Sunday, 2012-10-14

*** jstaniek has quit IRC00:04
*** lbt_away is now known as lbt00:19
*** karlzero has quit IRC00:26
*** karlzero has joined #mer00:26
*** yunta has quit IRC00:34
*** kaziklubey has joined #mer00:45
*** BeholdMyGlory has joined #mer00:46
*** kaziklubey has quit IRC00:49
*** Behold has quit IRC00:50
*** trbs has quit IRC00:57
*** trbs2 has quit IRC00:57
*** felipec has joined #mer01:14
*** furikku has joined #mer01:18
*** karlzero has quit IRC01:24
*** karlzero has joined #mer01:24
*** karlzero has quit IRC01:52
*** karlzero has joined #mer01:53
*** hazchemix has quit IRC02:13
*** karlzero has quit IRC02:50
*** karlzero has joined #mer02:50
*** DocScrutinizer05 has quit IRC03:03
*** DocScrutinizer05 has joined #mer03:03
*** karlzero has quit IRC03:19
*** karlzero has joined #mer03:19
*** beford has joined #mer03:28
*** ljp has quit IRC03:29
*** ljp has joined #mer03:30
*** phinaliumz has joined #mer03:50
*** Behold has joined #mer04:12
*** BeholdMyGlory has quit IRC04:12
*** Shaan7 has joined #mer04:15
*** karlzero has quit IRC04:17
*** karlzero has joined #mer04:17
*** araujo has quit IRC04:19
*** dakovaci has joined #mer04:25
*** karlzero has quit IRC04:46
*** dakovaci has quit IRC04:46
*** karlzero has joined #mer04:46
*** bigmeow has quit IRC04:47
*** clopez has joined #mer05:00
*** clopez has quit IRC05:40
*** karlzero has quit IRC05:44
*** karlzero has joined #mer05:44
*** M13 has quit IRC05:46
*** clopez has joined #mer05:54
*** karlzero has quit IRC06:12
*** karlzero has joined #mer06:12
*** FlameReaper has joined #mer06:22
*** araujo has joined #mer06:40
*** araujo has joined #mer06:40
*** kaziklubey has joined #mer06:43
*** felipec has quit IRC06:47
*** ortylp has joined #mer06:54
*** clopez has quit IRC07:04
*** phaeron has joined #mer07:05
*** Shaan7 has quit IRC07:05
*** merder has joined #mer07:06
*** karlzero has quit IRC07:10
*** karlzero has joined #mer07:11
vgrade_morning07:19
*** Shaan7 has joined #mer07:23
phaeronmorning07:30
vgrade_\o07:30
*** karlzero has quit IRC07:39
*** karlzero has joined #mer07:39
*** cat1 has quit IRC07:47
*** ortylp has quit IRC07:56
*** cat1 has joined #mer08:02
Stskeepsmor08:02
Stskeepsn08:02
*** M4rtinK has joined #mer08:07
*** phinaliumz has quit IRC08:12
iekkumorning08:16
*** Shaan7 has quit IRC08:17
*** Shaan7 has joined #mer08:18
*** Sfiet_Konstantin has joined #mer08:18
*** CosmoHill has joined #mer08:21
situHi all08:24
*** M4rtinK has quit IRC08:25
*** karlzero has quit IRC08:37
*** karlzero has joined #mer08:37
Stskeepsmorn situ08:37
situMorning08:37
*** kaziklubey has quit IRC08:38
*** kaziklubey has joined #mer08:41
lbtmorning08:59
CosmoHillhello .o/08:59
*** karlzero has quit IRC09:06
*** karlzero has joined #mer09:06
*** merder has quit IRC09:08
*** kaziklubey has quit IRC09:14
Stskeepsphaeron: btw, a lot of our QA tests has configuration settings09:18
Stskeepslike, use this alsa device, use this bpp ..09:18
Stskeepshave we considered how we vary this per device to device?09:18
phaeronnot really. I think they should have sensible defaults that apply to 80% of cases at least09:19
*** amjad_ has joined #mer09:20
Stskeepsok09:20
phaeronif there's one alsa device , then there's not much to configure :)09:20
*** NIN101 has joined #mer09:28
*** faenil has joined #mer09:29
lbtStskeeps: we're not using libproxy ?09:35
Stskeepswe're using pacrunner which is much more lightweight09:36
lbtbut doesn't support local static proxy settings09:36
Stskeepsmoment, there was a pretty good reason..09:37
lbtOK - I'm just setting up hostname and stuff for SDK to support PAC config and then using a local wpad server09:38
Stskeepsright, the issue is that libproxy depends on either a full webkit or mozjs09:40
Stskeepsneither of which i really want to load into apps09:40
lbtno09:40
lbthttp://code.google.com/p/libproxy/issues/detail?id=15209:40
lbtguessing the full mozjs is needed to support pac09:41
Stskeepsright, pacrunner uses v8 though09:41
lbtand wonder if that patch means you could ./configure without mozjs09:41
Stskeepsmaybe09:43
* lbt wonders about hacking pacrunner too09:43
lbtto look at /etc/pacrunner/wpdat.dat before doing a query09:43
StskeepsBuild-Depends: cdbs (>= 0.4.93~), debhelper (>= 8.1.3~), gnome-pkg-tools, autotools-dev, python-all-dev (>= 2.6.6-3~), network-manager-dev [linux-any], libdbus-1-dev, libmozjs-dev (>= 6), kdelibs5-dev, libqt4-dev, libwebkitgtk-dev, gconf2, libxmu-dev09:43
Stskeepsmakes me shiver though09:43
Stskeepshopefully some of those are optional09:43
lbt*g*09:44
lbtI like the /etc/pacrunner/wpdat.dat idea09:44
Stskeepsfwiw, where do you plan on putting a local wpad server..09:45
