18:00:36 #startmeeting 18:00:36 Meeting started Fri Nov 25 18:00:36 2011 UTC. The chair is ajmitch. Information about MeetBot at http://wiki.ubuntu.com/AlanBell/mootbot. 18:00:36 18:00:36 Available commands: #accept #accepted #action #agree #agreed #chair #commands #endmeeting #endvote #halp #help #idea #info #link #lurk #meetingname #meetingtopic #nick #progress #rejected #replay #restrictlogs #save #startmeeting #subtopic #topic #unchair #undo #unlurk #vote #voters #votesrequired 18:00:51 * stgraber waves 18:01:02 * highvoltage waves 18:02:39 agenda is at https://wiki.ubuntu.com/AppReviewBoard/Agenda, we'll mostly just go over what's been going on these last few weeks 18:03:32 firstly action items from last meeting - there was one for stgraber to check on packages being copied to the new release 18:04:09 * ajmitch had done a basic check that there were no packages copied to the precise dist on extras.ubuntu.com 18:04:16 so I had to initialize Precise to get extras.u.c working, so apparently everything works as expected (nothing gets copied) 18:04:28 ok, great 18:04:35 glad to know it was simple :) 18:05:44 #topic UDS session 18:06:06 just pulling up the blueprint link now 18:07:16 https://blueprints.launchpad.net/ubuntu/+spec/community-p-app-review-board has the work items there, most of them refer to getting access to myapps.developer.ubuntu.com & to the review shifts 18:08:04 does everyone have access now? 18:08:04 stgraber: did jono hand over the list admin? 18:08:08 next week is my first shift, I hope some of you will be available if I have questions 18:08:11 (which I will) 18:08:31 ajmitch: no, I poked him again two days ago, no news yet... 18:08:36 we have access but it's buggy & broken for accessing tarballs of apps copied from the non-ARB review parts 18:08:49 highvoltage: yeah, I'm taking the weekly shifts as a good time to do my own queue watching 18:08:52 oh, and in person at UDS, though that didn't help apparently :) 18:09:05 there's a LP bug open about the problems 18:09:18 highvoltage: more fun to do together 18:09:25 wendar: great :) 18:09:50 bug 886366 18:09:55 Launchpad bug 886366 in Developer registration portal "Cannot download some arb apps during review" [Medium,Confirmed] https://launchpad.net/bugs/886366 18:10:32 lfaraone: welcome, I just mentioned the bug you opened about not being able to access some arb apps 18:11:33 ah, that's an odd one 18:11:45 currently this bug is blocking reviewing of some apps, along with a few other problems which make the review page annoying 18:12:28 achuni has been on top of these things, so submit a bug (or bugs) 18:12:40 wendar: I'd like to hear from you how your shift went & then I'll get on to my issues with myapps :) 18:12:59 stgraber and I won't get all the same issues, since we have overall admin privs as well as ARB reviewer privs 18:13:33 first shift went well, it was nice to have some focused time 18:13:53 so, I packaged 'crabhack' 18:14:16 and was disappointed to discover at the end that there's a launcher bug in pyglet 18:14:25 (the toolkit it's written in) 18:14:43 otherwise, it's a nice little game 18:15:05 it also needed some fixes to the upstream tarball, so I sent those back to the developer 18:15:46 I have a couple of policy questions from other apps I reviewed 18:16:06 TagPlayer depends on a Last.fm account. The software itself is FLOSS, but the Last.fm terms are proprietary (http://www.last.fm/api/tos). 18:16:22 sure, I've got some as well 18:16:53 wendar: we have last.fm applications in the archive, including the official client. 18:16:56 afaik we have a few clients for proprietary services in main & universe 18:17:10 like ubuntu one. 18:17:18 See http://packages.debian.org/squeeze/last-exit and http://packages.debian.org/squeeze/lastfm 18:18:12 highvoltage: last.fm is a bit more proprietary than even that, it limits what you can do with the data you get 18:19:08 given the precedent of existing last.fm clients, I don't think it'll be a problem to include another app 18:19:11 ajmitch: ah, that doesn't seem much different than accessing a non-free web page from a free browser, does it? 18:19:28 ajmitch: right, but ftp-masters thinks the clients are DSFG-free. 18:19:46 highvoltage: though a browser isn't limited to just those pages :) 18:20:15 lfaraone: which is generally what counts - I think we follow the normal rules for what's allowed in extras 18:21:51 wendar: any other problems that came up? 18:22:15 ajmitch: that seems like a sensible ruling, we can put it in the notes for the meeting, and wider review 18:22:33 All the Quickly apps are coming in with python-support dependencies (because Quickly sets them up that way). Should we manually fix those up (it's an easy fix), or let them go for Oneiric, since python-support is still in the archive (though deprecated). 