[12:41] <DSG2490> hello
[01:27] <lifeless> BjornT: nag kiko about your branch - I'm giving it to him
[02:36] <crimsun> Is there a reason I should not be able to use ``bzr get sftp://~crimsun@bazaar.launchpad.net/~ubuntu-core-dev/alsa-utils/debian/''?  I'm receiving "Permission denied (publickey)", but my ssh keypair hasn't changed at all.
[02:37] <crimsun> (I'm using ``bzr get http://bazaar.launchpad.net/~ubuntu-core-dev/alsa-utils/debian/'' as a workaround.)
[02:38] <spiv> crimsun: "~crimsun@" looks wrong.
[02:39] <spiv> It should be "sftp://crimsun@..."
[02:39] <crimsun> haha
[02:39] <crimsun> I'm an idiot.
[02:39] <crimsun> thanks
[04:07] <milosz> hey guys i finaly got my po messages picked up by launchpad (woohoo) but now do i export the new translations
[04:07] <milosz> so they get included into my app
[04:07] <milosz> and can i message a user using launchpad
[04:11] <milosz> i'd like to write the guy who translated my app to thank him
[04:12] <lifeless> his details will be in the pot files
[04:12] <lifeless> s/his/their
[04:18] <jamesh> milosz: what's the name of your product?
[04:18] <milosz> drapes
[04:18] <milosz> i couldn't find a thing to go... download new po file
[04:20] <jamesh> milosz: found it.
[04:21] <jamesh> milosz: from https://launchpad.net/products/drapes I chose "translations" on the left
[04:21] <jamesh> that shows stats for the PO file you uploaded
[04:21] <jamesh> click on "View Template & All Languages..." in the main content
[04:22] <milosz> it seams like i only ask when i have problems with launchpad, but all the help is greatly appreciated and i like how i can upload a new translation po file now instead of having to email it to a group and then some shady (dj) guy
[04:22] <jamesh> this brings you to the page for your PO template, and there is a "Download Translations" link in the top corner
[04:22] <milosz> s\dj\jk\
[04:23] <jamesh> so the actual URL where you can request the export is https://launchpad.net/products/drapes/trunk/+pots/drapes/+export
[04:23] <jamesh> that should be easier ...
[04:23] <milosz> i couldn't find the "View Template & All Languages..."
[04:23] <jamesh> it is below the table in the centre of the page
[04:24] <milosz> this is kind of weird, you have to have it be sent to your email... but okay
[04:24] <milosz> thanks for your help jamesh 
[04:26] <milosz> where in my app file/directory should i put the credits to the translators?
[04:26] <bimberi> hi all, how do i request a new LP Project be set up?
[04:29] <spiv> bimberi: find a launchpad admin on IRC here, or mail the launchpad-users list.  But are you sure you want project, rather than a product (which anyone can register)?
[04:30] <bimberi> spiv: thanks, yes i'm sure - there are already a couple of products in place that require an "Umbrella", similar to the ubuntu-docs project
[04:31] <jamesh> milosz: the PO export can take a while for larger projects, so the work is handled by a batch job that emails you when it completes
[04:31] <spiv> Fair enough :)
[04:32] <milosz> okay
[05:00] <mpt> Goooooooooooood afternoon Launchpadders!
[05:05] <jamesh> mpt: we had another Rosetta usability issue come up: milosz trying to export translations for his product
[05:05] <jamesh> not particularly obvious how to get there from https://launchpad.net/products/drapes/
[05:11] <mpt> joy
[05:11] <jamesh> you can get there in three clicks, if you know which links to use
[05:13] <mpt> Yes, I found it
[05:14] <mpt> ugh, that listbox is misaligned
[05:17] <mpt> Anyway, I'm not sure what we can do to make downloading more obvious, other than the usual page layout problems of the facets menu in the wrong place and the actions menu existing when it shouldn't
[05:22] <milosz> man i love how whenever i come in here and have an issue you guys like fix it
[06:01] <Lord_Athur> i all, what do i have to do to become ubuntero in launchpad?
