Saturday, 2012-01-21

*** sandroandrade has quit IRC00:14
*** stepiro has quit IRC00:22
*** stepiro has joined #mer00:23
*** mdavey has joined #mer00:23
*** mlfoster has quit IRC00:33
*** timsamoff has quit IRC00:36
*** himamura has quit IRC00:44
*** phaeron has quit IRC00:45
*** csdb has quit IRC00:46
*** phaeron has joined #mer00:48
*** phaeron has quit IRC00:58
*** csdb has joined #mer01:04
*** norayr has quit IRC01:14
*** xtcx has joined #mer01:19
*** thomashc has quit IRC01:22
*** AIN2 has quit IRC01:29
*** M4rtinK has quit IRC01:31
*** himamura has joined #mer01:35
*** arcean has quit IRC01:37
*** bigbluehat has joined #mer01:38
*** himamura has quit IRC01:38
*** trbs has quit IRC02:01
*** plate has joined #mer02:04
*** gimli has quit IRC02:17
*** stepiro has quit IRC02:33
*** stepiro has joined #mer02:34
*** tmzt_ has joined #mer02:49
*** tmzt has quit IRC02:50
*** lynxis has joined #mer03:20
*** sigkill_ has joined #mer03:28
*** s1gk1ll has quit IRC03:32
*** lynxis has quit IRC03:32
*** stepiro has quit IRC03:39
*** stepiro has joined #mer03:39
*** pdz has quit IRC03:50
*** pdz has joined #mer03:53
*** lynxis has joined #mer03:56
*** lynxis has quit IRC04:07
*** lynxis has joined #mer04:07
*** onekenthomas has joined #mer04:08
*** tilgovi has joined #mer04:09
*** bigbluehat has quit IRC04:18
*** stepiro has quit IRC04:25
*** tilgovi has quit IRC04:26
*** stepiro has joined #mer04:26
*** stepiro has joined #mer04:26
*** plate has quit IRC04:31
*** stepiro has quit IRC04:44
*** stepiro has joined #mer04:45
*** plate has joined #mer04:49
*** mdavey has quit IRC04:52
*** plate has quit IRC05:05
*** disco_stu_droid has joined #mer05:12
*** disco_stu has quit IRC05:12
*** disco_stu_droid is now known as disco_stu05:12
*** plate has joined #mer05:22
*** stepiro has quit IRC05:31
*** stepiro has joined #mer05:31
*** plate has quit IRC05:42
*** plate has joined #mer05:42
*** sonach has joined #mer05:48
*** plate is now known as bourbon06:07
*** lynxis has quit IRC06:14
*** xtcx has quit IRC06:20
*** stepiro has quit IRC06:27
*** stepiro has joined #mer06:27
*** sonach has quit IRC06:28
*** stepiro has quit IRC06:36
*** stepiro has joined #mer06:37
*** stepiro has joined #mer06:37
*** stepiro has quit IRC06:46
*** stepiro has joined #mer06:46
*** sonach has joined #mer06:47
*** sonach has quit IRC06:53
*** rdqfdx has joined #mer07:02
*** Openfree` has quit IRC07:19
*** vakkov has joined #mer07:44
*** pdz- has joined #mer08:00
*** pdz has quit IRC08:01
*** leinir has joined #mer08:03
*** leinir has joined #mer08:03
*** KaIRC has quit IRC08:04
SageStskeeps: morning. Found the problem yet?08:21
Stskeepsi have no coffee and i just woke up 20 minutes ago, not yet :P08:21
Sagehehe :)08:22
Sagesry, can't help much today have quite a busy day coming.08:23
Stskeepsthat's ok08:23
StskeepsSage: so what was the exact issues we were supposed to have? bad touch?08:42
*** stepiro has quit IRC08:47
*** stepiro has joined #mer08:47
*** leinir has quit IRC08:49
*** leinir has joined #mer08:49
*** DocScrutinizer has quit IRC08:50
*** DocScrutinizer has joined #mer08:50
*** onekenthomas has quit IRC08:51
StskeepsSage: i can only blame high cpu usage / libmeegotouch on the touch issues i do see08:58
*** onekenthomas has joined #mer09:10
*** beford has quit IRC09:18
SageStskeeps: hmmp...09:34
Sagetouch and systemui going nuts09:34
Sagethe gestures crash was fixed by the xinput revert09:35
Sagethe touch on N900 have had problems before as well though09:36
Sagethe touch problem I had was on the apps menu and closing apps. The quick launchbar and the appmenu launche seemed to work nicely09:37
Stskeepshmm, ok09:37
* Stskeeps tries to build old version of libmtf09:37
Sagedid we have some libmtf touch patch as well at some point?09:39
Sageor was it only qt?09:39
Stskeepsqt only09:40
Stskeepsi think09:40
Stskeepsthere was the pan gesture thing09:40
*** macmaN has quit IRC09:41
Stskeepseither way, i don't think it's core related09:41
*** nsuffys has joined #mer09:43
Sagebtw, did you test the x86 image?09:45
*** macmaN has joined #mer09:46
Stskeepsno, that's next on my list09:46
Stskeepsmorn nsuffys09:46
nsuffysGood morning Stskeeps ;)09:49
*** disco_stu has quit IRC10:02
*** disco_stu has joined #mer10:02
*** norayr has joined #mer10:06
*** AIN2 has joined #mer10:09
*** smoku has joined #mer10:21
StskeepsSage: sysuid still acts up on x8610:53
*** tpn has joined #mer10:55
*** stepiro has quit IRC10:58
*** stepiro has joined #mer10:58
Stskeepsnsuffys: did you happen to have a wetab or exopc or something?11:01
*** stepiro has quit IRC11:07
*** stepiro has joined #mer11:08
*** M4rtinK has joined #mer11:13
smokuare my eyes right? we can finally build packages for Fedora and openSUSE on COBS? :)11:22
lbtthank phaeron11:23
lbtthe top level distros are there... that doesn't mean builds will work11:23
lbtno-one who cares *that* much about Fedora has helped to debug it11:24
lbtoh... morning all :)11:24