18:23:13 * ajmitch would rather fix them & be rid of python-support, but it means a little bit of extra work 18:23:38 the quickly template should be fixed to use dh_python2 for precise 18:23:48 ajmitch: +1 18:23:52 I submitted a bug to Quickly yesterday 18:23:58 great 18:24:00 but, not a specific deadline 18:24:23 ajmitch: you can note an action item for me to check with doko on when python-support will be removed from the archive 18:24:45 #action wendar to check with doko on when python-support will be removed from the archive 18:24:45 * meetingology wendar to check with doko on when python-support will be removed from the archive 18:24:52 (but, Quickly should migrate ASAP) 18:24:54 aha, that works :) 18:25:10 there are about 1k packages in oneiric using deprecated python-{support,central} 18:25:32 micahg: ok, we'd love to not add to that number :) 18:25:55 just saying, for oneiric, it probably doesn't matter, for precise, would be nice to drop entirely 18:26:09 but I don't think it should be a blocker for getting packages in - it should only take a few minutes to change for a package done with quickly 18:27:24 ajmitch: is it worth updating the packages before we put them in? It really is only a few minutes of work 18:27:39 * ajmitch has his review session today & will try & mark a few apps for rejection, or at least asking for them to submit source :) 18:27:53 wendar: if possible, then yes, I think it should be done 18:28:27 ajmitch: will do that for the two I've encountered so far 18:28:49 and, add in the review notes 18:28:53 ok 18:28:59 wendar: if its programmatically generated code, it should be easy to script the modification. 18:29:22 do we have a place to leave notes about packages for other reviewers? 18:29:26 lfaraone: for that, I'd rather spend the time patching Quickly itself 18:29:42 lfaraone: rather than parsing and rewriting the output from Quickly 18:29:53 lfaraone: (though, certainly possible) 18:30:08 lfaraone: IIRC, Quickly has some features for "upgrade my project" 18:30:31 nifty. might be worth looking in to. 18:31:11 the other project I worked on is 4digits 18:31:26 I discovered that it used to be in Debian, but was orphaned 18:31:49 so, rather than repackaging for Extras, I revived the debian packaging, and submitted it to the DEbian Games Team 18:32:13 with the upstream developer as the primary Debian maintainer 18:32:19 that's been accepted now 18:32:23 wendar: do you have a link to the bug you filed against quickly? 18:32:28 great, that's probably a better option for a number of these packages 18:32:44 so, will be flowing into Universe for Precise 18:33:02 I'll followup with a Backport to Oneiric once it comes through the DEbian sync 18:33:13 ajmitch: yes, I'm thinking about just doing that for crabhack too 18:33:39 (the developer explicitly asked if we could help him get into debian) 18:33:52 lfaraone: just a sec, pulling the link... 18:34:15 * ajmitch still has to finish sponsoring an upload to debian for a former arb submission as well, once I patch the build system a little 18:36:05 https://bugs.launchpad.net/python-distutils-extra/+bug/894582 18:36:06 Launchpad bug 894582 in python-distutils-extra "Python templates should use dh_python2" [Undecided,New] 18:37:15 lfaraone: ah, didrocks reports the fix isn't in Quickly, but further upstream 18:37:49 lfaraone: could be a fun weekend project 18:37:56 hmmm. 18:38:07 might only take a few min to fix there as well 18:38:25 ajmitch: that's all from my review shift 18:38:32 wendar: thanks 18:38:45 #action lfaraone to look into fixing bug 894582 18:38:45 * meetingology lfaraone to look into fixing bug 894582 18:38:46 Launchpad bug 894582 in python-distutils-extra "Python templates should use dh_python2" [Undecided,New] https://launchpad.net/bugs/894582 18:38:59 just so everyone knows, we're not restricted to only reviewing in our shift :) 18:39:26 wendar: if by "weekend" you mean "next weekend", not "this weekend", then