[06:02] <lifeless> have a look at wiki.ubuntu.com/Membership (IIRC)
[06:04] <Lord_Athur> lifeless, the wiki page doesn't exist
[06:59] <xordae> Hey guys, I tried registering at the launchpad page for an error report after a fresh Ubuntu 6.06 install, but it's taking long. Maybe anyone can help me with this error I got just before the install finished? http://img3.freeimagehosting.net/image.php?1da7c08625.jpg
[07:00] <xordae> sorry for the image quality, it reads /usr/lib/python2.4/site-packages/ubiquity/frontend/gtkui.py
[07:05] <jamesh> xordae: do you use greylisting or similar for your incomming email?
[07:05] <xordae> hmm.. it was a hotmail address I provided >< I dunno, tried it again with gmail. but I don't think so
[07:06] <stub> From recent reports gmail works but hotmail doesn't 
[07:06] <jamesh> people have successfully registered using gmail
[07:06] <xordae> oh, alright.. then I'll just report it there now.. hope it's readable
[07:14] <jamesh> xordae: so you should get an email to verify your account. Are you saying that you still haven't received such an email to your gmail account?
[07:14] <xordae> I just checked, got the mail and am making a report.. thanks ^^
[07:15] <jamesh> xordae: you can register additional emails to your LP account once it is set up (you go through a similar verification process)
[07:32] <stub> I'll be updating Launchpad as soon as soyuz as quietened down
[08:20] <stub> Launchpad will be going down in 15 minutes for its regular code update. Estimated downtime is 25 minutes.
[08:32] <SteveA> hi
[08:39] <jamesh> hi SteveA 
[08:46] <BjornT> hi SteveA. do you know when kiko plan to be around?
[08:49] <SteveA> jamesh: morning
[08:50] <SteveA> BjornT: kiko will be around soon.  he'll be busy in meetings, though
[08:51] <BjornT> ok, thanks
[09:04] <SteveA> BjornT: kiko says "what?"
[09:09] <BjornT> SteveA: lifeless put my bug tags branch in kiko's queue. i want to know if he will be able to review it in time so i can land it this week, or if he's too busy so i should have someone else review it.
[09:13] <lifeless> kiko has a note in his queue saying to nag him on IRC when a review is given to him
[09:16] <carlos> morning
[09:31] <stub> carlos: Any idea why you have 3 times more karma than anyone else in Launchpad?
[09:31] <carlos> stub: because I uploaded OO.org translations for breezy manually and I got credit for that (broken .po files)
[09:31] <carlos> stub: it will disappear over time
[09:32] <carlos> stub: you can also nuke all translations karma I have
[09:32] <stub> Will this keep happening?
[09:32] <carlos> I don't translate too much
[09:32] <carlos> it shouldn't
[09:32] <carlos> I got it two months ago
[09:33] <stub> So if you upload translations now, you won't get karma for them? I imagine it is something you and other rosetta-admins will need to do from time to time.
[09:33] <carlos> well, if we do it manually, yes, we will get more karma
[09:33] <carlos> but that's not usual
[09:33] <carlos> stub: anyway, I will add an option to admins to give ownership to rosetta-admins team
[09:33] <jamesh> would it be bad not to assign karma for manual PO file uploads?
[09:33] <carlos> so that karma is not given to anyone
[09:34] <carlos> jamesh: just talking to uploads done by admins
[09:34] <carlos> jamesh: and that only happens if the .po file lacks the metadata with the translator information
[09:34] <stub> It sounds like a way of gaming the system for normal users too
[09:35] <stub> hmm...
[09:35] <carlos> stub: why?
[09:35] <jamesh> carlos: I mean, would it be bad to only dish out karma for translations made through the web interface
[09:35] <carlos> if they have rights to upload such file
[09:35] <carlos> they will get karma for the new translations, nothing more
[09:36] <stub> so you would need to upload randomly generated translations, which would grant karma and poison the database at the same time.