lbtyou mean on device or my LAN ?09:45
Stskeepsah, so it's for your use case09:45
lbtno, I meant to support local proxy config09:46
Stskeepsok09:46
lbtpacrunner does an http call and feeds resulting js to v8 ... yes?09:46
Stskeepscorrect09:46
lbtso hack pacrunner to check for  /etc/pacrunner/wpdat.dat09:46
lbtit has a 'search' list of DNS to try09:47
Stskeepsok09:47
lbtso this becomes the local override - it would support local proxy settings for the rare case we need them09:47
Stskeepsi'm aware pacrunner looks a bit like a NIH from intel's side but i can somewhat agree with their more lightweight approach..09:47
lbtoh yes09:47
lbtand it's very mobile oriented09:48
lbtand I just implemented a js that says "proxy all rpm urls via my caching proxy" :)09:48
Stskeepsok09:48
lbtmy main concern is situations where a user wants to force a proxy for Mer, not Nemo09:49
Stskeepsi think if you can force a proxy for one rpm repository or the other, the ball's in your court to fix issues09:50
Stskeeps:P09:50
lbt*g* I'm hacking on SDK and when I wipe the image and reinstall sb2 and it downloads fresh each time... it's sloooooow09:51
lbtso I thought ... "set the proxy" ....09:51
Stskeepsbtw09:51
Stskeepsdid you restart the workers?09:51
lbtyes09:51
Stskeepsok09:51
lbtand the scheduler09:51
lbtvery late last night though - maybe 1:30?09:52
lbtthey were running the cold start last I checked before I left09:52
Stskeepsok09:52
Bostikbtw, ease telll me we have recent enough mesa for qtwayland's needs09:53
StskeepsBostik: else grab it from https://build.ci.merproject.org/project/packages?project=home%3Astskeeps%3Azephyr09:54
Stskeepsthere's patches to make it work properly09:54
Bostikphew09:54
Bostikat least somewhere :)09:54
Stskeepsi need to peek at mesa 9.0 as well09:55
Stskeepswiki.merproject.org/wiki/Zephyr may also be useful09:57
Bostikoh yes, that'll be a resource09:57
Stskeepshttp://releases.merproject.org/~carsten/qmlcomponentgallery.png - got it to this point, so fairly good base09:58
Bostikmesa is just one of specially evil things to package  - even hunting down cherrypicks is at times hideous09:58
Bostikand I'm pretty damn close to getting qtwebkit nicely done09:59
Stskeepscool09:59
*** Siosm has joined #mer10:01
*** karlzero has quit IRC10:03
*** karlzero has joined #mer10:03
*** Superpelican has joined #mer10:03
Bostikarm/thumb/nonthumb done; egl/gles stuff done; npapi exclusion WIP...10:04
Bostikmany of the 2.2.x build  hacks are now part of upstream and their source-local qmake/mkspec overrides :)10:05
Stskeepsgood :)10:05
*** Shaan7 has quit IRC10:08
*** mdfe has joined #mer10:10
Stskeepsmoin mdfe10:10
mdfegood morning :)10:11
*** mike7b4_lenovo has joined #mer10:11
Stskeepshow's stuff?10:11
mdfejust woke up...10:13
mdfewith bad headache10:14
mdfe;)10:14
* CosmoHill offers mdfe tea10:15
mdfethx :)10:15
CosmoHillI only got out of bed 10 minutes ago or so myself10:15
mdfehehe10:16
*** Behold has quit IRC10:26
*** BeholdMyGlory has joined #mer10:27
Skrydoes anyone here know about sgx 530 and 32bpp depth? I'm getting some strange behaviour here..10:27
Stskeepsuse 16bpp10:28
Skryjust experimenting here. when X is at 24/32 depth, whole screen gets "multiplied" so it consists of 4 small screens.10:30
Skryit's hard to explain10:30
*** karlzero has quit IRC10:31
Stskeepsok10:31
*** karlzero has joined #mer10:31
*** himamura has quit IRC10:34
*** himamura has joined #mer10:37
*** faenil has quit IRC10:50
*** CosmoHill has quit IRC11:00
*** kaziklubey has joined #mer11:02
*** kaziklubey has quit IRC11:06
*** morphis_ has joined #mer11:06
*** morphis has quit IRC11:07
*** morphis_ is now known as morphis11:07
*** rcg has joined #mer11:20
*** popey has quit IRC11:27
*** karlzero has quit IRC11:29
*** karlzero has joined #mer11:29
*** Behold has joined #mer11:30
*** faenil has joined #mer11:31
*** andreas12 has joined #mer11:33
*** BeholdMyGlory has quit IRC11:33
SuperpelicanStskeeps I read this yesterday: http://fedoraproject.org/wiki/How_to_create_an_RPM_package11:35
SuperpelicanIt says that you shouldn't use the "Copyright", "Packager" and "Vendor" tags11:35
SuperpelicanStskeeps:I'm assuming this are Fedora specific things?11:35
SuperpelicanStskeeps:That the Fedora Packagers don't want you to use when building packages for Fedora?11:35
SuperpelicanStskeeps:Are there similiar requirements for the Mer Project?11:36
AardSuperpelican: are we using any of those tags?11:36
Stskeepsjust use License: and that's enough11:36
SuperpelicanAard: I don't know11:37
Stskeepsrest is done automatically11:37
SuperpelicanThat's why I'm asking11:37