sure. :) 18:39:31 it'd be nice if people are looking at submissions as they come in 18:39:54 ajmitch: I just haven't had time since UDS, but I'm sure I'll get into the swing of things 18:40:04 this sort of relates to the 'health of the queue' item, which we can probably all agree is pretty poor still 18:40:09 lfaraone: I meant "generic weekend, as in sometime when you've got a bit of free time and interest" 18:40:24 wendar: works for me :) 18:40:43 I've looked at a few submissions in my spare time recently, and noticed that quite a few of them just don't have source 18:41:05 ajmitch: you can reply to those immediately, requesting the source 18:41:21 ajmitch: no need to wait for a longer review 18:41:22 Do we have published rules for rejection of packages, such as it needs to do something useful? 18:41:50 I even have a fixed blurb of text for "please submit source code" 18:42:06 wendar: right, I've been meaning to do that as I come across them, often I can't download the submission anyway due to the bug linked earlier 18:42:38 I usually do some hunting for the upstream project & look for a source download there if there is one 18:42:49 wendar: do we have things like https://wiki.ubuntu.com/Bugs/Responses ? 18:43:14 I'm collecting them here, but we could split out to a separate page: https://wiki.ubuntu.com/AppReviewBoard/Review 18:43:49 https://myapps.developer.ubuntu.com/dev/apps/399/ is probably one that can be rejected if I can think of a polite way to say "Submission must be more than just boilerplate code" 18:44:10 lolwut. I don't even. 18:44:41 ajmitch: Our guidelines are on that same page, but they don't say anything about "useful". 18:45:00 I wondered about Awesome Os too. 18:45:05 it looks like someone submitted the output of 'quickly create', I didn't spot anything original in it 18:45:19 I love the honest rationale. 18:45:29 * ajmitch is just hunting for the bug now that mpt wrote 18:45:38 I'm torn between "it doesn't do anything", and "well, it's just a silly app, like you'd find on Android, maybe it doesn't need to do anything" 18:45:56 https://bugs.launchpad.net/ubuntudeveloperportal/+bug/873443 18:45:57 Launchpad bug 873443 in Ubuntu App Developer site "No published rules for rejecting applications" [Undecided,New] 18:46:01 it certainly won't make its way into Debian or main/universe 18:46:35 wendar: Of course not. But we should have some sort of rule "somebody somewhere besides the author must find this useful" 18:46:38 probably won't even continue from Oneiric to Precise 18:46:46 wendar: silly app, but it really doesn't look like an original creation beyond running quickly 18:46:55 I don't think we should set the bar that low 18:47:25 ajmitch: see "long tail" in http://kitenet.net/~joey/blog/entry/the_popcon_problem/ 18:47:38 so, proposed addition to our review guidelines 18:47:47 otherwise we may as well accept the hello world 'app' that was used for testing :) 18:47:55 under "Content" 18:48:43 "App should be useful or interesting to a general audience" 18:48:47 not quite right 18:48:55 http://pad.ubuntu.com/arb-guidelines 18:49:00 wendar: we also had mention somewhere of submissions needing to be applications instead of data or documentation 18:49:21 which is important because I saw a recent submission of fonts 18:50:10 https://myapps.developer.ubuntu.com/dev/apps/196/ 18:51:28 right, I found the original reference to it: https://wiki.ubuntu.com/PostReleaseApps/Process 18:52:27 sorry to mess up the pad, I just pasted in the relevant section 18:52:52 no worries. 18:53:09 ajmitch: ah, I was just going to ask if you got that from https://wiki.ubuntu.com/ExtensionRepositoryPolicy 18:53:12 do we need content to be DFSG-free / under a free license? 18:53:41 ajmitch: but then, the second page started from the PostReleaseApps process 18:53:52 Like if I make a game with a non-free soundtrack, or propreitary maps. That's totally cool (explicitly allowed in the GPL FAQ!) but not suitable for main or universe, maybe. 18:54:05 wendar: there are a few too many wiki pages on this 18:54:12 lfaraone: yes, the ARB only reviews fully DFSG-free content 18:54:23 ajmitch: yes 18:54:31 ajmitch: we need to have a full-time wiki gardener. 