IRC logs for #baserock for Wednesday, 2014-10-29

*** radiofree [radiofree@unaffiliated/radiofree] has quit [Read error: Connection reset by peer]00:49
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock00:51
*** radiofree [radiofree@unaffiliated/radiofree] has quit [Ping timeout: 256 seconds]01:58
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock01:58
*** radiofree [radiofree@unaffiliated/radiofree] has quit [Read error: Connection reset by peer]02:17
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock02:20
*** radiofree [radiofree@unaffiliated/radiofree] has quit [Read error: Connection reset by peer]02:31
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock02:33
*** VVIBIN [~VVIBIN@] has joined #baserock04:29
*** radiofree [radiofree@unaffiliated/radiofree] has quit [Ping timeout: 256 seconds]05:35
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock05:35
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock05:39
*** hari__ [~hari@] has joined #baserock05:59
hari__Yesterday I had run the command "scripts/ systems/devel-system-x86_64-generic.morph clusters/upgrade-devel.morph"05:59
hari__I had left it overnight as it was taking some time06:00
hari__Today morning when came back to work, it looked as if it was trying to do a SSH06:00
hari__It ended up with the following error06:01
hari__ERROR: ssh-rsync.check failed with code 1: ERROR: Unable to SSH to root@localhost: Command failed: ssh root@localhost -- true06:01
hari__any idea anything i missed and how to continue or recover06:01
hari__seems I am able to proceed after retrying two more times06:21
hari__deployment completed06:31
*** zoli_ [~zoli_@linaro/zoli] has quit [Remote host closed the connection]07:08
*** mSher [] has joined #baserock07:34
*** petefoth_ [] has joined #baserock07:35
*** aananth [~caananth@] has joined #baserock07:54
*** radiofree [radiofree@unaffiliated/radiofree] has quit [Ping timeout: 256 seconds]07:58
aananthsam | paul:  I think I setup jetson for development by following Any quick test I can perform to confirm this?07:58
*** radiofree [radiofree@unaffiliated/radiofree] has joined #baserock07:58
*** petefoth_ [] has quit []08:01
*** petefoth_ [] has joined #baserock08:01
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock08:03
*** zoli_ [~zoli_@linaro/zoli] has quit [Read error: No route to host]08:06
*** zoli_ [~zoli_@linaro/zoli] has joined #baserock08:06
*** ssam2 [] has joined #baserock08:16
Mode #baserock +v ssam2 by ChanServ08:16
*** paulsherwood2 [uid50097@gateway/web/] has joined #baserock08:26
*** ssam2 [] has quit [Ping timeout: 272 seconds]08:28
*** zoli_ [~zoli_@linaro/zoli] has quit [Read error: Connection reset by peer]08:28
paulsherwood2morning campers08:28
*** abdul [~abdul@] has joined #baserock08:32
abdul is now known as Guest7223508:32
*** tiagogomes [~tiagogome@] has joined #baserock08:32
*** Guest72235 [~abdul@] has quit [Client Quit]08:33
*** Guest72235 [~abdul@] has joined #baserock08:33
*** Guest72235 [~abdul@] has quit [Quit: Leaving]08:40
*** abdul_ [~abdul@] has joined #baserock08:40
*** ssam2 [] has joined #baserock08:50
Mode #baserock +v ssam2 by ChanServ08:50
pedroalvarezgood morning!09:04
pedroalvarezpaulsherwood2, ssam2: aananth asked you a question when you were offline - you can see the log at the bottom of this page:
*** sambishop [] has quit [Quit: Leaving]09:08
*** sambishop [] has joined #baserock09:09
*** mSher [] has quit [Quit: Leaving]09:12
*** mike [] has joined #baserock09:12
mike is now known as Guest7651809:13
*** dutch [] has joined #baserock09:14
*** sambishop [] has quit [Quit: Leaving]09:18
*** dutch [] has quit [Quit: Quit]09:24
*** dutch [] has joined #baserock09:24
pedroalvarezI'm creating a baserock installer script, which is going to take a rootfs (deployed using the sysroot write extension), and run the rawdisk write extension (which I've upgraded to deploy also to devices). I've been taking a look at how morph runs the extensions and I don't know if is going to be easy to reuse that code09:25
pedroalvarezmainly because I don't understand it yet :)09:25
*** jonathanmaw [] has joined #baserock09:26
petefoth_What operations on can only be performed by admins? I know about enabling video embedding. Any others?09:30
*** Guest76518 [] has quit [Quit: Leaving]09:32
Kinnisonadding other admins09:32
Kinnisonadding ssh keys09:32
Kinnisonaltering the ikiwiki plugin configuration in general09:32
*** perryl [] has joined #baserock09:36
*** dutch [] has quit [Quit: Quit]09:37
*** sherm_ [] has joined #baserock09:37
*** mariaderidder [] has joined #baserock09:37
*** franred [] has joined #baserock09:39
*** Zara [] has quit [Remote host closed the connection]09:43
*** benbrown_ [] has quit [Remote host closed the connection]09:43
*** pdar_ [] has quit [Remote host closed the connection]09:43
*** ridgerun1er [] has quit [Remote host closed the connection]09:43
*** jmacs [] has quit [Read error: Connection reset by peer]09:43
*** SotK [] has quit [Read error: Connection reset by peer]09:43
*** paulsherwood [] has quit [Read error: Connection reset by peer]09:43
*** petefoth [] has quit [Read error: Connection reset by peer]09:43
petefoth_ is now known as petefoth09:43