*** FlameReaper has quit IRC11:37
SuperpelicanAre there any specific things that I should keep in mind when building RPM packages for the Mer Project11:37
SuperpelicanThat are Mer specific and not standard when building RPM packages for other distro's11:37
lbtSuperpelican: did you ever read the python 2.x packaging for Mer?11:38
*** CosmoHill has joined #mer11:38
lbtFYI pacrunner doesn't do the wpad bit ... connman does that. And connman already supports 'manual' configuration (somehow)11:39
CosmoHillmy laptop battery died so hard my laptop woke up in 197011:39
Superpelicanlbt:No, where can I find it?11:42
lbthttps://wiki.merproject.org/wiki/Source11:42
Superpelicanlbt: Can't find any information about python 2.x packaging11:43
lbtwhat does the first line on that page say? :)11:44
Superpelicanlbt:That python.spec is much shorter than the python3.spec of openSUSE!11:46
lbtI know11:46
*** Shaan7 has joined #mer11:47
lbtthe first bit of advice I gave you was to start with the current python 2.x package, build it and then go from there11:47
Superpelicanlbt:So all those .so's that are below %files lib go in the rpm package :011:47
Superpelicanlbt:So I've to manually check all the libs of python 311:48
lbtyep - also notice there is a .yaml file11:48
Superpelicanlbt:Which should go in the RPM11:48
lbtyes11:48
Superpelicanlbt:But was does the .yaml file do?11:48
Superpelicanlbt:It only gives information, right?11:48
lbthttps://wiki.merproject.org/wiki/Spectacle11:48
Superpelicanlbt:But why use .yaml files and convert them to .spec if you can use .spec straightaway?11:49
Superpelicanlbt:What the advantage?11:49
rcgSuperpelican, yaml is much easier to handle11:49
rcgalso spectacle gives you nice hints in case something is wrong or could be improved11:50
rcgfurthermore, handling patches is much easier this way11:50
*** Jonno_ has joined #mer11:50
rcgoverall, i would strongly suggest going with yaml + spectacle11:51
Superpelicanrcg, lbt:But how am I ever going to find out which libraries rpm should put in the .rpm?11:51
SuperpelicanAnd I probably can't copy that part of the python.spec in Mer GIT11:51
* rcg was really sceptic about yaml+spectacle before but when you get used to it you quickly learn to value its benefits ;)11:51
*** zenvoid has joined #mer11:51
lbtrcg: yep - it has a lot of sweet spots; especially for new developers11:52
SuperpelicanAs in Python 3 probably a lot of libraries are renamed added or removed11:52
lbtSuperpelican: packaging *means* deciding where to put files produced by building a package11:53
lbtuse 2.x as a hint11:53
lbtthe build process will tell you about any you miss11:53
Superpelicanlbt:So it should be a combination of trying to build & see what's missing + browsing through the untarred python-3.3.0.tgz archive?11:54
w00tonce you have the basics (compilation) working, trying to build the package will error if files aren't packaged11:55
lbtno need for browsing as such11:55
lbtlike I said - do a local build of python 2.x first.11:55
Superpelicanrcg, lbt:Should I install spectacle in Mer SDK or just openSUSE?11:55
lbtthen replace the tarball and make it configure+build11:55
lbtdo all work in the SDK11:56
SuperpelicanBut is spectacle even available in the Mer SDK repo's?11:56
Superpelicanlbt:But why should I first try to do a local build of 2.x?11:56
Superpelicanlbt:python 2.x for Mer is already done right?11:56
rcgSuperpelican, well, i guess that depends on the way you use it.. as i do all the stuff i do on obs in my normal os i installed it directly in my common os11:57
rcghttps://build.pub.meego.com/package/view_file?file=maliit-framework.yaml&package=maliit-framework&project=CE%3AMW%3AShared&rev=0363af95eb1ac25a113586649f1fd78011:57
rcgbut i think lbt has a valid point there as well :)11:57
rcgSuperpelican, see the link above for an example of a project with subpackages11:57
rcgthe syntax etc. should be straight forward11:57
*** karlzero has quit IRC11:58
Superpelicanrcg:Does python have subpackages?11:58
lbtyes11:58
*** karlzero has joined #mer11:58
lbtsee the yaml/spec for 2.x11:58
lbt(see the subtle hints I'm giving here?)11:58
SuperpelicanI've already locally run ./configure && make11:58
Superpelicanfor python311:58
SuperpelicanAnd got the interpreter running in Mer SDK11:58
lbtyep - that's kinda useful11:58
SuperpelicanWithout patches, hacks etc.11:59
rcgSuperpelican, packaging is not about compiling but putting the compilation results meaningfully into smaller packages11:59
Superpelicanrcg:But why should I divide the Python interpreter in different packages?11:59
w00tthe point is that you need to get the compilation done through packaging11:59
rcgbut having something that compiles and runs afterwards is a good step towards this :)11:59