*** JDLSpeedy has quit IRC11:24
lbtbut smoku if you have any problems/suggestions then do let me know11:24
smokushould I try and report bugs on bugzilla then?11:25
*** jargon- has joined #mer11:25
smokuwhich bugzilla?11:25
lbtmeego - do understand it's not my top priority at the moment - but if I feel it's translatable to Mer then I'll be much more motivated11:26
*** arcean has joined #mer11:26
smokuon the main page of COBS I see only "More about the MBS, License restrictions, policy, links to bugzilla etc" ;D11:26
lbtyeah .... well, meego's dead so tidying up wording is not a big issue :)11:27
smokuI'm building my cordia packages to build both for Mer and Fedora (which is my local development platform)11:27
Stskeepsmorn lbt11:27
lbtsmoku: ah, OK11:27
smokukeeping the Mer COBS package with my local rpmbuild package is PITA11:27
*** JDLSpeedy has joined #mer11:27
*** JDLSpeedy has joined #mer11:27
lbtthat raises the prio :)11:27
smokuit would be nice if I could just use COBS to install packages to my devel machine11:28
lbtreport bugs on the meego bz and make sure I'm cc'ed (if I'm not actually assigned)11:28
smokuok. cool. :)11:28
smokumorn all btw :)11:29
Stskeeps\o/ libqtwebkit4-2.1.0-1.1.mipsel.rpm11:29
lbtcongrats :)11:30
smokuStskeeps: were there any interest in packaging dawati for Mer?   If not, I'm going to do it - I need an upgrade for my Ideapad. But maybe I should sync with someone?11:31
Stskeepssmoku: i'm personally interested, but i haven't seen anyone packaging it yet11:31
Stskeeps(for others, dawati:
smokuok. so I will just start by "upgrading" moblin packages and we'll see what's what11:32
lbtStskeeps: so ... config packages...11:33
*** Dotti has quit IRC11:34
Stskeepsget a version number on the tarball?11:35
*** mdavey has joined #mer11:35
lbtwell, I didn't want to make a git repo etc if I was going to merge them11:36
lbtso that's just minimal packaging11:36
lbtI guess as a 1st step I could package independently11:37
lbtthen merge all the mer config to one place as a task11:37
Stskeepssmoku: things i'm curious about: ability to run on glesv2 for that netbook ux11:41
Stskeepssmoku: think cheap arm/mips netbooks :P11:43
*** plate has joined #mer11:58
*** bourbon has quit IRC11:59
smokuStskeeps: I had some issues building NetbookUX for ARM11:59
smokuStskeeps: mostly related to hardcoded glx ;-)12:00
smokubut that's workable12:00
*** phaeron has joined #mer12:08
*** sonach has joined #mer12:14
lbtmorning sonach12:15
sonachmorning lbt:)12:16
*** sonach has quit IRC12:23
*** stepiro has quit IRC12:28
*** stepiro has joined #mer12:29
*** AIN2 is now known as NIN10112:32
nsuffysStskeeps: I've two Nokia N900 (azerty & qwerty) and two Nokia N9 (Black & White), I think investing in a WeTab, but it isn't sold in France. WeTab can run Nemomobile / Mer and perhaps Tizen ?12:33
*** sonach has joined #mer12:37
*** stepiro has quit IRC12:37
*** xtcx has joined #mer12:38
*** stepiro has joined #mer12:38
nsuffysoh find it at grosbill for 329€ !12:42
lbtphaeron: ask smoku about cordia12:42
phaeronI saw the video and it made me feel nostalgic about the n900. would be nice to have it on this tablet :)12:44
smokuphaeron: which tablet?12:52
phaerondell xt2 (i586)12:52
dm8tbrSage: looks like FOSDEM might simply invalidate the need for that proprietary driver... :D12:55
smokuphaeron: grab and it should work :)12:56
phaeronI was thinking more installing it on top of base mer :)12:56
smokuthe image was built from Stskeeps' OBS repos, which are no more...12:57
phaeronbut I noticed you are reorganizing the projects/repos12:57
phaeronok I'll wait while you are creating the new ones12:57
*** norayr has quit IRC12:58
smokuand the packages do not build with current Mer12:58
phaeronoh :(12:58
phaeronthere was an extra linker flag that was needed12:58
smokuback that12:59
smokufor i586 they are fine12:59
phaeronlbt: upower needs pm-utils wich is not available13:00
phaeronlbt: zless needs less which is not anywhere in core either13:00
smokuphaeron: I'm in a process of moving from GTK2 to GTK3, so it may take few weeks to get it working again13:00
phaeronthat has to be a lot of effort13:01
smokua bit13:02
smokuI'm almost dome porting Hildon.  hildon-desktop should be easier13:03
* phaeron wants to offer help, but not sure what I can do :)13:04
phaeronlbt: your sample kernel config is missing fanotify ?13:04
*** stepiro has quit IRC13:05
*** stepiro has joined #mer13:06
lbtphaeron: yes .. how did you know13:08
phaeronlbt: sifting through dmesg13:09
phaeronI don't see ill effects yet13:09
lbthere's my kernel build log
lbtyou like ?13:11
phaeronnice :)13:14
*** Dotti has joined #mer13:14
*** niqt has joined #mer13:20
phaeronthe only audio playback app for pa is crashing when I try to play something13:23
phaeronand the image viewer crashes too ..13:23
*** raignarok has joined #mer13:25
* lbt suspects he needs > kernel 2.2.17 for Mer...13:25
lbtso my car stereo is out then .... :)13:26
