/srv/irclogs.ubuntu.com/2012/07/03/#ubuntu-arb.txt

dholbachgood morning06:52
dholbach_ajmitch, thanks for the review06:56
=== dholbach_ is now known as dholbach
dholbachI just commented on it again, as I wasn't quite sure how to proceed06:57
PaoloRotoloGood morning!07:04
dholbachhey PaoloRotolo07:07
PaoloRotolodholbach, hi!07:11
PaoloRotolodholbach, I tried to post a feedback on the app page: https://myapps.developer.ubuntu.com/dev/apps/1032/  but it says that I haven't permission07:18
dholbachoh, I had no idea :/07:19
dholbachcan you put up your comment on a pastebin again, then I'll try to put it in there07:20
PaoloRotolodholbach, ok07:22
PaoloRotolodholbach, here it is: http://paste.ubuntu.com/1072675/07:27
dholbachPaoloRotolo, ok, this time it worked: https://myapps.developer.ubuntu.com/dev/apps/1032/feedback/07:29
dholbachyou might have to be member of ~ubuntu-app-review-contributors07:29
PaoloRotolodholbach, good, thanks :)07:40
ajmitchdholbach: hi, sorry I didn't comment earlier08:51
* ajmitch picked a bad time to get sick08:51
dholbachajmitch, oh damn :-/08:51
* dholbach hugs ajmitch08:51
dholbachI hope you're going to be better soon again!08:52
dholbachajmitch, I updated the merge proposal again - I hope that's what you intended it to be08:52
ajmitchyeah, I seemed to be a bit better tonight08:52
* ajmitch shall have to take a look08:52
ajmitchmutable default arguments are something of a python gotcha :)08:52
dholbachajmitch, it'd be great if we could get it in - then I'd trigger another build and we'd have most of the regular checks covered by a small tool :-D08:53
ajmitchthe list is bound to the function definition & not reset to [] when called with no argument08:53
ajmitchok, will look now then08:53
dholbachajmitch, I'm not quite sure I understand the problem with the default argument08:55
ajmitchlet me throw together a quick test case08:55
dholbachajmitch, I guess I'm just a bit dense :)08:55
ajmitchor I'm just wrong08:58
ajmitchyou're not changing the argument within the function, so you get away with it afaik, so ignore my MP comments :)09:03
dholbachyeah, I just pass it through :)09:03
dholbachajmitch, does "ignore my MP comments" mean "it's approved, merge it" or are you still having a look at it?09:23
ajmitchdholbach: merge it09:27
dholbachgreat09:27
dholbachthanks09:27
ajmitchsorry, I should have marked it as such on the MP09:27
dholbachno worries09:28
* dholbach triggers another build09:29
dholbachthat'd hopefully make reviewing apps a bit more fun :)09:29
dholbachespecially once we add examples for successfully fixed apps09:29
dholbachPaoloRotolo, there should be a new arb-lint version in the ppa now, which also checks if the /opt installation is correct09:55
PaoloRotolodholbach, great!09:57
PaoloRotoloI'll install it at home today09:57
dholbachsuper09:57
PaoloRotolodholbach, the app review is fun :P09:59
dholbachcool, I'm glad you like it :)09:59
PaoloRotolo:)10:00
dholbachajmitch, are we going to ignore things like usr/share/python/runtime.d/harvestwidget.rtupdate and etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf?10:06
ajmitchdholbach: they're added by dh_python2 & dh_apport respectively, from what I know10:07
ajmitchI think we'd need to override those to stop them being created10:07
dholbachajmitch, and that's what we'd want?10:09
dholbachI guess what I'm asking is: is it unacceptable to have these files in there?10:10
ajmitchto avoid potential file conflicts, it is, since we're trying to stick close to the /opt requirement10:11
dholbachok10:11
ajmitchtechnically I don't know if it's a problem either way, as they have the package name10:11
dholbachhum, I can't find dh_apport10:17
ajmitchit's shipped in dh-apport10:17
dholbachI don't have it installed, but still the crashdb.conf file gets shipped in the package :)10:18
ajmitchthen I'm probably wrong again & should stop misleading you10:18
dholbachno no :)10:18
dholbachaha, python-distutils-extra again10:19
ajmitchoh is it?10:20
ajmitchfun10:20
* didrocks should set an autojoin :)10:20
dholbachhey didrocks10:21