w00tyou shouldn't be running configure and make yourself11:59
SuperpelicanYes I know12:00
Superpelicanrpmbuild should12:00
rcgSuperpelican, well, it depends... some software has lots of additional packages like help files or libraries that are not frequently used.. people tend to split such things up into distinct packages12:00
SuperpelicanCurrently I'm having problems with %setup macro part of %prep12:00
rcgthis way "users" can choose what they want12:00
SuperpelicanBut what's the point of having not working python interpreter?12:01
SuperpelicanI've never seen various "python3" packages in the openSUSE & Ubuntu packagemanager12:01
SuperpelicanOr is "python3" just a metapackage?12:02
lbtSuperpelican: please. Go and build the 2.x package locally. Then read the spec. Then read the yaml. Then think... a lot ... about why it might be laid out like that.12:02
lbtafter that you can do things to begin to move to 3.312:02
lbtyou need a foundation to build upon12:02
lbtthis will help you12:03
Superpelicanlbt:With build you mean package in RPM?12:03
SuperpelicanNot just makeinstall?12:03
lbtI mean run "osc build Mer_Core_i486 i586"12:03
Stskeepshttp://gitweb.merproject.org/gitweb?p=mer-core/python.git;a=blob;f=python.spec;h=00c1b9967ca08597c8eae3d40101d69d5a9ebabf;hb=HEAD12:03
Superpelicanlbt:but osc isn't locally, right?12:04
lbtwrong :)12:04
lbtosc build is local12:04
lbtalso I assume you found and read : https://wiki.merproject.org/wiki/Contribution_in_detail (please get approval here before doing a gerrit push though)12:05
Superpelicanlbt:Confused, when should I use osc utility, when SB2 and when just plain rpmbuild12:05
lbtalways, for cross compiling, never12:05
*** ortylp has joined #mer12:06
*** ridikulus_rat has joined #mer12:06
lbtactually   :  always, never: it's magically used for cross compiling, never12:06
SuperpelicanBut why did Stskeeps keeps point me to Max RPM book then?12:06
lbtbecause it describes the spec file12:06
SuperpelicanK12:07
w00t*(sb2 and rpmbuild can be used, if you know what you're doing, in special cases)12:07
lbtrpmbuild *is* used internally12:07
lbtw00t: ain't OCD about accuracy a bitch :)12:07
w00tbut osc is usually the right choice12:07
w00tlbt: sorry :P12:07
Superpelicanlbt:Does osc also provide addinational information about what's happening at building time?12:08
SuperpelicanDoes it have a sort of "debugger"?12:08
* lbt curses zypper search .... wtf does it produce an ascii table?12:08
lbtSuperpelican: yes12:08
SuperpelicanBecause osc commands seem a lot more userfriendly to me12:09
lbtdo what I said above ... I was going to type more but lets wait until you get that far first12:09
Superpelicanlbt:So I should first branch the Mer Python 2.x package?12:09
Superpelicanlbt:And then do a local build with osc?12:09
lbtas per the Contribution in detail package12:09
SuperpelicanK12:09
lbtuse i48612:09
lbtslight risk of hitting Intel Atom'isms on i58612:10
lbtStskeeps: can we consolidate i486/i586? the only delta was glibc iirc and it could have runtime detection12:11
lbtqt did RT detection anyway according to Thiago12:11
Stskeepsi486 is ultra generic and there would be a significant loss for not optimizing at atom type level12:12
StskeepsIMHO12:12
*** ridikulus_rat has left #mer12:12
Stskeepsvgrade_: how is the eglfs module going?12:13
lbtOK12:13
*** Superpelican has quit IRC12:14
*** savaaZ has quit IRC12:14
*** savaaZ has joined #mer12:16
*** CosmoHill has quit IRC12:18
*** cxl000 has quit IRC12:23
*** imunsie has joined #mer12:23
*** imunsie has quit IRC12:29
*** Superpelican has joined #mer12:44
Superpelicanlbt:why not i586 or i686?12:45
SuperpelicanWho uses i486 these days?12:45
lbthttps://wiki.merproject.org/wiki/OBS_architecture_naming12:45
lbtwe're limited to the names we can use12:46
Superpelicanlbt:But I can choose to use i586/i68612:46
SuperpelicanSo why choose i486 than?12:47
lbtdo you have ssse3 hardware?12:47
lbtie Intel Atom based?12:47
lbtthe description column is important12:47
SuperpelicanWell I do have a netbook, but don't use it for development12:47
lbtright .. so ... i48612:47
SuperpelicanBut why is the naming so confusing?12:48
lbt*g*12:48
SuperpelicanARMv6 is totally different from ARMv712:48
lbthistory12:48
SuperpelicanBut how can you tell osc the difference between ARMv6 and ARMv712:49
SuperpelicanI mean you can't mix those12:49
SuperpelicanThere not compatible12:49
lbtthe description column is important12:50
SuperpelicanBut what's going to happen when real ARMv8 devices show up12:50
lbtthe description column is important12:50