lbtof course reading the Mer wiki kernel page doesn't confirm this ....13:26
sonachlbt: I make hi3716m_rootfs.cramfs from mer-core-rootfs, and burn it to nandflash. When mounting the rootfs, error. maybe this error is releated to linux3.0.  Please look at
* Stskeeps looks13:29
sonachStskeeps: thanks!13:30
lbt possibly?13:30
Stskeepssonach: it looks like a corrupted file system somehow - is your kernel command line including "quiet" somehow?13:31
lbtsonach: oops .. you can't read that13:31
Stskeepslbt: pastie it13:31
Stskeepsthe blog post seem like it may be the case13:32
sonachStskeeps, bootargs is: bootargs=mem=96M console=ttyAMA0,115200 root=/dev/mtdblock9 rootfstype=cramfs mtdparts=hinand:1M(fastboot),1M(stbid),1M(loaderdb),1M(bakloaderdb),5M(ma13:32
sonachinloader),5M(bakloader),1M(logo),1M(logo1),5M(kernel),58M(rootfs),3M(apparam1),3M(apparam2),8M(ui),35M(app),-(other) mmz=ddr,0,0x86000000,160M13:32
Stskeepssonach: ok, just wanted to check we got all the information we needed13:33
sonachStskeeps: after mkfs.cramfs, the hi3716m_rootfs.cramfs is burn to the 58M rootfs region.13:34
lbt58M ?13:36
sonachlbt: the region is 58MB, and hi3716m_rootfs.cramfs is about 47MB13:38
lbthmm the root cause is "if (compr_len > (PAGE_CACHE_SIZE << 1))"13:40
*** stepiro has quit IRC13:43
sonachlbt: what do you mean by this?13:43
*** stepiro has joined #mer13:43
Stskeepssonach: what parameters did you use to make cramfs?13:47
lbtsonach: what kernel ?13:47
sonach just like this: mkfs.cramfs mer-base-rootfs hi3716m_rootfs.cramfs.13:48
sonachlbt: what do you mean?13:48
lbtwhat kernel version13:48
lbt"urrently, mkcramfs #define's PAGE_CACHE_SIZE as 4096 and uses that for blksize, whereas Linux-2.3.39 uses its PAGE_CACHE_SIZE, which in turn is defined as PAGE_SIZE (which can be as large as 32KB on arm)"13:48
sonachlbt: Linux version (zqhuan@localhost.localdomain) (gcc version 4.4.1 (Hisilicon_v200(gcc4.4-290+glibc-2.11+eabi+nptl)) ) #1 Fri Nov 25 08:513:49
sonach9:50 CST 201113:49
lbtOK - ta13:49
sonachlbt: it is
sonachlbt: what'is compr_len? is it the size of my cramfs?13:51
*** stepiro has quit IRC13:52
lbtreading the code I'm guessing it's the (un?)compressed length of a cramfs file13:52
lbtcompr_len = (*(u32 *) cramfs_read(sb, blkptr_offset, 4) - start_offset);13:52
Stskeepssonach: can i see the command you wrote the cramfs to flash with?13:53
*** stepiro has joined #mer13:53
lbtlet me co the correct kernel version13:53
lbtno, it's a chunk of filesystem13:55
*** sandroandrade has joined #mer13:56
*** mdavey has quit IRC13:57
lbtthose numbers seem very large too - corruption?13:57
Stskeepslbt: i was thinking not-zeroed blocks13:59
lbtyeah ...14:01
sonachStskeeps, wait a minute,14:04
Stskeepsok, no rush :)14:04
Stskeepsok, and that is u-boot14:07
sonachit calls "fastboot"14:07
lbt-> food ... bbiab14:10
*** xtcx has quit IRC14:11
sonachlbt: what do you mean by "it's a chunk of filesystem"?14:11
lbtreading the code, the values you get 417793690 and 3329639899 represent the length of a 'chunk' of the filesystem14:13
lbtthese are supposed to be less than about 8K14:14
lbtit doesn't tell us a huge amount other than we're reading bad data14:15
sonachReading bad data? Maybe the command I wrote the cramfs to flash is wrong?14:16
Stskeepssonach: and hi3716m_rootfs.cramfs is smaller than 60817408 bytes?14:16
lbtpossible causes are incompatible mkcramfs (not likely) or corruption (possible - can you do a nand read  and compare the result to the nand write?) or some issue in writing14:16
lbtStskeeps: he said 47M earlier14:17
sonachStskeeps: 4733337614:17
Stskeepsok, commands look fine then14:18
sonachlbt: Now I understand what you said.14:18
* lbt wonders what write.yaffs could do other than dd14:18
sonachStskeeps, what is the different between write.yaffs and write?14:19
sonachthere is not command like "write.cramfs" at all...14:19
lbtand just using write.yaffs with a cramfs image...14:19
Stskeepseither way14:19
lbt(assuming you can't brick it)14:19
Stskeepswe shouldn't use cramfs, it's read-only14:20
Stskeepssonach: do you have mkyaffs2image ?14:20
sonachStskeeps: no mkyaffs2image. On my ubuntu11.10, I cannot install mkfs.yaffs/ysffs2, so I didnot try yaffs2,14:21
lbtanyhow. really, food14:21
Stskeepssonach: ok, let me look for yaffs2 in ubuntu14:21
*** ScriptRipper has joined #mer14:22
*** ScriptRipper has joined #mer14:22
sonachStskeeps: Do you strongly recommend me to make a yaffs2 filesystem?14:22
Stskeepssonach: well, cramfs is read-only which isn't going to make Mer happy14:23
sonachStskeeps: Why?14:23
Stskeepssonach: because it relies on being able to write to root file system during bootup14:23
sonachStskeeps: Mer will write something to rootfs when mounting?14:23
sonachStskeeps: I decide to use cramfs because the whole flash size is only 128MB. I doubt it is not enough for yaffs2.14:25
Stskeepsmaybe jffs2? it has compression14:26
lbtMer shouldn't really write to /14:26