*** dutch [] has joined #baserock09:50
*** dutch [] has quit [Client Quit]09:50
*** Krin [] has joined #baserock10:01
*** Krin [] has quit [Client Quit]10:04
*** mariaderidder [] has quit [Quit: Ex-Chat]10:04
*** mariaderidder [] has joined #baserock10:04
*** petefotheringham [] has joined #baserock10:05
*** sherm_ [] has quit [Quit: Leaving]10:08
*** ridgerunner [] has joined #baserock10:08
*** sherm_ [] has joined #baserock10:08
*** benbrown_ [] has joined #baserock10:09
*** perryl [] has quit [Quit: Lost terminal]10:10
*** tiagogomes [~tiagogome@] has quit [Quit: Leaving]10:10
*** Krin [] has joined #baserock10:10
*** mariaderidder [] has quit [Quit: Ex-Chat]10:10
*** mariaderidder [] has joined #baserock10:10
*** perryl [] has joined #baserock10:11
*** tiagogomes [~tiagogome@] has joined #baserock10:11
*** tiagogomes [~tiagogome@] has quit [Client Quit]10:11
*** tiagogomes [~tiagogome@] has joined #baserock10:12
*** pdar [] has joined #baserock10:12
*** zarazaimeche [] has joined #baserock10:14
*** franred [] has quit [Quit: Leaving]10:14
zarazaimeche is now known as Zara10:14
*** sherm_ [] has quit [Quit: Leaving]10:14
*** mike [] has joined #baserock10:14
mike is now known as Guest5672010:14
*** franred [] has joined #baserock10:15
*** dutch [] has joined #baserock10:15
*** visteon [~visteon@] has quit [Quit: Leaving]10:17
*** dutch [] has quit [Client Quit]10:20
radiofreejjardon: tried your gtk branch but it failed, need to upgrade wayland10:20
petefothKinnison: for what operations on w.b.o do you need to have ssh keys uploaded? To be able to push to the git repo?10:20
*** dutch [] has joined #baserock10:21
*** ssam2 [] has quit [Ping timeout: 240 seconds]10:23
abdul_i built baserock development image for x86_64 machine. may i know where the kernel source is located?10:23
jjardonradiofree: what did fail? If its only gtk we can merge the rest until Wayland gets updated?10:23
petefothKinnison: can one *only* push with ssh keys, or is it possible to push via a git:// url ? (Sorry about the dumb questtions :)10:23
*** Guest56720 [] has quit [Quit: Leaving]10:23
*** sherm_ [] has joined #baserock10:23
KinnisonI believe you can only push via ssh10:24
Kinnisonwe didn't (I believe) enable mob push10:24
*** mariaderidder [] has quit [Quit: Ex-Chat]10:24
*** mariaderidder [] has joined #baserock10:24
petefothKinnison: thanks? 10:24
pedroalvarezabdul_: this is the kernel repository:
radiofreejjardon: gtk failed yes, i'm rebuilding with wayland 1.6.010:25
jjardonI'm happy to merge the rest of the upgrades and left the gtk one pending10:27
abdul_pedroalvarez: may i know the location in my baserock system i mean inside the /src10:28
* jjardon doesn't remember gtk to fail when he tested that though10:28
*** jmacs [] has joined #baserock10:29
pedroalvarezabdul_: sorry, maybe I'm not understanding what are you trying to achieve :/10:30
paulsherwood2abdul_: by default morph gets the source for each chunk direct from git. the local git repositories are held in /src/cache/gits/*10:31
paulsherwood2if you want to edit the source, you can do 'morph edit linux-x86-64-generic'10:31
paulsherwood2that will check out a local copy. 10:31
*** DavePageCT [] has quit []10:31
paulsherwood2from your definitions repo it will be stored at ../../../upstream/linux10:31
madhuIn yocto, the sources are copied to build directory before it gets built, I guess abdul refers that10:31
paulsherwood2right. in baserock, the build directory is normally off-limits :)10:32
straycatCan we lorry setuptools from bitbucket as well? ( )10:32
paulsherwood2no user-serviceable parts inside :-)10:32
paulsherwood2what are those?10:32
paulsherwood2straycat: ^^10:33
jjardonstraycat: I'd not put bitbucket in the module name10:33
straycatWe already have python-setuptools10:34
pedroalvarezindeed, I'd put mercuirial or hg10:34
paulsherwood2so if you want to modify the kernel, add patches etc, you do the work in the ../../../upstream/linux directory, then morph will take that and put it into the build directory when it needs it10:34
paulsherwood2abdul_: ^^10:34
pedroalvarezstraycat: is what we are doing for other lorries10:34
pedroalvareznss-hg.lorry  nspr-hg.lorry  libyaml-hg.lorry ...10:35
madhupaulsherwood2, you mean it  still exists in morph, can you point me the location10:35
straycatpaulsherwood2, It's setuptools' development repo10:36
paulsherwood2madhu: what still exists?10:36
straycatI don't see why I can't lorry it as -bitbucket, that seems more meaningful than -hg to me10:36
madhupaulsherwood2, build directory where the sources copied10:37
straycatthey could move their hg repo to some other random host10:37
paulsherwood2madhu: if the build succeeds, it is removed10:37
madhuhmm, can it be configurable not to delete them10:38
pedroalvarezstraycat: what about if  they move to git in bitbucket?10:38
paulsherwood2madhu: no10:39
paulsherwood2why do you want that?10:39
paulsherwood2if the build fails, the whole build environment including the source directory will be saved as /src/tmp/failed/$randomstring/10:40
paulsherwood2unfortunately morph reports this as, for example 'ERROR: In staging area /src/tmp/staging/tmpPHjEtD'10:41
paulsherwood2the linux source in the above example (assuming it's linux that has failed) will be in /src/tmp/failed/tmpPHjEtD/linux-x86-64-generic.build10:42
madhuok, I dont have usecase right now, but I thought having build/ is useful at some times.10:42
paulsherwood2you could chroot into /src/tmp/staging/tmpPHjEtD, then cd into 10:43
paulsherwood2and from there manually run make etc10:43
straycatpedroalvarez, I'm not entirely certain which situation is the most likely, but no one's given me a good reason not to use -bitbucket as the suffix.10:44
paulsherwood2i'm still trying to understand what this is, i.e. why we should lorry it?10:44
