
*** CosmoHill has joined #mer23:15
*** gabriel9 has joined #mer23:15
*** himamura has quit IRC23:17
*** arcean_ has joined #mer23:18
*** hasselmm has quit IRC23:19
*** arcean has quit IRC23:20
*** M4rtinK has joined #mer23:28
*** hasselmm has joined #mer23:32
*** niqt has quit IRC23:37
*** Aristide has quit IRC23:38
*** jstaniek has quit IRC23:39
*** pvanhoof has quit IRC23:41
*** denexter has joined #mer23:52
*** rdqfdx has quit IRC23:57
*** VDVsx has quit IRC23:57

Generated by 2.11.0 by Marius Gedminas - find it at!