lbtwhat does it write?14:26
Stskeepslbt: think /home /tmp ..14:26
lbtagreed - but in general a ro / is a good goal14:27
lbtand if we define what is rw then we can handle that using either a tiny rw or tmpfs14:27
lbtor something14:27
lbtI'm not saying version 1.0 though .... just a general point - it's really nice to be able to reflash / and not lose anything14:28
sonachStskeeps: When cramfs is compressed, Even it is not read-only,  how do Mer write to /home or /tmp?14:28
*** raignarok_ has joined #mer14:28
*** raignarok has quit IRC14:28
Stskeepssonach: i think most people do a combination of cramfs and unionfs when they use cramfs14:28
Stskeepseither way, cramfs is not good for flash, if there is one bad block it ceases to function14:29
sonachStskeeps: well, no command like "mkfs.jffs2" is found on ubuntu11.10...14:30
Stskeepssonach: that one is easier to solve.. it's in package mtd-tools14:31
sonachStskeeps: OK. It is my own work to make jffs2 image:)14:31
*** raignarok_ is now known as raignarok14:32
*** jargon- has quit IRC14:32
Stskeepsi think your hisilicon manual probably talks about mkfs.jffs2 :)14:34
Stskeepsat least i found a manual that describes it14:34
phaeronlbt: there's a gma500 driver in 3.214:35
*** raignarok_ has joined #mer14:35
*** raignarok has quit IRC14:35
lbtphaeron: pretty sure it's not accelerated14:36
phaeronI think the acceleration is xorg side ?14:36
lbtI saw it in the config14:37
phaeronwell it says 2D KMS framebuffer14:37
*** raignarok_ is now known as raignarok14:37
sonachStskeeps, I am trying jffs2, I will report the result ASAP14:37
Stskeepssonach: ok14:37
phaeronmaybe 2D means not accelerated14:37
lbtphaeron: it does14:37
lbtwe need the drm/dri stuff14:37
lbtdrm is the kernel side14:38
phaeronkms is used for mode setting afair14:39
phaeronthen there is another stub driver that depends on a poulsbo.ko14:41
Stskeepslbt, phaeron: , second day14:41
*** vakkov has quit IRC14:42
*** raignarok_ has joined #mer14:42
*** raignarok has quit IRC14:42
lbtwhich basically gives OSS but 2D only14:43
phaeronlbt: ok14:43
phaeronStskeeps: which one is of interest :)14:44
lbtthanks though :)14:44
Stskeepsphaeron: obs, obs, obs..14:44
smokulbt: could you take a look at ?  Why does it say that repository publishing is disabled?14:44
sonachStskeeps: the command "mkfs.jffs2 –d ./meego-mer-core-armv7l-base- -l –e 0x20000 -o hi3716m_rootfs.jffs2" seems very very slow... Is something wrong??14:44
Stskeepssonach: no, that's normal14:44
smokuStskeeps: ouch.  thanks :)14:45
Stskeepssonach: it compresses quite heavily14:45
smokuStskeeps: it gets copied during copypac?14:45
Stskeepssmoku: seemingily14:45
sonachwhen I execute "mkfs.jffs2 --root=meego-mer-core-armv7l-base- --output=hi3716m_rootfs.jffs2", It is very fast,14:45
*** raignarok_ has quit IRC14:46
Stskeepssonach: i think it's because -l -e sets it to have eraseblocks14:46
*** raignarok__ has joined #mer14:46
Stskeepssonach: which is needed for NAND14:46
Stskeeps-l, --little-endian14:46
StskeepsCreate a little-endian JFFS2 image. Default is to make an image with the same endianness as the host.14:46
Stskeeps-e, --eraseblock=SIZE14:46
StskeepsUse erase block size SIZE. The default is 64 KiB. If you use a erase block size different than the erase block size of the target MTD device, JFFS2 may not perform optimally. If the SIZE specified is below 4096, the units are assumed to be KiB.14:46
sonachStskeeps: OK. I should wait... another question: does fastboot support NFS?14:52
*** raignarok__ has quit IRC14:52
*** raignarok__ has joined #mer14:54
Stskeepssonach: not sure about fastboot :) but there shouldn't be much problem in booting mer from NFS with a initrd or the likes, for easier development14:54
Stskeepsprobably a good guide for lbt to make at some point14:54
Stskeepswe did it with meego and N90014:54
*** raignarok__ has quit IRC15:02
*** afiestas has joined #mer15:02
sonachStskeeps: so currently, mer cannot boot from NFS?15:05
*** nsuffys has quit IRC15:05
lbtnothing prevents it - that's a kernel config15:05
sonachStskeeps: hehe, the mkfs.jffs2 stuff is still working, it takes about half an hour...15:06
Stskeepssonach: i consider it a bit over in hardware adaptation :) there should be nothing that stops it15:06
sonachOK. NFS is another choice later for development:)15:07
*** tomeff has quit IRC15:07
*** afiestas has quit IRC15:07
*** afiestas has joined #mer15:08
*** tomeff has joined #mer15:09
smokuare you going to copy relevant documentation from MeeGo wiki to Mer wiki?15:09
lbtsmoku: are you volunteering?15:10
smokulbt: yes15:11
lbtgreat - thanks15:11
smokuany guidelines, or whatever I feel like? :)15:11
lbtwell, not en-masse15:11
smokuwhere to?  some namespace/category?15:12
lbtwhat pages were you thinking of?15:12
smokuI'm asking because I needed to google Spectacle documentation *once again* on meego wiki15:12
*** afiestas has quit IRC15:13
*** afiestas has joined #mer15:13
lbtcool -
lbtalso ...