* paulsherwood2 is not a bitbucket user, is all10:44
radiofreepedroalvarez: did you get chance to test my nouveau patches?10:44
radiofreei know richard has been using them, rdale, +1 them!10:45
petefothwhat is the preferred way to contact w.b.o admins to add ssh keys? What would be the criteria for deciding whether or not a user should have direct push access? 10:45
pedroalvarezradiofree: I didn't, and I should do it10:45
* pedroalvarez looks for a jetson10:45
straycatpaulsherwood2, I'm planning to monkey patch setuptools so I can get build dependencies from a python package, regardless of whether that works out I thought it would be useful to have the development setuptools in baserock.10:46
straycatIf there's some doubt about the usefulness of lorrying this then I can work on my patch first then resubmit the lorry request if I need to.10:46
*** ridgerunner [] has quit []10:46
paulsherwood2petefoth: has rule on how to become a maintainer, which is effectively what having push access makes you. previously new candidates have been cited on the mailing list10:47
* straycat decided he's wasted enough time trying to lorry the stuff he wants and forks the repo on bitbucket10:47
madhupaulsherwood2, thanks for the information, so in case if there is failure, the current way is to browse through  the /src/tmp/failed to know which chunk/chunks failed.10:47
paulsherwood2pedroalvarez: i have one i could loan you10:47
pedroalvarezpaulsherwood2: how much?10:47
paulsherwood2to you, free :)10:47
*** ridgerunner [] has joined #baserock10:47
rdaleradiofree: are you in the office today - i can't seem to get my jetson build to boot anymore10:48
paulsherwood2i'll need it tomorrow morn gin though10:48
radiofreerdale: yes i am10:48
paulsherwood2madhu: if a chunk fails, morph will make it very clear which chunk it is. and the build directory will be setup for only that chunk10:49
paulsherwood2s/build directory/failed build directory/10:49
* straycat uses bitbucket mainly because it's free (as in price)10:49
paulsherwood2vs github?10:49
*** persia_ [quassel@ubuntu/member/persia] has joined #baserock10:49
straycatAt the time, github didn't offer private repos to non-paying customers, where as bitbucket did, so I went with bitbucket.10:50
paulsherwood2so to get back, setuptools. are we lorrying the wrong upstream, straycat ?10:53
straycat"In addition to the PyPI downloads, the development version of setuptools is available from the Bitbucket repo, and in-development versions of the 0.6 branch are available as well."10:54
petefothI’ve pushed a test file to w.b.o using the anonymous url. It works even though says that anonymous pushes ‘will be subject to verification.’ Do we need to tighten this up?10:54
straycatthat "0.6 branch" is at and we're lorrying from svn.python.org10:54
straycatSo I'm not really sure we're lorrying _the wrong_ one but most of the development seems to happen in the bitbucket repo10:55
persiapetefoth: Only in the event that we begin to discover trouble with content management.  Being too tight in advance doesn't help much.10:55
persiastraycat: So trunk is the authoritative place, but everyone is using git, and occasionally someone sends a batch of commits to svn?10:56
straycatpersia, Probably but I'm not certain, I haven't asked.10:57
petefothpersia: Fair enough. Do we need to change the text that says it *will* be subject tio moderation? I can pout the right words in the "Editing the Wiki" page I’m working on , but I don;t think I can edit the text on
persiapetefoth: Ah, yes, perhaps we ought use "may" :)10:58
abdul_paulsherwood2 : i can see linux repo in /src/cache/gits/git___git_baserock_org_delta_linux. but i could find the kernel source10:59
persiastraycat: I'm a bit nervous about having two different setuptools, for confusion reasons, but if there's a real difference that upstream indicates is the case, any uncertainty I would express would be removed.10:59
abdul_paulsherwood2 : i can see linux repo in /src/cache/gits/git___git_baserock_org_delta_linux. but i could not find the kernel source11:00
straycatpersia, *nod* I just figured I didn't want to start patching in a repo that hasn't been updated in over a year11:01
straycat(the svn hasn't been updated since 2013)11:01
abdul_paulsherwood : whereas in yocto i can find the kernel source in "build4.0/tmp/work/eagle_imx6-oe-linux-gnueabi/linux-imx/3.0.35-r73.14/git"11:01
paulsherwood2abdul_: that's a bare git repo. again, no user-servceable parts inside11:01
paulsherwood2if you want the linux source, either do morph edit, or git clone 11:02
paulsherwood2bare git repo means nothing is checked out11:02
* paulsherwood2 would be ok with *switching* to take setuptools from bitbucket, but less keen on having both lorries separately11:05
paulsherwood2abdul_: actually another trick that i use sometimes if i'm debugging a build is to just put 'exit 1' as the first instruction in configure-commands: in the morph file for the chunk i care about. this causes the build to fail.. then i can chroot into the failed area and explore, run the build etc 11:11
madhumorph edit linux-x86-64-generic takes long time. 11:12
paulsherwood2first time, it's cloning the linux repo from git.baserock.org11:13
madhuhmm, I must be having them at /src/cache right?11:14
paulsherwood2shouldn't take more than a minute or so normally... it's doing a git checkout11:14
paulsherwood2actually, yes - i think so11:14
madhuabdul_, how about you11:14
paulsherwood2how long is 'a long time'11:14
madhuI guess its 5 mins already, I didn't put time.11:15