18:54:39 ajmitch: the PostReleaseApps pages are deprecated, I should add a note 18:54:47 * ajmitch isn't volunteering for that role, fwiw :) 18:54:54 hello all! I'm sorry I came late to the meeting 18:55:06 coolbhavi: glad you could join us, we're still going :) 18:55:10 ajmitch: but, they also include content that isn't anywhere else yet (though, it's also partially out-of-date) 18:55:11 :) 18:56:11 wendar: as long as it doesn't conflict with what we do now it should be ok to refer to 18:56:25 How about linking to the ExtensionRepositoryPolicy from ours, but giving easier short explanations? 18:56:48 I like "Apps should not be forks of existing applications in the archive." 18:56:51 sounds fair 18:56:55 it's quick and easy to understand 18:57:35 wendar, sounds good +1 here :) 18:57:45 wendar: which also covers things which are obviously just GNU hello. 18:58:09 wendar: would it be sensible also to say "should not be a clone of something in the archive"? 18:58:29 wendar: like, if I pull a gnote and implement another application feature-for-feature, etc. 18:58:34 depends on what clone means 18:58:53 'useful or interesting to a general audience' is suitably vague that it gives us some discretion as to what to let in 18:58:53 we'd certainly allow two independent implementations of, say a Last.fm client 18:59:18 but, a straight copy-and-resubmit we wouldn't 19:00:25 there are some specialised applications that most people wouldn't use but would still be useful to some, I think we would let those through 19:00:26 lfaraone, fair enough like we give scope for a wider variety of apps by saying that it shouldnt be an exact clone but as wendar said clone should have an unambiguous meaning 19:01:13 hmmm. 19:03:21 Okay, scratch that. I was trying to come up with a case that would make this rule relevant and it ended up sounding much too contrived. 19:03:45 about that tools & libraries one I just added - we don't really want apps bundling libraries, how far should we go in disallowing it? 19:03:51 * lfaraone will have to duck out. 19:04:05 lfaraone: thanks for being here 19:04:06 ajmitch: you should not bundle a library already packaged. 19:04:06 * highvoltage too in a minute or so) 19:04:35 the problem is when a package needs a compiler which isn't in ubuntu :) 19:04:59 ok we should probably take this one to the list then 19:05:12 ajmitch: like, something for a language we don't support, or just a newer version of an exsiting lang? 19:05:38 lfaraone: new language, a couple are built with some obscure BASIC dialect 19:05:55 ajmitch: it's a good one to add to the list 19:06:01 they're submissions in the queue at the moment 19:06:31 wendar: "must use existing languages"? 19:06:47 wendar: I'm just not sure where to draw the line on libraries, since some can be large 19:07:06 a definite +1 on not bundling new versions of boost or Qt for example 19:08:04 here too I ve seen some of them being sizeable 19:09:16 I can guess the answer to this, but I'd just like to be 100% sure, may an ARB app depend on something in the -backports repository? 19:10:07 highvoltage: we haven't discussed that afaik, I'd vote yes though 19:10:42 ajmitch: it sounds like it could potentially be a sane way to deal with things like new versions of qt 19:11:00 there's a technical problem at the moment, I don't think an Extras package has any way to automatically pull something from backports 19:11:11 highvoltage, yes, but if thats optional, then a +1 here 19:11:31 right, though it's usually tricky to backport libraries because all reverse (build-)depends need to be tested with it 19:11:45 hmmm... even if it could, that's potentially dangerous, updating a system library to a backports version 19:12:15 wendar: afaik it should technically work if there's a versioned dependency satisfied only by backports, as long as both extras & backports are kept enabled 19:12:18 I lack a little history, has the topic of appimages ever come up before? 