vgrade_Stskeeps: slowly, but have playground now, https://build.pub.meego.com/project/show?project=home%3Avgrade%3AraspPiQt5Plugin12:50
SuperpelicanK, will read that first in that case12:50
lbtthe rest are system naming conventions and are just letters12:50
Stskeepsvgrade_: cool12:50
* lbt -> kitchen ... bbiab12:51
vgrade_Stskeeps: so we have a -rpi directory in the Qt5 source populated with a copy of eglfs for now, so I could test the packaging12:52
*** andreas12 has quit IRC12:53
vgrade_Stskeeps: so next step is to include the Pi gfx adaptation package and hack in the hooks12:53
Stskeepsye?12:54
Stskeepsp12:54
vgrade_Stskeeps: first bit took me longer than expected but learning loads12:55
vgrade_Stskeeps: handling two source tar files in packaging etc12:56
Stskeepstwo?12:56
*** karlzero has quit IRC12:56
Stskeepsyou mean two specs12:56
*** karlzero has joined #mer12:56
Superpelicanlbt:Is it also possible to only branch python.git instead of the whole Mer repo?12:57
vgrade_Stskeeps: I have Qt5 source tar and my plugin tar12:57
vgrade_Stskeeps: two specs as well12:58
Stskeepsah, ok12:59
StskeepsSuperpelican: yes13:00
SuperpelicanStskeeps:OSC is complaining after I logged in that my home dir or any other dir "is not an osc project dir or working copy"13:03
SuperpelicanI've no idea how to fix it13:03
*** ced117 has quit IRC13:04
*** BeholdMyGlory has joined #mer13:06
*** Behold has quit IRC13:07
*** ced117 has joined #mer13:09
*** ced117 has joined #mer13:09
SuperpelicanStskeeps:Looks like there's a bug in OSC13:11
situStskeeps: which repository contains blts-alsa-core ?13:12
SuperpelicanStskeeps:Running "osc -A https://api.pub.meego.com" and entering an username and password results in: http://pastebin.com/exqtf5KM13:14
SuperpelicanStskeeps: Looks like some OSC bug13:14
Stskeepssitu, mer:qa:tests:testing13:14
situohh the package is called blts-alsa-core-tests13:14
dm8tbrSuperpelican: might need to fix your oscrc manually13:14
SuperpelicanDoes OSC keep my password and username in plain text?! :013:21
SuperpelicanThat must be a bug13:22
Superpelicandm8tbr:And how I'm supposed to fix my oscrc?13:23
Stskeepsdo you have a login/password to the meego community obs?13:24
*** karlzero has quit IRC13:24
Superpelicanno13:24
*** karlzero has joined #mer13:24
Stskeepsthat's probably a good start13:24
SuperpelicanI was just trying to do a local build with osc13:24
Stskeepsget a meego.com login and tell lbt the username13:24
dm8tbrthere is a way to store the pw obfuscated13:25
SuperpelicanStskeeps:But for how long is meego.com going to exist?13:25
dm8tbrbut you can also disable storing and enter it each and every time13:25
dm8tbrSuperpelican: that's where COBS lives right now, so you need that account.13:26
*** Behold has joined #mer13:45
*** BeholdMyGlory has quit IRC13:46
*** amjad_ has quit IRC13:46
*** khetzal has joined #mer13:49
*** mdfe has quit IRC13:49
*** Hoolxi has joined #mer13:50
*** ortylp has quit IRC13:52
Superpelicanlbt:my meego.com username = superpelican13:55
lbtdone ... have fun :)13:55
*** karlzero has quit IRC13:55
Superpelicanlbt: But Stskeeps told me I should give you my username and that you would then fix: http://pastebin.com/exqtf5KM13:56
SuperpelicanI get that error on pastebin every time I try to login into obs/osc13:57
*** amjad_ has joined #mer13:57
SuperpelicanIt all boils down to "ValueError: invalid interpolation syntax in '<mypassword>' at position 9"14:00
*** CosmoHill has joined #mer14:00
*** VDVsx has joined #mer14:06
lbt*g*14:08
lbtosc bug I bet - does your passwd have a % in it?14:08
*** FlameReaper has joined #mer14:11
Superpelicanlbt: yes14:11
Superpelicanlbt: but I just "fixed" it14:11
Superpelicanlbt:There was an if statement in the .py14:11
lbtplease report a bug for that, with a patch if you have one14:12
Superpelicanlbt:I just replaced the 'contents' of the statement with 'pass'14:12
lbtit should go upstream to suse too14:12
Superpelicanlbt:I hope this "fix" keeps working14:12
SuperpelicanAnd that I don't get trouble with it later on :(14:13
SuperpelicanBecause that statement probably was there for a reason14:13
SuperpelicanOr it was just outdated14:13
SuperpelicanAnd everybody forgot to remove it14:13
Superpelicanlbt:*sigh, osc still complains: "Error: '/home/jurre/rpmbuild/SPECS' is not an osc project dir or working copy"14:15
lbtthe instructions on the wiki are pretty clear ... they don't mention rpmbuild/SPECS14:16
Superpelicanlbt:K, will check wiki.merproject.org again in that case, but which page?14:16
lbthttps://wiki.merproject.org/wiki/Contribution_in_detail14:16
Superpelicanlbt:Yes but was first trying to do a local build14:18
lbthttps://wiki.merproject.org/wiki/Contribution_in_detail#Build14:18