paulsherwood2 11:16
paulsherwood2madhu: which is the last line of morph output you are seeing?11:16
paulsherwood2it took only 27 seconds for me as you can see from the above paste. sometimes it can take longer though. it does check some stuff with the trove11:17
straycatpaulsherwood2, I don't feel confident removing the existing lorry, from what little I've been able to gather it is still maintained for those who need it; some baserock users might be using it.11:21
straycatit - the svn repo we're lorrying from at the moment11:23
franredstraycat, IMHO you can lorry it, I don't like the -bitbucket suffix but it is just an appreciation. If we try to replace the svn one with the mercurial one, not sure if the sha1s will match (surely not) and we couldn't reproduce previous built AFAICT11:26
pedroalvarezindeed, we shouldnt remove the old one11:27
franredstraycat, +1 to merge your lorry file regardless of the name11:28
KrinI swear ant is trolling me, failed on the last build step for the system ¬.¬11:29
pedroalvarezKrin: build log?11:29
Krinwhere would i find the buildlog? i'm just seeing the error message11:30
franredKrin, you can copy the error message in a pasterbin and also /src/morph.log, tough11:31
radiofreerdale: revert
radiofreerdale: and for mesa, upgrade the ref in strata/mesa-common.morph to 3b6a4758fa8958db4b76e6d7efccc93b12b1da0611:35
rdaleok thanks11:35
*** hari__ [~hari@] has quit [Quit: Ex-Chat]11:35
*** ssam2 [] has joined #baserock11:37
Mode #baserock +v ssam2 by ChanServ11:37
Krinfranred, pedroalvarez should i pass the log file to you or is there somewhere i should put it into the bin from? it's huge even after clearing it out and doing a fresh run11:39
franredKrin, we only need the error trace to see if we can help you11:40
franredplease, use or other public pasterbin11:40
abdul_madhu: with "morph edit" i could able to get linux source in less than 30 secs11:41
madhupaulsherwood2, it stopped at "updating git repository upstream:linux in cache", I guess I did something wrong11:41
madhuI execute the morph edit from another ssh session, 11:42
aananthhow to find the version of baserock running on a machine or vm11:43
madhubtw if I cancel that and run again in the ssh shell where I built, I am getting this11:43
Krinnot sure how much help it's going to be, i'm installing it the same way that i installed it manualy before, but now it's telling me that a file is missing, if you can spot something though be much appreciated11:43
madhu2014-10-29 11:40:54 Editing linux-x86-64-generic in bsp-x86_64-generic stratum11:43
madhuERROR: Command failed: git config -z morph.repository11:43
ssam2aananth: there are metadata files in /baserock that tells you what commit everything was built from11:43
ssam2aananth: the pattern '/baserock/*system*-rootfs.meta' will find the metadata for the system11:44
ssam2aananth: the 'repo' and 'sha1' fields are the ones you want: you can use those to look up exactly which commit was built11:45
franredKrin, could you paste also your ant chunk please?11:46
ssam2madhu: hmm, that seems like something went wrong with the previous 'morph edit' command11:46
madhuis there a way to recover11:46
ssam2delete the upstream/linux folder and run 'morph edit' again11:46
radiofreepaulsherwood2: you've used ssh-rsync to jetson boards recently?11:50
radiofreeERROR: ssh-rsync.write failed with code 1: ERROR: Command failed: ssh root@ -- blkid -s UUID -o value /dev/root11:50
radiofreehmm there's no /dev/root11:51
ssam2/dev/root :(11:51
ssam2doesn't exist on my Fedora system and I believe is generally being excised from Linux11:52
ssam2so I think we should fix that code to not rely on the existance of it11:52
madhussam2, it hangs now at "Updating git repository upstream:linux in cache", is there a verbose mode to see what's happening.11:53
franredKrin, you have a typo on "sh -Ddist.dir=/urs/lib/antdist dist"  I think you need "/usr/lib..."11:53
ssam2madhu: in this case no because it's just running `git remote update origin`11:53
franredKrin, but the error message does not match with that11:53
paulsherwood2radiofree: yup, but not with any of your latest work.11:54
ssam2madhu: but you can try doing 'morph edit linux --no-git-update'11:54
franredKrin, also if your install commands are empty they should have "[]" after ":" --> install-commands: []11:54
paulsherwood2may last attempt, which worked, was screen captured on sunday .... see
paulsherwood2radiofree: ^^11:54
radiofreeblkid -s UUID -o value works for me11:55
ssam2madhu: oh, `morph edit` seems to ignore the --no-git-update flag, sadly11:55
ssam2madhu: I shall look at fixing this for you guys today, it seems this hasn't been tested well behind a high-latency internet connection11:55
franredKrin, does java/ant chunk also depends on java chunk in your stratum? (could you paste your zookeper stratum too)11:55
ssam2madhu: sorry for the fact that you have to keep waiting around for it!11:56
pedroalvarezKrin: can you paste here the chunk morphology that you are using to install ant?11:56
radiofreepaulsherwood2: i doubt it will work with my never kernels11:56
radiofrees never/newer11:56
paulsherwood2well it worked with mainline?11:57
pedroalvarezKrin: oh, you already did that, sorry11:57
Krinfranred, what is the typo? i cant see it? i'll add the []'s and which one would you like me to give you franred? the whole strata or something else?11:57
paulsherwood2radiofree: what dark magic are you weaving? ;)11:57
franredKrin you put "urs" rather than "usr"11:57