Stskeepssmoku: feel free to transfer that to somewhere, yeah15:15
smokuI'm not that familiar with mediawiki15:15
smokushould I create ?  or somewhere else?15:16
lbtread the links first15:16
smokuis Category: a namespace, or some magic grouping query?15:16
lbtthey have details of what to create and how to do it15:16
smokulbt: the point is I do not understand how Category:About works15:17
Stskeepssmoku: it's just a tag you put in the bottom of a page15:17
Stskeepsthe "About_category_guidelines" page illustrates how people should use it15:17
*** afiestas has quit IRC15:18
smoku'Create a normal Category page and [...]' - this may be obvious for someone familiar with mediawiki15:18
smokubut I do not know how to create a category page15:19
lbtlet me clarify it a bit15:19
*** afiestas has joined #mer15:19
lbtyep, smoku I see the problem15:19
lbtmy fault15:19
sonachStskeeps: the mkfs.jffs2 command is bad. since the output is about 30GB now!! I interrupt this command. And I find that if I specify "-l" the size of hi3716m_rootfs.jffs2 will increase quickly!!15:20
smokuI come from DokuWiki, and I'm a bit lost ;-)15:20
Stskeepssonach: that doesn't sound right indeed15:21
Stskeepssonach: let me just try to do the same here15:22
sonachStsKeeps: --root=DIR --output=file is not equal to "-d DIR -o file"!15:23
smokulbt: is mediawiki flat, or can I crate hierarchy ?15:24
Stskeepssonach: i assume you didnt by accident have /proc and /dev/pts or /sys mounted inside the directory?15:25
*** afiestas has quit IRC15:25
*** afiestas has joined #mer15:25
sonachStskeeps: Now the work is finished. I just replase "-d DIR" with "--root=DIR" and alike...15:26
sonachand I will try to burn flash now,15:26
sonachIt is very strange...15:26
Stskeepssonach: ok - i think you managed to get a unicode character in by accident, your copy of the command contained a long "-"15:27
Stskeepsnot a short one15:27
sonachStskeeps: I examine my command. Exactly what you said!!15:28
sonachI copy the command from the Chinese manual so with a long "-".15:28
lbtsmoku:  try this now
*** afiestas has quit IRC15:32
*** afiestas has joined #mer15:32
smokulbt: got it now :)15:32
lbtgood - thanks for insisting :)15:33
*** sonach has quit IRC15:34
*** KaIRC has joined #mer15:34
*** afiestas has quit IRC15:38
Stskeepslbt: i haven't been able to spot the actual talk..15:38
*** afiestas has joined #mer15:38
lbtkeep an eye out .. or ask paulsherwood....15:40
Stskeepsyeah, ok15:41
smokulbt: so, is fine?15:41
*** dionet has joined #mer15:41
Stskeepslbt: also there is the big question about vendors using those drivers and doing products with them..15:41
Stskeepspatents and so on15:41
lbthmm - why not start a 'copying meego wiki page' smoku - you did the right thing to copy the original, attribute and edit15:42
lbtStskeeps: yes, it struck me as an 'interesting' area15:42
lbtsmoku: yes, looks good.15:43
*** afiestas has quit IRC15:43
lbtsmoku: you could do  by the tools area links too15:43
*** afiestas has joined #mer15:44
* Stskeeps glances at his new n915:44
lbtand thanks15:44
lbtStskeeps: really...15:44
lbtblatant ...15:44
smokulbt: you mean a page to track meego wiki copying progress?15:44
lbtsmoku: that too - I meant mainly to note that we should attribute source15:44
Stskeepslbt: i have to get a microsim first, but in all honesty, the signal strength on n950 is crap fr me :P15:44
* lbt looks at phaeron and growls15:45
*** arcean_ has joined #mer15:45
smokulbt: you mean like a note in the footer?  I don't quite get what you mean exactly ;-)15:46
*** sonach has joined #mer15:47
smokuStskeeps: I just got Lumia800 and it looks... niiiceee... makes me wanna get N9 too ;-)15:48
*** arcean has quit IRC15:48
*** vakkov has joined #mer15:48
sonachStskeeps: When I write jffs2 to flash, encounter error: And message when mounting rootfs is
sonachIs there something wrong with my mkfs.jffs2 command?>15:49
*** afiestas_ has joined #mer15:49
lbtanyone want to buy a lumia ?15:50
lbtsmoku: we should have a page called "Importing meego wiki"15:50
lbtand it should document what you did for the spectacle page15:51
Stskeepssonach: hmmm15:51
smokulbt: oh. OK. got it :)15:51
lbtsome people would not remember to attribute15:51
Stskeepssonach: try with write.jffs2 ?15:51
lbtyou're right that it could also have a list of pages we'd like to import and have already imported (with a link(15:51
sonachI have tried it, same error,15:52
* lbt goes back to drm15:52
Stskeepssonach: and the command line you used to mkfs.jffs2 == ?15:52
Stskeepsis what, i mean15:52
*** afiestas has quit IRC15:53
sonachStskeeps: sudo mkfs.jffs2 --root=meego-mer-core-armv7l-base- --little-endian --eraseblock=0x20000 --output hi3716m_rootfs.jffs215:53
Stskeepslooks correct15:53
sonachStskeeps, this command is equal to what is provided by hisilicon's manual...15:53
sonachStskeeps: "Attempt to write non page aligned data" is the problem?15:54
sonachStskeeps: I don't quite familiar with jffs2 or alike...15:55
Stskeepssonach: ls -l hi3716m_rootfs.jffs215:56
Stskeepshow big?15:56
Stskeeps(in bytes)15:57
*** raignarok__ has joined #mer15:57
sonachStskeeps, 5127027215:58
*** sonach has left #mer15:59
*** sonach has joined #mer16:00
sonachStskeeps: is the file size relevant?16:00
Stskeepssonach: yes, the error means that it's not aligned with eraseblocks16:01
Stskeepsso i'm just reading some documentation on how to proceed16:01
*** afiestas_ has quit IRC16:02
*** afiestas has joined #mer16:02
Stskeepssonach: try to add '-p' parameter to mkfs.jffs216:03
Stskeepsthis will pad it to erase block size16:06
*** shanem has quit IRC16:07
*** afiestas has quit IRC16:07
*** sandroandrade has quit IRC16:08
*** sonach has quit IRC16:11
*** shanem has joined #mer16:11
*** afiestas has joined #mer16:11
*** sonach has joined #mer16:15
sonachStskeeps: It seems that the rootfs works! but I don't quite confirm that:
Stskeepssonach: ok, now we're at a better state :)16:18
Stskeepslbt: where is your config flags that need to be enabled?16:19