dholbachwe just found files like etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf10:21
dholbachwhich look like they're shipped through python-distutils-extra10:21
didrocksindeed :)10:22
dholbachas many folks who just submitted apps didn't modify these files at all, it might make sense to just override whatever ships them somehow10:22
didrocksthis is to enable running ubuntu-bug <project> against the upstream project10:22
didrockshum, they shouldn't need to modify it10:23
dholbachbut for the future it'd be nice if apport and python-distutils-extra would do something like etc/apport/crashdb.conf.d/extras-harvestwidget-crashdb.conf or opt/extras.ubuntu.com/harvestwidget/etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf10:23
didrocksit should just do the right thing10:23
dholbachah ok10:23
dholbachwell then, I don't know :)10:23
didrockswell, apport doesn't read extras.ubuntu.com10:23
didrocksso the bug is apport to read from there?10:23
dholbachapport and python-distutils-extra10:23
dholbachfirst for apport to check an alternate location10:23
dholbachand then for p-d-e to install to it10:24
didrocksyeah, then to push in the alternate location10:24
dholbachyep10:24
dholbachI'll have a chat with pitti10:24
didrocksI still think that if we wanted to be serious about /opt, we would change all desktop service to run from there10:24
dholbachajmitch, which location would you prefer?10:24
didrocksrun/listen10:24
didrockslike dbus and such10:24
dholbachetc/apport/crashdb.conf.d/extras-harvestwidget-crashdb.conf or opt/extras.ubuntu.com/harvestwidget/etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf?10:24
dholbachyeah10:24
dholbachopt sucks balls10:24
dholbach:)10:24
ajmitchagreed10:24
didrockswell, it's only 2 years that I'm advocating for this, since the ARB was created :)10:25
* dholbach urgently looks for his valerian tea, to calm down again10:25
didrocksno issue with /opt as long as we support it :)10:25
ajmitchdidrocks: advocating for more stuff in /opt, or not at all? :)10:25
didrocksajmitch: everything can go in /opt, but before making this a requirement, we should ensure that apport, dbus, unity lens files, and such are listening to /opt/<whatever>10:26
dholbachajmitch, which of the two location would make more sense to you? etc/apport/crashdb.conf.d/extras-harvestwidget-crashdb.conf or opt/extras.ubuntu.com/harvestwidget/etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf?10:26
ajmitchdholbach: I *suspect* we'd need to ask for an explicit exception to have it in /etc still, going by the letter of the law "_10:26
dholbachthe good thing is: pitti is on the TB ;-)10:26
ajmitchyes, and so is stgraber :)10:26
didrockseverything where it makes sense to have application hooked up and where you put exceptions :)10:26
dholbachalright10:27
ajmitchdidrocks: and then you still run into problems where people want to write a plugin for a certain app10:27
ajmitchwhether it be nautilus, or something else10:27
didrocksajmitch: well, do we want plugins in all our apps?10:27
didrocksajmitch: this is a control gate in fact :)10:27
didrocksto see what we want to enables for ARB apps or not10:28
didrocksenable*10:28
ajmitchright, but we don't support them right now except for lenses as a special case10:28
didrocksyeah, and I think it should stay that way10:28
didrockslike just supporting that for some special cases10:28
didrocksbase system, like dbus, apport, desktop files detection10:29
ajmitchthe restrictions on this could probably be made clearer for the next app developer competition, too :)10:29
didrocks(so bamf, unity…)10:29
didrockssome special ones like unity lens10:29
* ajmitch needs to bug unity people some more to not have to restart for new lenses :)10:41
dholbachok, I talked to pitti10:41
dholbachhe said we should just move the file to opt/extras.ubuntu.com/harvestwidget/etc/apport/crashdb.conf.d/harvestwidget-crashdb.conf10:41
dholbachand apport and p-d-e will need to be fixed10:41
ajmitchproblem is, my friends who work on it tell me to submit a patch10:41
ajmitchdholbach: ok10:41