SuperpelicanAnd besides it doesn't matter which dir I cd to it keeps complaining14:18
lbtnot if you follow the instructions it doesn't14:18
lbtthat's why we have them you know14:18
*** Free-MG has joined #mer14:19
* Bostik apparently just entered the Twilight Zone14:21
Superpelicanlbt:Tried all the commands that are in the #build section but it keeps complaining about "not an osc project dir or working copy"14:21
lbtSuperpelican: seriously?14:21
Superpelicanlbt: yes14:21
lbtread the instructions ... all of them, not a random subset.14:21
lbtBostik: uh huh...14:22
Superpelicanlbt:But all the instructions are about git and branching code14:23
Bostiklbt: yep, I have two distinct "wtf" errors and they both look like they're independent14:23
lbtSuperpelican: you say that like I didn't write them and don't know that...14:23
lbtBostik: we got hit by a cosmic ray yesterday - beat that for wtf...14:24
Bostiklbt: did anyone at the company win the lottery too?14:24
lbtBostik: I think fate is working with 'or'14:24
Superpelicanlbt:But I just don't understand what branching and git has to do with doing a local .spec build. I'm not trying to offend anyone14:24
lbtSuperpelican: some stuff you can just follow and undertand later14:25
Superpelicanlbt:So you mean that copying a branch automatically creates a dir structure suitable for use by osc?14:26
Bostiklbt: in x86 build; "g++ -Wl,--no-undefined -Wl,--no-keep-memory [....] -o libQtWebKit.so.5.0.0 [...]" -> usr/bin/ld: failed to set dynamic section sizes: Memory exhausted14:26
Bostikand that's the easy one14:26
lbtSuperpelican: I said, way back, that this is a very complex area and you're jumping in at the deep end. This is what happens when you do that :) OTOH you learn a lot very quickly *IF* you search and solve (most of) your own problems14:27
lbtBostik: it does use insane amounts of RAM14:27
lbthow much swap do you have?14:27
Bostiklbt: that was from COBS :)14:28
lbtinteresting14:28
lbtwhat worker?14:28
Bostikyou can run away screaming now if you want to14:28
lbtit may not have been picked up for pubworker414:29
Bostikpubworker04 according to log14:29
lbtouch14:29
Bostiklet me repeat myself14:29
Bostikyou can run away screaming now if you want to14:29
* lbt runs away screaming....14:29
*** billmania has joined #mer14:30
*** Hoolxi has quit IRC14:31
*** FlameReaper has quit IRC14:31
billmaniaStskeeps: Confirmed that the microSD I used with MeeGo on my N900 yesterday is only Class 2. That explains the poor performance.14:32
billmaniaI don't have a Class 10 4GB for Nemo so I can't do that experiment yet.14:32
lbtBostik: which Mer ?14:32
lbtmmm nm ... toolchain was released14:33
billmaniaI also see that I can try Nemo with my Pandaboard, which I will try after the N900.14:33
*** FlameReaper has joined #mer14:34
*** faenil has quit IRC14:36
*** Superpelican has quit IRC14:37
lbtBostik: try again14:39
lbtit was sitting at 8Gb RAM 12Gb tmpfs and 4Gb swap on a machine with 16Gb RAM14:39
lbtso I've upped swap to 8Gb14:40
lbtswap is in tmpfs14:40
lbttmpfs is backed by real swap14:40
lbtit's fiddly14:40
*** mike7b4_lenovo has quit IRC14:43
*** himamura has quit IRC14:46
*** himamura has joined #mer14:49
Bostikoookay, let's try again14:50
*** faenil has joined #mer14:53
*** rubdos has joined #mer14:59
Stskeepsrandom question to everyone: what's the cheapest android phone you know / lowest spec?15:11
faenilalso chinese phones, or common phones?15:11
Stskeepsideally ones that you know you can hack / have kernel source for15:12
faeniloh...mm15:12
Bostiklow-end ZTE devices come to mind, as they at least provide the sources without much griping15:12
faenilI'd say ZTE Blade15:14
situI would say the cheapest chinese phones won't have a name on it ;)15:21
faenil:D15:22
* Stskeeps is looking for good phones for lowend work, so15:25
Stskeepsalways interesting to see how your system behaves under constraint15:25
faenilStskeeps, +10015:26
vgrade_T-Mobile Pulse , Orange SanFransisco15:31
CosmoHillStskeeps: good man15:32
CosmoHillit sucks when people make software and they say it works fine but it would work for them, they have 2000 cores and a gillian bytes of RAM15:33
Stskeepsyeah.. mobile is a shock for most platforms15:33
CosmoHillTrain Sim 2012 ran like a dog with four broken legs on most people's computers15:34
*** kaziklubey has joined #mer15:34
vgrade_Stskeeps: I have booted both with custom kernel and userspace in the past15:34
Stskeepsvgrade_: cool15:34
Stskeepsstill have it out of curiousity?15:34
vgrade_Stskeeps: have both15:35
Stskeepskernel version?15:35
vgrade_Stskeeps: around .32 I think. Been ages since I did anything on them15:36