Stskeepssonach: what next is that you need to enable some settings in your kernel Mer needs/systemd needs16:19
StskeepsFailed to mount /dev: No such device indicates you need something called DEVTMPFS16:19
lbtStskeeps: currently in the DATA section at the end of the script - matches
*** afiestas has quit IRC16:20
*** sledges has quit IRC16:20
lbtsonach: we have a script that verifies your config matches that page16:20
Stskeepssonach: - you need to add CONFIG_DEVTMPFS=y and CONFIG_DEVTMPFS_MOUNT=y and CONFIG_GROUPS=y16:21
Stskeepsat least16:21
Stskeepsafter that kernel has built, it should continue to boot16:21
Stskeepssonach: to get rid of the CLEANMARKER warnings, add '-n' option16:22
Stskeepsto mkfs.jffs216:22
sonachStskeeps: OK, so first I should rebuild the kernel with "CONFIG_DEVTMPTS=y" etc; second I should copy drivers of hi3716m to /lib/modules. Is that right?16:23
sonachStskeeps: but I don't know why "Failed to open /dev/kmsg for logging: No such file or directory" ?16:24
Stskeepssonach: that's normal, it comes with CONFIG_DEVTMPFS=y16:24
Stskeepsthe fix comes with, i mean16:24
Stskeepsthe important error is the one about /dev16:25
sonachStskeeps: OK! Thank you & lbt very much for help:)16:26
sonachStskeeps: I will try it tomorrow. It's 00:25 of my localtime and I should goto sleep:)16:26
Stskeepsgood night16:26
sonachSee you:)16:27
Stskeepslbt: one more for the hardware guide.. utilizing ubifs16:34
Stskeepsit's my impression that the merits of ubifs hasn't reached many embedded companies16:35
lbtI was thinking something along those lines as we discussed16:35
*** sonach has left #mer16:39
*** harbaum has joined #mer16:44
*** harbaum has quit IRC16:47
*** xtcx has joined #mer16:47
*** xtcx has quit IRC16:52
*** thomashc has joined #mer16:52
*** arcean_ is now known as arcean16:56
*** stepiro has quit IRC16:59
*** stepiro has joined #mer16:59
Stskeepsif there's no objections, i'll start the prep for a final release of 0.2012012017:00
Stskeepsthink we've established any regressions seen in images weren't caused by core changes17:00
*** ScriptRipper has quit IRC17:01
vgradeStskeeps, I was hoping to get conf from mdfe_ on his PA next image but not seen it yet17:05
*** niqt has quit IRC17:05
*** stepiro has quit IRC17:09
*** stepiro has joined #mer17:09
Stskeepsi personally tested with nemo on n900 and exopc17:10
Stskeeps4.8.0 will be the stable qt for a long time to come anyway17:10
Stskeepsany if anyone wants to 'stay' on old core, that's entirely possible17:10
*** trbs has joined #mer17:10
w00tStskeeps: what about the nemo input problems?17:11
w00tor were they unrelated17:11
Stskeepsw00t: the problematic ones were reverted, and the other ones were attributable to systemui / libmeegotouch upgrade cpu usage17:13
*** tomeff has left #mer17:14
w00tcpu usage?17:14
w00tit uses more, or?17:14
Stskeepsno, there's a bug17:14
Stskeepspixmap / status bar sharing in a secure manner17:14
w00tah.. great.17:14
Sagedm8tbr: yes saw the tweet :)17:17
SageStskeeps: :(17:17
Stskeepsi still need to really confirm it's because of the upgrade but i tried qml components without difficulty17:18
SageStskeeps: that is cause of something in the core-next as it works fine in the last release :(17:18
SageMTF packages are the same in Thursdays release and in those core-next images17:18
StskeepsSage: hmm17:19
SageI didn't do any updates after Thursday to repos so only thing changed is the core AFAIK17:19
Stskeepswell, at least the systemui issue exists in thursday image too17:20
Sageit does?17:20
Stskeepsat least veskuh claimed so17:21
Sagewell, I installed that image on N900 and N950 and didn't see any problems17:21
* Sage gets his n950 17:21
Stskeepsit doesn't always show that easily, for me, on x86, it shows by a stuck 'charger not connected'17:21
Sagewell, I have used my n950 with 19.1 release now day and rebooted several times without problems17:23
Sagechecked also top and no cpu usage17:23
Stskeepsperhaps it's because the x libraries run better, or something17:24
Stskeeps! , btw17:24
StskeepsSage: this does show why we need per-change QA/automatic tests though17:27
Stskeepsit's bloody difficult to root cause a full release :)17:27
SageBut in my tests I would say that this is caused by change in core17:28
Stskeepsi think i'll release the core and we'll have to patch it when we find out more what happens17:28
Stskeepsif anything, we've learnt a lesson17:28
Sageif you can verify that to test e.g. ix86 image that would be cnice17:28
Stskeepsi tested the x86 image earlier, at least17:28
Sagesounds reasonable17:29
*** raignarok__ has quit IRC17:31
*** raignarok__ has joined #mer17:31
Stskeepsi think i'll switch method to that the moment i first pick up changes, that becomes a prerelease17:32
Stskeepsso we have the ability to make images way earlier17:32
*** raignarok has joined #mer17:32
Stskeeps.. or something17:32
* Stskeeps has to think17:32
*** raignarok has quit IRC17:32
*** raignarok_ has joined #mer17:32
SageStskeeps: what if we would make prerelase after every merged change?17:33
*** raignarok__ has quit IRC17:33
*** raignarok_ has quit IRC17:34
*** raignarok_ has joined #mer17:34
smokuStskeeps: can I copypac package from Mer17:36
vgradeStskeeps, missed that17:36
Stskeepssmoku: copypac Mer:fake:Core:i586 packagename17:37
Stskeepsvgrade: just happened17:37
smokuStskeeps: thx17:38
smokuStskeeps: any idea why this does not come up on COBS search?17:38
lbtStskeeps: I will be picking up the IMG and BOSS tasks next week I hope17:42
lbtI want to wrap up the HA properly17:42
lbtincluding the kernel checks and packaging (would have caught sonach's config problem)17:42
*** stepiro has quit IRC17:46
*** stepiro has joined #mer17:46
Stskeepssmoku: it's a remote obs17:50
Stskeepssmoku: technically17:50
Stskeepslbt: ok17:51
*** mdavey has joined #mer17:52
*** raignarok_ has quit IRC18:00
*** dionet has quit IRC18:01
phaeronI can help with that18:03
phaeronso we can divide the work18:03
lbtsounds good18:04