franredKrin, yes could you paste the whole stratum?11:58
paulsherwood2ssam2: i really like your email on  the ansible approach to conditionals11:58
paulsherwood2in fact it's easily my favourite email in that whole thread :-)11:58
madhussam2, no problem, I am trying now  with no-git-update, 11:58
pedroalvarezKrin: also, what does `sh -Ddist.dir=/urs/lib/antdist dist` do? Install ant in  /urs/lib/antdist?11:59
radiofreeERROR: Command failed: ssh root@ -- mount /dev/root /tmp/tmp.RAZimO11:59
radiofreemount: special device /dev/root does not exist11:59
Krinyes pedroalvarez that is the function of the build.sh11:59
ssam2paulsherwood2: I quite like Ansible :)12:00
madhussam2, but mine again hangs at he same place with "--no-git-update"12:00
* radiofree manually creates the symlink for now12:01
radiofreei think we need a better way of finding the root device12:01
pedroalvarezKrin: are there any command in the install-commands?12:01
ssam2madhu: sorry, yes. That flag is not working for morph edit :(12:01
Krinpedroalvarez, franred, stratum here and god damnit i swear i fixed that typo 3 times already >.<12:01
paulsherwood2ssam2: me too12:01
ssam2madhu: you can see what Morph is up to by doing 'tail -f /src/morph.log'12:01
ssam2madhu: it'll probably be sat in a `git remote update` command12:02
paulsherwood2i guess i should offer a patch to fix morph edit?12:02
ssam2or i can12:02
paulsherwood2you'll be faster :-)12:02
paulsherwood2i hereby +1 your result :-)12:02
pedroalvarezKrin: can you try this one instead/
pedroalvarezKrin: note I've removed "java/" from the name of the chunk12:03
Krinok pedroalvarez i'll try that one *is feeling rather useless now that he seems to need so much help*12:04
franredKrin, try to not use "/" in names, replacing them by "-"12:05
ssam2madhu: what happens if you 'git remote update' in an existing checkout that you have, like definitions.git ?12:05
ssam2Krin: in fairness I did tell you the other day to use / in the name of the Lorry entry. / is OK in repo names but it is not OK in chunk names12:05
franredKrin, also Camelcase "javaAnt.morph" should match with the name of the chunk, I would prefer somethink like java-ant.morph, but this is personal12:06
ssam2Krin: although morph should not crash in that case like it did!12:06
Kringah so many naming rules that have no consistency across things >.<12:06
franredyep, it looks a nasty error12:06
ssam2Baserock is pretty consistent with regards to camelcase: don't ever use it, ever :)12:07
richard_mawexcept class names in the morph codebase itself12:07
* Krin facedesks12:07
ssam2that's TitleCase :)12:07
madhussam2, no issues, let me wait till it completes, btw here is the tail message, 12:07
madhu<ssam2> Baserock is pretty consistent with regards to camelcase: don't ever use it, ever :)12:08
petefothKrin: that’s faceDesks - its a method!12:08
madhusorry about my  clipboard error12:08
ssam2no problem!12:08
madhu2014-10-29 11:57:37 DEBUG run external command: [['git', 'rev-parse', '--verify', 'd67a0e110187abd560a1de63fa172894a52839d5^{commit}']]12:08
madhu2014-10-29 11:57:37 DEBUG # git rev-parse --verify d67a0e110187abd560a1de63fa172894a52839d5^{tree}12:08
madhu2014-10-29 11:57:37 DEBUG run external command: [['git', 'rev-parse', '--verify', 'd67a0e110187abd560a1de63fa172894a52839d5^{tree}']]12:08
madhu2014-10-29 11:57:37 DEBUG # git clone /src/cache/gits/git___git_baserock_org_delta_linux /src/workspace/default/upstream/linux12:08
madhu2014-10-29 11:57:37 DEBUG run external command: [['git', 'clone', '/src/cache/gits/git___git_baserock_org_delta_linux', '/src/workspace/default/upstream/linux']]12:08
madhupastebin is access denied here :)12:09
ssam2oh :(12:09
ssam2well, it's cloning from your disk to your disk12:10
ssam2which should be really fast12:10
pedroalvarezmadhu: ?12:10
madhuhehe, that works12:10
pedroalvarezwe should create a page12:10
ssam2madhu: could you try the command that it's doing manually? so `git clone --verbose /src/cache/gits/git___git_baserock_org_delta_linux /src/test-linux-clone`12:11
ssam2it's strange that this command wouldn't work, maybe the IO of your laptop is really overloaded12:11
* paulsherwood2 wonders if now would be a good time to decry the over-use of classes....12:13
* paulsherwood2 decides not :)12:13
madhuyes, "Checking out files" taking lots of time12:13
madhuI am running i5+4gb pc12:14
straycatyes let's use closures for everything12:15
paulsherwood2straycat: trolling is unbecoming of one so intelligent :)12:15
straycatso you're saying it's fine for me to troll here? :)12:16
rjekI like closures.12:16
rjekMore than classes, anyway.12:16
paulsherwood2no, quite the reverse. but smileys make it all ok, allegedly :-)12:16
paulsherwood2rjek: given that you've repeatedly told me 'i'm not a software engineer' how should i value that opinion? :)12:17
rjekpaulsherwood2: I'm not an architect either, but I prefer brick over concrete for exterior rendering.12:17
*** thecorconian [] has joined #baserock12:18
Krinpedroalvarez, i love you i think it worked XD12:18
pedroalvarezit's easy once you understand what is going on12:20
* paulsherwood2 googles closures, notices that it's 'functional programming', momentarily flashes back to his university course on functional languages in the early eighties which just highlighted that mortals can't understand them, reminds himself of the software commandments, and drops closures back out of his mental lexicon.12:20
ssam2just read some node.js code12:20