dholbachI'll add a test and docs to arb-lint about it10:42
ajmitchapport would need an SRU for it to look in the right place then10:42
dholbachyep10:42
dholbachand p-d-e too10:42
ajmitchyippee?10:42
dholbachlet's go shopping10:43
ajmitchsorry for sarcasm, I'm a little jaded on /opt still :)10:43
* dholbach makes a valerian tea for ajmitch too10:43
ajmitchheh :)10:43
* ajmitch really needs to go & sleep soon10:48
ajmitchI'm really sorry I haven't been around much for this apps sprint, this just turned out to be one of my busier weeks10:48
dholbachwe can do it again10:51
dholbachI think we all learnt a lot10:51
dholbachhttps://bugs.launchpad.net/ubuntu/+bugs?field.tag=arb10:51
ajmitchnot too many, but some will take some work10:52
dholbachok, I just went ahead and added tests with docs and examples11:13
dholbachok, lunch time :)11:29
dholbachcoolbhavi, if you install arb-lint from the PPA now, it should do a lot of the testing for you14:13
dholbachhopefully also the /opt stuff14:13
coolbhavidholbach, sure will check it out :) Thanks!14:14
dholbachmhall119, so I should run arb-lint on arb-lint itself? ;-)14:14
dholbachE: The package seems to contain no files in '/opt/extras.ubuntu.com/arb-14:15
dholbach   lint' which indicates a problem with the installation of files.14:15
dholbach;-)14:15
coolbhavi:)14:15
mhall119dholbach: lol14:24
mhall119dholbach: so now we can take a tarball, put it in a bzr branch, create packaging for it, and check that the package follows the extras guidelines14:25
mhall119that's great!14:25
dholbachyeah, we're slowly getting somewhere :)14:30
dholbachhey highvoltage :)14:30
highvoltagehowdy dholbach14:36
dholbachhow are things?14:36
highvoltageI have a scumbag brain.14:36
dholbachwhat what?14:36
coolbhavihey highvoltage14:37
highvoltageI'm a victim of this again today: https://plus.google.com/118060934321492774758/posts/P6Ryggx81MU14:37
highvoltagegot millions of ideas last night and couldn't sleep and now my head is just a black hole.14:37
highvoltagedholbach: "scumbag brain" is a popular meme: http://www.quickmeme.com/Scumbag-Brain/14:38
highvoltagedholbach: and you, how are you?14:38
highvoltagehey there coolbhavi14:38
coolbhavihighvoltage, how are you doing buddy?14:38
dholbachah ok, I had no idea about that meme :)14:39
dholbachhighvoltage, I had a similar phase and two things helped me back then: I kept pencil and paper (not a laptop!!!) next to the bed and pinned the ideas down - also doing just a little bit of sports (not too much) during the day14:40
highvoltagedholbach: I'll try the pencil. I do more and more excercise but that seems to make it worse (especially if I do it later in the day).14:41
highvoltageI finally joined the trello board. It's pretty cool.14:42
highvoltageI wanted to spend more time with arb stuff this weekend but only ended up reviewing one app :-/14:43
dholbachI'll blog about what we did in the last few days later on14:43
dholbachhopefully that'll get us some more interested folks :)14:43
coolbhavi:)14:44
highvoltagegreat. I've been meaning to ask about that, so does that go on the app developers blog?14:44
dholbachhighvoltage, I don't have access to it14:44
dholbachbut I guess we could put something up on an etherpad and get it up on the app dev blog if we all think that's the best place for it14:45
PaoloRotoloHi all!15:07
highvoltagehi PaoloRotolo15:08
PaoloRotolohighvoltage, hi15:08
PaoloRotolodholbach, The new release of arb-lint is cool15:35
PaoloRotolodholbach, it reported 2 errors against koza15:36
PaoloRotoloI'm solving them...15:36
dholbachsweet15:37
dholbachI'm in a  call - I'll be back in a few15:38
PaoloRotolook15:38
PaoloRotoloFixed :D15:50
PaoloRotoloSuccessfully updated my branch16:10
dholbachPaoloRotolo, shall I add a comment for you to some app?16:26
PaoloRotolodholbach, yes, thanks!16:33
dholbachah, that's koza again?16:33
dholbachPaoloRotolo, ^?16:34
PaoloRotolodholbach, yes :) I've other two errors16:34
dholbachperfect16:34
PaoloRotoloafter the arb-lint update16:34
dholbachgood work16:36
dholbachPaoloRotolo, dh_python2 and which other one?16:36
PaoloRotolodholbach, only python16:37