lbtwe have a machine from stefan that's not being used18:04
lbti'd earmarked it for VMs for wiki18:04
lbtbut IMG may be a better use18:04
lbtfirst.... /home/abuild/rpmbuild/BUILD/emgd/drm/emgd_drv.c:2371:2: error: too many arguments to function 'drm_pci_init'18:05
lbtwhy can't drivers define *something* about what kernel version they expect to work with18:05
smokuStskeeps: I need Pango built with gobject-introspection18:07
smokuStskeeps: for now I will just rebuild it, but maybe we can think of something better18:07
Stskeepssmoku: that's normal, yeah18:08
Stskeepssmoku: in practice, we do imagine that vendors will modify core18:08
smokuStskeeps: it may break things if someone upgrades to Mer package until Cordia catches up18:08
Stskeepssmoku: yeah.. generally we speak of it as running your own 'core' program18:09
Stskeepsand hoping to make this a lot more lightweight in the future18:09
lbtI'm hoping git pull --rebase18:10
vgradelbt, you always know what kernel a driver needs, "not the one you want it too"18:11
lbtthis ie emgd-drm.tgz btw18:12
lbtand is "not it, please guess again"18:12
lbtshall we try 3.1 folks ?18:14
Stskeeps3.0 may be more likely?18:14
lbtvgrade: yes18:14
Stskeepsdoesn't it say it in README.txt?18:14
lbtand no18:14
lbtwhat README ?18:14
Stskeepsthe emgd one18:14
lbtyeah ... what README18:15
vgradepdf mentions 2.6.37 but thats for the older xorg18:15
lbtit may say in the tarball I extracted the tarball from18:15
Stskeepscheck what version the IVI kernel has in meego 1.2/1.3..18:16
lbtbmc is down18:16
lbtah... back now18:16
Stskeepsit's not me this time i hope, i transfer data with 750kb/s rate limit18:17
Stskeeps(importing the new mer release)18:17
lbtI'm not using meego stuff18:17
lbtit's too undocumented18:17
lbtand scattered18:17
lbtthe intel official tarball says 2.6.3718:18
lbtbut it also says Xorg 1.9.X18:19
vgradethe wayland dir18:19
lbtand provides an Xorg 1.1018:19
lbtthe wayland dir18:19
lbtprovides binaries with no comments18:19
lbtembedded in rpms18:19
lbthowever, the X driver works with Mer's X18:20
vgrade2209 rings a bell with the ivi kernel patches18:20
lbt3226 ?18:21
lbtof course by the time anyone has navigated the maze they are so lost that they say "fuck it it works" and post a youtube video as proof and then get pissed and forget how they did it18:22
lbteh vgrade?18:22
vgradeI can remember just used the bits and pieces from ivi kernel and emgd meego package18:23
lbtI did look - there was no attribution of the binaries18:24
vgradenow that only worked with a particular kernel version, Xorg version18:24
lbtI downloaded the ppa versions from ubuntu18:24
lbtthat claims ( ) to work on 3.0.018:25
vgradeonce you want to use another kernel version or xorg version you have to start all over again18:25
lbtbut I don't have joggler patches for that18:25
lbtvgrade: yep - so having found a suitable Xorg driver I'm trying to get other code from the same tarball to work18:25
vgradeok yes seems sane.  But the patches you have don't match the kernel18:27
vgradeyou want to use18:27
lbtwell, it's an out-of-tree driver18:29
lbtso it should be a lot more flexible than patches18:29
vgradeso you have to become an expert in X, drm, dri, Linux kernel memory management to forward port the patches or use a kernel version the patch was made from18:29
lbtpretty much18:29
lbtStskeeps: OK18:29
vgradeno wonder I want to forget all that18:29
StskeepsFWIW, yocto seems to carry emgd patches as well18:30
lbtStskeeps: from what I see there are no patches18:30
Stskeepsok, so, horror18:30
lbtI meant it in a kind of good way18:31
lbtit's 'just' an out-of-tree module18:31
lbtand it actually almost builds with my HA kernel-devel package too18:31
lbtso moving back to 2.6.37 will let me fix a couple of broken symlinks in -devel18:32
*** Venemo has joined #mer18:35
*** Venemo has quit IRC18:35
*** Venemo has joined #mer18:35
vgradeand being Mer modules are allowed unlike MeeGo18:38
Stskeepsallowed but not recommended, rather, it's HA, we don't really care :P18:40
vgradelbt, I did make an attempt at documentation but as always you can never have enough,
*** eebrah has joined #mer18:43
lbtvgrade: OK ... useful thanks18:44
vgradethat was based on the IVI way of doing things with kernel patches so not so relevant to the new kernel module methid18:45
*** eebrah has quit IRC18:52
*** eebrah has joined #mer18:54
*** pohly has joined #mer19:26
lbt-> tango (again) .... until tomorrow19:41
Stskeepshave a good one :)19:41
lbtalways :D19:41
*** harbaum has joined #mer19:54
*** swerden has joined #mer19:55
*** beford has joined #mer19:55
*** swer has quit IRC19:57
*** eebrah has quit IRC20:04
*** Venemo has quit IRC20:06
*** stepiro has quit IRC20:06
*** jstaniek has joined #mer20:07
*** stepiro has joined #mer20:07
*** harbaum has quit IRC20:11
*** harbaum has joined #mer20:11
*** jstaniek_ has joined #mer20:12
Stskeepsevening jstaniek_20:12
*** nsuffys has joined #mer20:12
*** jstaniek_ has quit IRC20:13
*** stepiro has quit IRC20:16
*** stepiro has joined #mer20:16
phaerondid mer core just get switched ?20:16
Stskeepsbad time?20:16
*** harbaum has quit IRC20:17
phaeronnah , was just starting to get worried the OBS was broken20:17
phaeronnow I can go make coffee :)20:17
Stskeepsi botched the release once, out of disk space20:18
Stskeepsbut now it's all better..20:18
phaeronok :)20:19
phaeronso I went through the core + adaptation + plasma active I think this is quiet promising20:20
phaeronI have a better feel of what vendors might want20:20
Stskeepsjust a lot of work to be done first, but we already have something quite nice if we get basics going20:21
phaeronbased on a minimal mer core, we could use kickstarts to install everything too. using a script to parse it or something20:22
phaeronnot sure if that would be helpful20:22
*** beford has quit IRC20:22
*** beford has joined #mer20:23
Stskeepscould be a slippery slope of user expecting a finished system20:23
Stskeepsherro beford20:27