rjekClosures are often available in functional languages.  But it's more about variable scoping than functional headaches.12:20
* paulsherwood2 has successfully understood node.js code without grokking closures. maybe he did it wrong12:21
pedroalvarezKrin: so, the thing is, if you are trying to *install* something, you have to use the $DESTDIR prefix.12:21
rdaleruby uses closures a lot, but calls them 'blocks'12:22
rjekpaulsherwood2: IMO, closures are actually simpler to understand than non-closures.  They're just functions whose local variables are related to where you define the function, not some other scheme.12:22
paulsherwood2aha :)12:22
paulsherwood2in that case, just another long name for something i think i do understand, which is scope12:23
rdaleyou can pass closures around and they still capture their original scope12:24
paulsherwood2like 'idempotence' is a smartass word for 'make this safe so i can kill this, and or run it repeatedly without the sky falling in'12:24
rjekpaulsherwood2: Lexical scoping is wonderful!  Something Ruby gets right.12:24
* paulsherwood2 should do some actual work :)12:25
* pedroalvarez looks at to deploy a paste.baserock.org12:31
radiofreejjardon: "error while loading shared libraries: libatk-1.0.so0: cannot open shared object...."12:32
radiofreefrom gtk3-demo12:32
jjardonradiofree: you have to add gtk-deps in your system12:33
* jjardon manages to enagle the EGL context in qemu12:33
jjardonits extremely slow though :P12:34
radiofreejjardon: where's the gl demo?12:52
jjardongtk3-demo -> opengl area12:52
jjardonradiofree: run GDK_ALWAYS_USE_GL=true gtk3-demo13:00
*** tiagogomes [~tiagogome@] has quit [Ping timeout: 260 seconds]13:02
*** SotK [] has joined #baserock13:02
aananthsam: I added strata/genivi-foundation.morph to 'devel-system-arm7lhf-jetson.morph' file and gave a I get "ERROR: <urlopen error [Errno -2] Name or service not known>" Any clues, why? How do I debug this?13:19
ssam2aananth: hmm! could you paste the tail of /src/morph.log to or somewhere13:20
ssam2last 30 or 40 lines of the log file should be enough13:21
aananthsam: here is the url
ssam2aananth: seems it hit a network error in the `morph build` command13:25
radiofreewrong keyboard...13:26
ssam2aananth: if you try the command again after checking your networking is OK, it will hopefully succeed13:26
aananthsam: I could ping google.com13:26
ssam2can you ping ?13:26
ssam2could be that the problem is there ...13:26
aananthsam: yes "64 bytes from seq=0 ttl=53 time=225.198 ms13:27
rjek225ms?  Ouch!13:27
rjekThat is not going to make git pleasing to use, with that sort of round trip time.13:27
ssam2rjek: Yes. I believe that's why these folk are having issues13:27
ssam2Nobody has done work to make Morph usable on high-latency network connections13:28
aananthOk, let me put jetson inside visteon network.13:28
madhussam2, can I see the chunks compilation log live when I build the system image, I see I can read it by /src/cache/tmp*** for individually, 13:29
aananthsam: abdul sent you an email regarding booting baserock image on our target system.13:29
radiofreegtk3 and opengl on a tegra
rjekQuake 3, now?13:30
ssam2aananth: I've just replied to Abdul, thanks for pointing it out13:32
radiofreerdale: +1 my nouveau patch!13:33
ssam2madhu: regarding a live compilation log, you can 'tail -f /src/morph.log' to see live build output13:33
madhuyes, but that did not show the compile log for each chunks, I mean what argument, and which files it compiles right now13:34
paulsherwood2radiofree: which version of GStreamer are you using on Jetson?13:38
radiofree0.10 for qt5 and 1.0 for everything else13:39
radiofreegstreamer 1.2.4 even13:40
paulsherwood2so 1.2.4?13:40
radiofreeapart from qt, yes13:40
paulsherwood2so we have two gstreamers coexisting? and is there any reason for 1.2.4, as opposed to later versions, e.g. 1.4.3 ?13:42
paulsherwood2radiofree: ^^13:44
ssam2madhu: oh, sorry, I thought that would show you the full build output13:45
ssam2let me investigate13:46
radiofreepaulsherwood2: probably no reason, we don't actively use it so i guess no one has got round to upgrading it13:46
madhussam2, sure. thanks13:50
ssam2seems that I was wrong, it only shows the exact build commands if something went wrong13:52
ssam2would you find it useful if there was a way to see the build log 'as it happens' ?13:53
madhuyes, I think it is a useful feature13:54
madhuI find that gives better control on what we do, at least I feel that way in "bitbake -v"13:55
ssam2OK! I'll note that done on my list of things we should fix in Morph13:55
ssam2*note it down13:55
madhuok thanks13:56
paulsherwood2ssam2: from my point of view, i prefer the tidy output of morph as is. if i want to track all of the splurge from a build, i'll tail it13:58
paulsherwood2(in the log)13:58
ssam2you can't tail it in the log13:58
*** mariaderidder_ [] has joined #baserock13:58
paulsherwood2are we talking about the same thing? maybe i'm confused13:59
*** mariaderidder [] has quit [Ping timeout: 256 seconds]13:59
ssam2the log just shows the `linux-user-chroot --blah make` command being executed13:59
ssam2it doesn't show any actual build output, unless the build failed13:59
ssam2I agree dumping all build output to stdout by default wouldn't be great14:00
paulsherwood2i repeat my point about preferring tidy morph output - it's too noisy already for me :)14:00