Stskeepsok15:37
Bostiklbt: my build's been "scheduling" for quite some time, do you need to kick the scheduler after a host config tweak?15:39
lbtI'll look15:39
lbtdispatcher seemed to have lost it - it's building now15:41
lbtit=worker415:41
*** cristi_ has joined #mer15:41
Bostikright, looks like something I've seen with earlier obs versions too15:43
*** ortylp has joined #mer15:44
*** cristi has quit IRC15:45
*** Sfiet_Konstantin is now known as mike7b4_lenovo15:55
*** mike7b4_lenovo has joined #mer15:56
*** mike7b4_lenovo is now known as Sfiet_Konstantin15:56
*** Sfiet_Konstantin is now known as SfietKonstantin15:56
*** SfietKonstantin is now known as Sfiet_Konstantin15:56
iekkuhmmm, i might have another meeting tomorrow same time than bug triage15:59
iekkucan someone chair meeting if that happens?15:59
iekkulbt, Stskeeps?16:01
lbtyep16:01
iekkuthanks16:02
*** karlzero has joined #mer16:05
*** jussi has quit IRC16:09
situAnyone following http://www.redbullstratos.com/live/ ?16:10
*** jussi has joined #mer16:10
faeniljust read tweets about it :)16:14
faenilstarting soon?16:14
situHe is going up.16:15
slowsitu: me16:17
slowit is rising fast16:17
slow15km in 45min16:17
faenilyeah read "the mission" and everything is clear now :) thanks for the link :)16:19
*** zeq has quit IRC16:20
iekkuhmmmm, do you belive he jumps this time?16:23
ka6soxyes16:24
situI think so.16:25
faenilhas he already tried?16:27
iekkumy "wife" reported at phone about that on tue when i was walking home... :D16:27
iekkui think they never got him to the air earlier16:27
faenil:)16:29
*** karlzero has quit IRC16:34
*** karlzero has joined #mer16:34
*** kaziklubey has quit IRC16:34
*** mikhas has joined #mer16:38
*** boss14420 has joined #mer16:50
*** boss14420 has quit IRC16:53
Stskeepslbt: snapshot please16:54
faenildo you hear the commentary on redbull's live?16:56
faenilmine says "Please stand by" "Starting soon"16:57
faenilI though that was ok! :(16:57
faenilthought16:57
Stskeepslbt: let's make it 0.20121018.0.116:57
situYeah.. I can.16:59
*** ALoGeNo has quit IRC17:00
*** nsuffys has joined #mer17:00
faenilsitu, works on my gf' notebook...wonder what's going on on mine instead...-.-17:00
*** faenil has quit IRC17:03
*** M13 has joined #mer17:07
*** Shaan7 has quit IRC17:07
*** popey has joined #mer17:15
vgrade_Stskeeps: I get "have choice for libGLESv2.so.2 needed by qt5-qtgui: gfx-rpi-libGLESv2 mesa-llvmpipe-libGLESv2" when I introduce gfx-rpi into the OBS project.  How can I get it to use gfx-rpi or stop mesa version?17:15
Stskeepsvgrade_: Prefer: gfx-rpi-libGLESv217:17
lbtStskeeps: OK17:17
Stskeepsi guess17:17
*** karlzero has quit IRC17:17
vgrade_Stskeeps: yea I tried that, let me try again17:20
*** khertan has quit IRC17:20
Stskeepsvgrade_: in prjconf17:20
vgrade_Stskeeps: no.  I'll try in projconf17:21
*** panda-z has joined #mer17:31
*** rcg has quit IRC17:33
*** NIN101 has quit IRC17:39
*** clopez has joined #mer17:40
*** FlameReaper has quit IRC17:42
*** faenil has joined #mer17:46
vgrade_Stskeeps: did the trick17:52
Stskeepsgood17:57
vgrade_Stskeeps: now need to round up the bcm/vcos headers/libs17:58
vgrade_Stskeeps: ah, hiding in /target/usr/include/ not /usr/include18:04
lbtFYI http://www.redbullstratos.com/live/  about to jump18:04
*** BeholdMyGlory has joined #mer18:08
*** Behold has quit IRC18:08
*** CosmoHill has quit IRC18:13
vgrade_mad man18:13
lbt"I need directions" :)18:14
lbt"down"18:14
*** felipec has joined #mer18:15
*** khetzal has quit IRC18:16
Bostiklbt: I think I'm probably hitting the 32-bit limit for linker's memory addressing18:18
Bostik"yow"18:18
*** CosmoHill has joined #mer18:18
lbt:/18:18
*** rcg has joined #mer18:19
*** HazardousWaster has quit IRC18:20
StskeepsBostik: do you have -Wl,no-keep-memory?18:21
lbthe does18:21
lbtI upped the worker to have more swap too18:22
*** kaziklubey has joined #mer18:23
Stskeepshmm18:23
Stskeepsbuild log = ?18:23
lbthttps://build.pub.meego.com/package/rawlog?arch=i586&package=qtwebkit&project=home%3Abostik%3AQt5a1&repository=Mer_Core_i58618:25
*** NIN101 has joined #mer18:26
*** kaziklubey has quit IRC18:27
StskeepsBostik: http://gitweb.merproject.org/gitweb?p=mer-core/qtwebkit.git;a=blob;f=03_hide_std_symbols.diff;h=1e68b06a04a95206917e50dad934a3cb5298767d;hb=HEAD18:28
Stskeepssymbols filter may be useful?18:28
*** jstaniek has joined #mer18:35
befordhey Stskeeps. I compiled libhybris on the pi but I was unable to run the test_ginger as it claimed the .so was an invalid elf file. do you know if the test.so is armv7?  pi supports only armv618:36
Stskeepsit probably is yes18:37
Stskeepsbut there's no android gpu libs for r-pi yet is there?18:37