dholbachah ok16:37
PaoloRotolodholbach, and I have fixed a typo in debian/control16:37
dholbachah cool16:38
dholbachgood work16:38
PaoloRotolodholbach, thanks :)!16:38
PaoloRotoloSo, Koza is ready for the software center (IMHO) :D16:42
dholbachhighvoltage, wendar, stgraber, ajmitch: ^ say Hi to a great new contributor to the apps world! :)16:43
highvoltagehi PaoloRotolo!16:44
highvoltagePaoloRotolo: and thanks for all the work, it's much appreciated!16:45
dholbachI will have to rush out in a bit because a friend of mine will have a surprise birthday party in a bit and I'll have to get a present still 0:-)16:45
PaoloRotolohighvoltage, dholbach Thanks a lot :)16:46
dholbachso if somebody could have a quick look over lp:~paolorotolo/ubuntu-app-reviews/koza that'd be cool16:46
PaoloRotolodholbach, Have a nice evening16:47
PaoloRotolothanks for all16:47
wendarhi PaoloRotolo, and thanks!16:47
PaoloRotolo:)16:47
dholbachI just blogged about the last two days: http://daniel.holba.ch/blog/2012/07/making-better-apps-possible/16:47
dholbachthanks everybody for your great work :)16:47
dholbachif I missed something, I'll write another blog post tomorrow :-)16:47
PaoloRotolodholbach, nice, thanks again!16:49
highvoltagewendar: around?17:17
wendarhighvoltage: yup17:17
highvoltagewendar: I think it was you who commented that even MOTU apps only need two +1 from MOTUs17:18
highvoltagewendar: do you think it should be like that for ARB apps too?17:18
highvoltagewendar: I'm thinking that if the arb pretty much agrees with that then we should get the policy changed to keep the vote queue as small as possible17:18
wendarhighvoltage: I think 3 is pretty reasonable for arb apps17:19
highvoltageindeed.17:19
wendarhighvoltage: we just need to get better about actually voting17:19
highvoltage3 is still easy to get to at least.17:19
wendarhighvoltage: I mean, I'm assuming one of the +1s is actually the person who submitted it for vote17:19
wendarso, three all together is really only the same as the MOTU "get at least two others to agree this is a good idea"17:20
highvoltagewendar: ah yes, but that could also be someone from arb-helpers who submitted it17:20
highvoltage(but it would still be fine I guess)17:20
wendaryes, that's why I was making the assumption explicit17:20
highvoltageok, sorry for stating the obvious :)17:21
wendarI'd say at least two of the +1s should be ARB members who have done a code review/security review on the package17:21
wendarthe last one could be more of a sanity check17:21
wendarand really 3 will be easier to get once we recruit a few more ARB members17:22
wendarmainly, I just don't want to get bogged down in policy changes :)17:23
wendarthree +1s is the current rule17:23
wendarso, we can just run with it17:23
wendaranother way of thinking about it (just thinking out loud here)... the MOTU rule of two +1s assumes a third experienced packager in the process: the person who made the package17:24
wendarwe don't assume the *developer* is at all experienced17:25
wendarso we add one more17:25
wendarbut... hmmm17:25
wendarI could see changing the policy so the third is an arb-helper17:25
wendarso, each package has to get two +1 votes from ARB board members, and one +1 vote from the guide who raised it for vote17:26
wendarand, the guide can be either a board member or a helper17:26
wendarthe ARB board members are required to be ubuntu-devs17:27
wendarbut the helpers aren't17:27
wendarso, that makes the whole process more parallel to MOTU17:27
highvoltagewendar: yeah that would be good17:27
wendarhighvoltage: I'd be in favor of that17:28
wendarhighvoltage: we could do it "inofficially" for now, and just have one board member +1 for experienced arb-helpers, without doing a full code review17:29
wendarhighvoltage: like, really, I'd confidently +1 anything dholbach submits as an arb-helper :)17:30
wendarhighvoltage: especially after a couple other board members have done a code review and +1'd it17:31
highvoltagewendar: *nod*17:31

Generated by irclog2html.py 2.7 by Marius Gedminas - find it at mg.pov.lt!