*** phaeron has quit IRC20:27
*** phaeron has joined #mer20:28
phaeronstupid battery died20:28
*** pohly has quit IRC20:30
befordStskeeps,  does the mer kernel support gadgetfs ?20:31
Stskeepsthere's no mer kernel :)20:31
Stskeepson what device, rather?20:31
timophstill couldn't get sb2 to work with mer's toolchain :(20:31
befordn900 :P20:31
befordoh yea, I meant nemomobile, wrong channel20:32
timophworks with codesourcery though20:32
timophdamn. it's been a while since I last used my n810. i used be a fast typist with this. not anymore20:33
Stskeepstimoph: yeah, it probably doesn't work nicely outside mer, it seems20:33
timophcould be20:34
Stskeepsi presume you didn't try inside the chroot?20:35
timophjust the usuall sb2 way20:35
timophso you're thinking chrooting to mer's rootfs might work?20:36
Stskeepsyes, and installing sb2 and the cross compiler into that20:37
timophI'll try that next20:37
timophI wonder what's needed to make it work w/o chrooting20:40
Stskeepsi'm browsing around in my codesourcery too20:40
*** pohly has joined #mer20:41
Stskeepsseems like it's possible to statically link mpfr20:45
* Stskeeps files a task bug20:46
*** pohly has quit IRC20:46
Stskeepstimoph: with sb2 and the right tools distro it's possible to do it without chrooting..20:49
Stskeepstimoph: hmmm..20:49
*** nsuffys has quit IRC20:51
Stskeepspoint to that sb2-tools-distro/ directory with -t <PATH> when doing sb2-init20:51
Stskeepsor was it -T .. well, whatever points to the tools directory20:53
timophack. I'll try that in the morning21:01
Stskeepstime for some sleep soon for me21:02
timophyeah. already 23 here so I'm off to bed21:03
timophtoday's bed time book: dive into python21:04
*** Venemo has joined #mer21:08
*** Venemo has quit IRC21:08
*** Venemo has joined #mer21:08
* phaeron is looking for something to do21:10
Stskeepshmm, probably very late to ask anyone here, but anyone who's burning for QA?21:10
Stskeepsand heading that up21:10
Stskeepswe have a rather large base of code and tests to start from21:10
phaeronas in automated QA ?21:11
Stskeepsfor example, yeah - but also test cases for people to run through21:11
phaeronI would be interested in the automated QA. deploying OTS or similar is on my list21:12
phaeronfor mer it's basically "does it boot" ?21:13
phaeronbut even that depends on the hardware adaptaion21:14
Stskeepsi think it's a bit more advanced, we also need to determine exactly what we test21:14
phaeronwell testing the toolchain is interesting.21:15
phaeronI am thinking mer only , nothing else.21:15
Stskeepsyes, of course, and same tools useful for vendors too21:16
Stskeepseven if we're only 322 packages, it is still a lot to test through21:16
Stskeepsshould probably prioritize how much gets tested21:17
Stskeepsqt, qtwebkit, qt mobility are the biggies21:17
phaeronsomething like qtdemos + tdriver21:18
Stskeeps is what ~mer atm21:19
w00ti do think that automated vendor testing is an interesting (and perhaps more rewarding, long term) idea.. i.e. does this image boot to a ux, and can i run y test applications, and perform z testcases on them21:20
phaeronYes, I agree. the main problem is how to get image --> device21:22
w00twell, a VM would be a nice start (and probably easier to automate)21:23
w00tpast that you start getting into device specifics i guess21:23
Stskeepsthat's actually not a bad idea, what if vendors could report code coverage back on the mer packages?21:23
Stskeepser, test coverage21:23
phaeronw00t: I guess with the llvm stuff Stskeeps is working on , that is actually almost possible now21:24
Stskeepsfor x86 for sure21:24
Stskeepsjust for the logs:
*** Anssi138 has joined #mer21:36
* phaeron looks at gnome-shell dep graph and gives up21:48
*** Anssi138 has quit IRC21:50
phaeronI was only able to get something viewable at max-depth=121:51
Venemophaeron, :D22:02
*** rdqfdx has quit IRC22:02
Venemophaeron, couldn't you just copy the necessary packages from Fedora?22:04
phaeronI could , but even that would take a lot of time :D22:04
phaeronmaybe not tonight :)22:05
Venemoor you could write a script that has rpm list the full dependency list, copies the spec files to a Mer chroot and rebuilds them there22:05
phaeronprobably cobs will do the building22:08
Venemowhatever :P22:09
*** IanWizard-Cloud is now known as IanWizard-Cloud|22:16
*** IanWizard-Cloud| is now known as IanWizard|GOPROV22:16
*** IanWizard|GOPROV is now known as IW|GOPROVINCE22:17
*** dionet has joined #mer22:19
*** vakkov has quit IRC22:22
phaeronVenemo: it does feel like stealing effort though22:26
*** stepiro has quit IRC22:26
*** stepiro has joined #mer22:27
Venemophaeron, why?22:28
phaeronsomeone does all the effort and I steal it22:28
Venemoreusing someone's free software is not stealing IMO22:29
phaeronI just have to make sure to properly attribute22:29
*** raignarok_ has joined #mer22:30
*** IW|GOPROVINCE is now known as IanWizard-Cloud22:38
phaeronStskeeps: new repo missing patterns ?22:39
Stskeepsdoes createrepo support patterns?22:40
Venemogood night guys22:40
phaeronStskeeps: zypper se -t pattern doesn't show them ..22:41
Stskeepsphaeron: ok, i see it, but it would have to go to next release22:41
phaeronkickstarts depending on them will fail22:41
Stskeeps--pkgmgr=zypp doesn't work anyway22:41
phaeronI don't have a problem personally :)22:41
phaeronso yum will see them ?22:42
Stskeepsphaeron: can you file a task bug listing bottom of for me?22:42
phaeronI mean yum would see them , while zypp doesn't22:42
phaeronsure will report a bug22:43
*** Venemo has quit IRC22:43
phaeronStskeeps: 13122:46
*** raignarok_ is now known as raignarok22:47
phaeronnemo black screen ..22:53
*** NIN101 has quit IRC22:57
*** NIN101 has joined #mer22:57
*** norayr has joined #mer23:13
*** ALo has joined #mer23:14
*** ALoGeNo has quit IRC23:16
*** stepiro has quit IRC23:23
*** stepiro has joined #mer23:23
*** jstaniek has quit IRC23:40
*** dionet has quit IRC23:44
*** otep has quit IRC23:47
*** otep has joined #mer23:50

Generated by 2.9.2 by Marius Gedminas - find it at!