*** Shaan7 has joined #mer18:38
befordno, just wanted to check loading the example test library18:38
Stskeepsok18:38
Stskeepsyes, it's armv718:38
*** HazardousWaster has joined #mer18:39
Bostikaaahhhh, so that's what it's for18:43
*** shmerl has joined #mer18:49
shmerlHi!18:50
vgrade_Stskeeps: that's it for me this weekend. Thanks for the help as always18:50
vgrade_Stskeeps: cristi_ kulve zuh https://wiki.merproject.org/wiki/Community_Workspace/RaspberryPi#Qt5_Platform_Plugin18:50
*** clopez has quit IRC18:51
shmerlShould I add a bug to Mer bugzilla to add opus and vp8 codecs and gstreamer plugins? They are be used in WebRTC, and in Jingle too, so they should be important for Mer. I'm surprised Jolla didn't pay attention to it.18:53
*** clopez has joined #mer18:54
*** Behold has joined #mer18:54
*** Behold has joined #mer18:55
shmerlAlso, how will H.264 support work for example? I.e. will it require a separate repository, to avoid patent issues?18:55
shmerlSome Jingle clients use H.264, though VP8 should be becoming more widespread now.18:56
*** Shaan7 has quit IRC18:56
*** Shaan7 has joined #mer18:57
*** BeholdMyGlory has quit IRC18:57
*** blaroche has quit IRC18:58
*** blaroche has joined #mer18:59
*** furikku has quit IRC19:00
*** nitrate_ has joined #mer19:02
*** nitrate__ has quit IRC19:05
shmerlI think libgstvpx.so should be in the gst-plugins-good.19:06
*** rdqfdx has joined #mer19:07
shmerlAnd opus probably in gst-plugins-bad.19:08
*** rcg has quit IRC19:10
*** mikhas has quit IRC19:11
*** pohly has joined #mer19:12
*** morphis has quit IRC19:16
*** M4rtinK has joined #mer19:17
*** VDVsx has quit IRC19:18
lbtw00t: any suggestion where qmlplugindump is?19:21
*** ant has left #mer19:25
*** amjad_ has quit IRC19:29
*** M13 has quit IRC19:29
kulvevgrade_: could you try using the pkgconfig -files to "resolve the header files/libs paths"? The vcos_platform_types.h is in /usr/include/interface/vcos/pthreads/ that BCM examples add explicitly to include paths. That is also added by the egl.pc (although the pkgconfig files are not really tested)19:43
kulvevgrade_: using the pkgconfig files would be good testing and maybe avoid some unneeded hardcoded paths elsewhere19:44
*** blauzahl has quit IRC19:44
*** nsuffys_ has joined #mer19:44
*** cxl000 has joined #mer19:45
*** nsuffys has quit IRC19:47
*** rubdos has quit IRC19:51
*** nsuffys_ is now known as nsuffys19:55
*** ALoGeNo has joined #mer19:57
*** ALoGeNo has joined #mer19:57
*** nsuffys has quit IRC20:00
*** disco_stu has quit IRC20:02
*** Jay_BEE_ has joined #mer20:02
*** Jay_BEE has quit IRC20:03
*** Jay_BEE_ is now known as Jay_BEE20:03
*** Jay_BEE has joined #mer20:04
*** merder has joined #mer20:06
*** Ian--- has quit IRC20:07
*** Ian-- has joined #mer20:07
*** ced117 has quit IRC20:20
*** Sfiet_Konstantin has quit IRC20:21
*** kaziklubey has joined #mer20:21
beford20:22
*** asdfafew has joined #mer20:22
*** niqt has joined #mer20:25
vgrade_kulve: thanks, the existing qmake .pro file for raspPi expects things still in /opt/vc but I'll have a look at the pc files20:25
*** wubudubudubudu has quit IRC20:25
*** blauzahl has joined #mer20:26
vgrade_kulve: I've not used those before and need to see how they fit in with qmake .pro20:27
vgrade_kulve: have you used .pc files within a qmake .pro?20:30
vgrade_kulve: http://www.developer.nokia.com/Community/Wiki/Using_pkg-config_with_qmake helps20:32
lbtStskeeps: snapshot done, no space or cpio errors20:38
lbtSage_: phaeron any objections to an import to cobs now?20:38
phaeronnone from me20:39
*** NIN101 has quit IRC20:43
lbtok .. here we go then20:45
lbtoh yeah - mips too20:50
*** pohly has quit IRC20:54
*** rcg has joined #mer21:05
CosmoHillhttp://black-flag.co.uk/lfs-qdb/?id=18121:13
CosmoHilloops, wrong tab21:13
*** ortylp has quit IRC21:16
*** ka6sox is now known as zz_ka6sox21:31
CosmoHillnight night21:33
*** Free-MG has quit IRC21:33
*** niqt has quit IRC21:35
*** billmania has left #mer21:36
*** CosmoHill has quit IRC21:41
*** cxl000 has quit IRC21:44
*** BeholdMyGlory has joined #mer21:54
*** yunta has joined #mer21:54
*** Behold has quit IRC21:57
*** faenil has quit IRC22:07
*** himamura has quit IRC22:15
*** zz_ka6sox is now known as ka6sox22:16
*** zenvoid has quit IRC22:25
*** kaziklubey has quit IRC22:36
*** shmerl has quit IRC23:04
*** M4rtinK has quit IRC23:05
*** yunta has quit IRC23:16
*** rcg has quit IRC23:18
*** popey has quit IRC23:39
*** imunsie has joined #mer23:49
*** jstaniek has quit IRC23:51
*** Behold has joined #mer23:56
*** BeholdMyGlory has quit IRC23:56

Generated by irclog2html.py 2.11.0 by Marius Gedminas - find it at mg.pov.lt!