19:12:28 like, what if the app depends on a newer version of Qt, so pulls it in from backports, and the version of Unity the user is running breaks on the newer version of Qt? 19:12:47 wendar: then that version of Qt wouldn't be allowed into backports 19:12:58 backports tests "build, install, run", but doesn't test that it works with all dependencies 19:13:26 i.e. they don't test every package in the archive that uses the backported package 19:13:42 they should, people are asked to do that when submitting a backport request 19:15:12 https://help.ubuntu.com/community/UbuntuBackports#How_to_request_new_packages has the rules for libraries 19:15:24 "A backport should never cause any other package on the system to break" 19:15:27 that's good 19:15:39 but also "New versions of libraries are strongly discouraged" 19:15:40 it's not a complete cowboy operation ;) 19:15:42 which makes sense 19:16:03 but, also means it's not very useful for the scenario where an app needs a new version of a library 19:16:29 right, but there's no easy way to handle that without shipping a private copy of a system library 19:16:31 because, if the changes between versions are so minimal that they have no disruptions 19:16:50 then the ARB app will likely still support the old and new versions simultaneously 19:17:36 but, in general, it seems fine to allow backports, if we can solve the problem of pulling in backports as a dependency 19:18:19 I don't know of a good way to test it, but I think the biggest problem is in building the package rather than installing on a user's system 19:18:21 for example, a new library that isn't in Oneiric, but is in Precise and gets backported to Oneiric, should be no problem 19:19:25 ajmitch: As I understand it, you have to take a manual step to enable backports for a specific package 19:19:33 PPAs can depend on backports, I just checked on LP 19:19:44 ajmitch: so, how would the Extras package pull in the backports package? 19:20:16 I think this is a question for the backports team 19:20:25 wendar: afaik it's a PPA change rather than per-package - if apt can satisfy the dependency only by backports it will pull in the package 19:20:34 does anyone want to volunteer to chat with them before next meeting? 19:20:40 sure, I will 19:20:58 ajmitch: that part isn't a problem, backports is enabled by default in Oneiric 19:21:14 ajmitch: but, it's enabled by default with no packages turned on 19:21:28 ajmitch: cool, thanks! 19:21:43 wendar: yep, that's set in the Release file as NotAutomatic: Yes 19:22:01 I need to check on the other flag that I know of for that 19:22:25 #action ajmitch to ask about ARB apps depending on backported packages 19:22:25 * meetingology ajmitch to ask about ARB apps depending on backported packages 19:22:34 yeah, specifically "is there a way to specify the backports version of X in the Depends line of a package" 19:22:52 or Build-Depends 19:22:56 we should probably wrap up this meeting soon 19:23:07 a simple versioned (Build-)Depends should cover it 19:23:31 it's the same situation in debian for experimental, I saw a discussion on it a couple of days ago 19:23:32 I'll copy our Content requirements from the etherpad into the wiki page 19:23:38 thanks 19:23:53 AOB? 19:24:18 oh I almost forgot 19:24:25 as an FYI, Anthony Lenton is working now on making our MyApps queue public 19:24:41 (public read-only) 19:24:46 is it worth having a meeting next month, since it falls between christmas & new year? 19:25:08 bad timing 19:25:16 but, I'd hate to go 2 months between 19:25:30 yeah, I may be around but I don't know who else may be 19:25:38 could we do say 2 weeks from now 19:25:40 I can be there. 19:25:44 and mid january? 19:26:01 yes why not have it 2 weeks from now and mid jan? +1 19:26:02 wendar: when do you fly out for LCA? 19:26:31 ajmitch: probably saturday (so the day after that mid-jan meeting) 19:26:45 wendar: remember that it's saturday here for me now 19:26:58 me too 19:27:06 aye, that's why I added the clarification 19:27:16 early morning 1 am here 19:27:17 (my saturday, not your saturday) 19:27:22 wendar: ah right :) 19:27:27 I'll be around and free that Friday between Xmas and New Year too 19:27:49 ok, 2 weeks from now is 9th december, 18:00 UTC 19:27:55 who volunteers to be chair? 19:28:41 wendar: I'll miss the january meeting just because I get to leave ballarat at about 7AM to get to the airport 19:29:03 that's the middle of a Review shift, so should work out (the 9th, I mean) 19:29:10 ok 19:29:24 ajmitch: okay, we can decide at Dec 9th if it's worth a mid-Jan meeting 19:29:27 highvoltage: want to chair the next one? :) 19:30:02 ajmitch: I guess I can chair. We may also want to switch to having the next chair be the next one in alphabetical order (that's how the DMB and TB do it now and saves quite a bit of time) 19:30:26 stgraber: ok, sounds good 19:30:55 with that, I guess we're done, unless anyone has something to add 19:31:29 thanks ajmitch! 19:31:47 thanks everyone, sorry it took so long :) 19:31:54 #endmeeting