*** VVIBIN [~VVIBIN@] has quit [Ping timeout: 240 seconds]14:05
*** paulsherwood [] has joined #baserock14:11
*** paulsherwood2 [uid50097@gateway/web/] has quit []14:11
richard_mawI've pondered this before, and we could have it display the last 10 lines of the current build as it's in progress, but once it's finished blank those lines and move the cursor back up, so it's not left in the terminal after the build is over, so you just get a clean bit about the build succeeding14:12
straycatSo many differing opinions, I was thinking yesterday that morph should by default output what's going on with committing uncommitted changes etc14:13
Kinnisonrichard_maw: without a curses-like UI, that kind of behaviour is fraught against line width etc14:13
paulsherwoodrichard_maw: that sounds like an interesting approach14:13
straycatI don't think morph is that noisy at all, it already hides most of what's really going on14:14
jjardonpaulsherwood: 1.4.x is the new stable release and its api compatible with the previous 1.x version, so not reason to not upgrade14:25
jjardonI can do a patch tonight if no one is faster than me ;)14:26
*** tiagogomes [~tiagogome@] has joined #baserock14:28
*** aananth [~caananth@] has quit [Ping timeout: 240 seconds]14:35
paulsherwood+1 from me, already :)14:38
*** thecorconian [] has quit [Quit: Leaving.]14:42
*** aananth [~caananth@] has joined #baserock14:52
rdalefor each qt5 module i have a chunk morph with custom build instructions for installing html docs. the commands are always the same - is there anything wrong with just having one of them and refering to it for each of the qt5 modules?14:52
richard_mawcurrently we can't do that because the name: field in the chunk needs to be the same as the name in the stratum14:54
pedroalvarezrichard_maw: nope, it doesn't need to be the same14:54
pedroalvarezI mean. morph won't crash because of that14:54
*** franred [] has quit [Quit: Leaving]14:55
pedroalvarezthe problem is that you won't know what module you are building because morph uses the name written in the chunk14:55
*** thecorconian [~thecorcon@] has joined #baserock14:58
rdaleok, i'll leave them as they are14:59
*** franred [] has joined #baserock15:00
*** thecorconian [~thecorcon@] has quit [Ping timeout: 245 seconds]15:03
*** genii [~quassel@ubuntu/member/genii] has joined #baserock15:07
jjardonradiofree: \o/15:08
petefothI’ve reworked the SideBar on w.b.o, but I’m not sure about the results. All opinions / constructive criticism welcome :)15:21
* Kinnison approves of the sections15:21
Kinnisonpetefoth: shouldn't 'get baserock' go above 'quick start' ?15:22
petefothKinnison: QuickStart includes instrcutions to Download. Get baseroc is just a link to ‘Downloads’15:23
petefothKinnison: for people who want to do stuff woithout follwing uinstructions :)15:24
Zaraooh, I was going to suggest using 'troubleshooting' as a catch-all for errors but it's already there! :)15:24
Zarathe black hilighting on the word is a little odd, though, because it makes it look like you're on that page15:25
Kinnisonpetefoth: fair enough, then the current ordering is good15:26
petefothZara: I know what you mean - it’s black because it isn’t a link. Maybe I shoudl create a ‘Troubleshooting page which just links to the two sub-pages15:26
Zaramaybe it'd be clearer without the bullet point, but still black?15:27
Zarabut that would be fine too15:27
petefothbut then it’s not obviously aprt of the ‘Using Baserock’ section - give me a minute15:27
petefothZara:  is that better?15:31
ZaraI think so; and the nice side-effect of the indentaton below is that it's easier to find at a glance15:33
petefothZara: ta!15:34
* pedroalvarez notices that wikicat is not here anymore...15:36
* pedroalvarez brings cat food to #baserock15:36
straycati don't think wikicat's coming back, it left in despair you see, so i don't think it will ever cat wikis again.15:41
pedroalvarezstraycat: was its source open?15:41
radiofreeafter upgrading baserock on my jetson things look... at bit odd
radiofreeif i go back to factory it's all lovely and nice again15:44
rjekANSI escape codes?15:44
pedroalvarezradiofree: is this a devel system?15:44
rjekTERM=vt100 ?15:44
Kinnisonlooks like the pager doesn't understand the terminal15:44
radiofreei don't really remember doing anything to it before15:45
KinnisonI'd hazard a guess at 'different pager'15:45
Kinnisonor failing that, different terminal setup15:45
radiofreeKinnison: i'm not sure what that means :\15:45
Kinnisonis TERM the same in both cases15:46
Kinnisonis there terminfo for the TERM value15:46
Kinnisonis the pager (less, more, most, whatever) the same one in both cases?15:46
radiofreeTERM is xterm on both15:47
radiofreeless looks the same on both as well15:48
Krinhmm, how do you permanently (alter / add to) the path on baserock as I don't see a .bash_profile to edit?15:48
Kinnisoncreate one15:48
Kinnisonwe don't have a skeleton homedir15:48
radiofreehmm i don't have vim on this image though15:48
Kinnisonthat's all15:48
radiofreepedroalvarez: it's a genivi image without strip-gpl315:49
* Krin gives a thumbs up "so just .bash_profile in ~15:49
pedroalvarezradiofree: then the version of less is different15:49
KinnisonKrin: probably15:49
pedroalvarezradiofree: that's why I asked, I had this situation a week ago and it was because the system was a base-system, without "less" installed15:49
pedroalvarezso, the genivi system is using busybox less I believe15:50
radiofreei do have less installed though?15:50
*** sherm_ [] has quit [Quit: Leaving]16:03
*** franred [] has quit [Ping timeout: 244 seconds]16:06
*** aananth [~caananth@] has quit [Quit: Leaving]16:16
*** franred [] has joined #baserock16:17
ssam2richard_maw: I just noticed that the Yarn step 'GIVEN a git server' sets 'trove-host= []' in morph.conf16:30
ssam2is that some magic that I don't understand ?16:31
ssam2I made a change to how 'wget' is called and it seems to have broken the yarns, but I don't really understand how it worked before, either16:31
ssam2oh, I wonder if the issue is just that something is expected there to be nothing on stderr16:32
*** Krin [] has quit [Remote host closed the connection]16:34
radiofreejjardon: upgrading wayland (not weston) to wayland 1.6.0 works with wayland-ivi-extensions as well16:38
jjardonradiofree: nice to know, thanks for checking it. I will send my patches tonigth then16:39
ssam2richard_maw: ok, it's actually a problem in the distbuild code that it thinks if serialise-artifact produced any text on stderr then it must have failed16:39
ssam2another case where having distbuild in the test suite saved me from actually breaking distbuild !16:39
pedroalvarezssam2: heheh, you are not allowed to break distbuild twice in a week16:40
richard_mawssam2: not too sure what trove-host= [] does, it may be some ipv6 thing for all I know16:50
Kinnisonmakes an empty list perhaps?16:50
richard_mawthis is in the morph.conf16:51
richard_mawI don't think that's valid .conf syntax16:51
*** thecorconian [~thecorcon@] has joined #baserock17:02
rdalewe don't seem to have got libs like gio, gobject and gio-unix for building gnomey libs like geoclue that i am looking at. they could go in 'foundation.morph', but i'm not sure what that is supposed to be a foundation of17:09
paulsherwoodrdale: jjardon is working on getting gnome tidy here, he may advise17:10
jjardonradiofree: thats part of glib17:10
jjardonso its already in foundation17:10
jjardonrdale: ^17:10
rdaleah ok, i was having a look and couldn't see gio for instance17:14
rdaleit is inside the glib chunk?17:14
jjardonrdale: yes17:14
KinnisonI believe gio is part of glib17:14
Kinnisonas is gobject IIRC17:14
rdaleah ok17:14
* paulsherwood wonders what's involved in getting baserock running on Acer CB5-31117:17
radiofreepaulsherwood: ask in #tegra?17:17
radiofreeor maybe cyndis will know?17:18
* paulsherwood was hoping radiofree would know :)17:18
cyndissorry, i don't have one (yet)17:18
cyndisbut i think you should be able to boot u-boot on one the same way as jetson17:18
cyndisas for flashing u-boot, i believe you have to unscrew a screw to make the bootloader flash read-write17:19
franredpaulsherwood, wasn't Adnan the last who run baserock in an chromebook?17:19
cyndisit might also be possible to boot using google's bootloader (depthcharge? coreboot?), but i'm not familiar with that17:20
cyndissomeone on #tegra probably knows better :)17:20
richard_mawfranred: Adnan didn't re-flash the bootloader, he just copied the genivi rootfs's files on17:21
radiofreedepthcharge? is it a requirement to make some kind of submarine reference when writing a bootloader?17:22
cyndisapparently :p17:22
*** ssam2 [] has quit [Quit: Leaving]17:23
cyndisthe tegra chromeos kernel is much closer to upstream compared to the L4T kernel, so i wouldn't be surprised if the stock bootloader could be used as well17:23
franredrichard_maw, oh, I see17:25
rdalehmm, our glib in foundation.morph is too new to build gstreamer 0.1017:26
radiofreecyndis: thanks for the helpe17:30
cyndissure, no problem :)17:30
*** tiagogomes [~tiagogome@] has quit [Ping timeout: 244 seconds]17:38
*** jonathanmaw [] has quit [Quit: Leaving]17:46
pedroalvarezhm... setting up the environment before calling a write extension doesn't do anything :(17:56
pedroalvarezI mean, the write extension doesn't get the environment that I've set up17:56
richard_mawpedroalvarez: how are you running it?17:58
pedroalvarezrichard_maw: that question gives me an idea of what I'm doing wrong17:59
pedroalvarezos.system("/path/to/write-exts arg1 arg2")17:59
pedroalvarezmaybe is better to reuse the logic of ExtensionSubprocess18:00
pedroalvarezI thought I was being clever implementing a context manager to set up the environment.18:01
pedroalvarezAt least now I know how to create context  managers :)18:01
*** mariaderidder_ [] has quit [Quit: Ex-Chat]18:04
richard_maweww, os.system is a gross api18:17
richard_mawis there any reason you're not just generating a shell script that runs `env FOO=BAR… /path/to/write/ext /path/to/sysroot /path/to/location`?18:18
pedroalvarezthere isn't any18:25
pedroalvarezI shall do it in that way18:25
franredrichard_maw, how can I increment the size of a created btrfs image?18:26
pedroalvarezfranred: create a new image, mount both, and copy the content?18:26
franredpedroalvarez, could be, but I think there is a way to extend a btrfs image18:33
richard_mawif it's a disk image for a virtual machine, shut it down, use truncate to make the disk larger18:35
richard_mawthen once you've booted up again, run `btrfs filesystem resize max /`18:35
* richard_maw is a little amused by how many patches have made it to the mailing list after 6PM18:43
straycatradiofree, looks like you accidentally sent as so your mail is pending moderation19:07
*** franred [] has quit [Quit: Leaving]19:47
*** dutch [] has quit [Quit: Quit]19:54
*** VVIBIN [~VVIBIN@] has joined #baserock20:01
*** VVIBIN [~VVIBIN@] has quit [Ping timeout: 245 seconds]20:06
*** thecorconian [~thecorcon@] has quit []21:06
*** SotK [] has quit [Quit: Lost terminal]21:14
*** genii [~quassel@ubuntu/member/genii] has quit [Read error: Connection reset by peer]22:43

Generated by 2.14.0 by Marius Gedminas - find it at!