[09:36] <carlos> jamesh: that will not happen, if a normal user uploads a .po file, we are giving karma 
[09:36] <carlos> stub: sure, but that's an abuse of the system and you are able to do it using the web interface
[09:36] <stub> (although in that extreme case, we are going to have to do manual cleaning anyway)
[09:36] <carlos> stub: it's more difficult, but it's possible
[09:36] <jamesh> carlos: I know that's what happens now.  I'm asking whether it would be bad to change it so you only get karma for translations done through the web interface
[09:36] <jamesh> i.e. would it "disenfranchise" any LP users? :)
[09:37] <jamesh> (not that karma gets you anything ...)
[09:37] <carlos> jamesh: the only change would be that the admins will get a 'give ownership of this upload to the rosetta-admins team' option
[09:37] <mpt_> bradb!
[09:37] <carlos> and thus, if the .po file metadata is broken, no one will get that karma
[09:37] <carlos> normal users will get karma as usual
[09:40] <jamesh> carlos: sure.  But would there be any downsides to normal users not getting karma for manual PO file uploads?
[09:41] <carlos> no, they will not see any change
[09:41] <carlos> well, there will be such change
[09:41] <carlos> because I will not earn so huge amount of karma ;-)
[09:43] <jamesh> carlos: ignoring your proposed change, suppose Launchpad was changed to not dish out any karma for manual PO file uploads (for everyone)
[09:44] <jamesh> would there be a problem with that?
[09:44] <carlos> jamesh: yes, people doing offline translations will not get karma for their contributions
[09:45] <lifeless> carlos: what % of translations is that ?
[09:46] <jamesh> carlos: okay.  So there is two classes of manual PO file uploads: people who are importing translations from upstream, and people who've downloaded the PO file from LP, done some translations locally and uploaded the result back to LP
[09:46] <jamesh> is the first case done by more than just rosetta admins?
[09:46] <carlos> jamesh: for Ubuntu, yes, for products, no
[09:47] <carlos> jamesh: anyway, for upstream imports we don't give karma for the translations, just for the fact that they did the upstream import giving us more up to date information
[09:48] <carlos> lifeless: I don't keep track of that, but I'm sure it's more than the 20% due some UI limitations we are fixing over time
[10:04] <carlos> danilos: ping
[10:05] <danilos> carlos: pong 
[10:05] <carlos> danilos: meeting time?
[10:05] <danilos> carlos: indeed ;)
[10:06] <SteveA> BjornT, lifeless: kiko will be too busy this week.
[10:06] <lifeless> SteveA: are you also too busy ?
[10:06] <SteveA> lifeless: kiko and I will not be available to do reviews this week
[10:07] <lifeless> SteveA: then both of you need your knuckles rapped, you should have updated your PendingReviews status.
[10:07] <lifeless> I will do so for you now.
[10:07] <jamesh> SteveA: the zope specs are loaded up at https://demo.launchpad.net/products/zope/+specs
[10:09] <SteveA> thanks jamesh 
[10:09] <lifeless> jamesh: I've swapped malcc's branch you had with bjorn
[10:09] <lifeless> jamesh: and given you bjornts
[10:10] <jamesh> lifeless: okay.
[10:18] <Kamion> Hi - does RF 3796 fix the way bug comments are displayed backwards, i.e. newest-first?
[10:18] <Kamion> Or was that a recent intentional change?
[10:24] <BjornT> Kamion: i didn't know that the comments were displayed backwards, now i see that they are. it looks ok on staging, though, so some patch has fixed it.
[10:24] <lifeless> I like it ;)
[10:25] <LarstiQ> ah, now I see it too
[10:25] <LarstiQ> lifeless: there is something about breaking the natural flow of conversation :P
[10:27] <Kamion> lifeless: it's debbugs circa 1994
[10:27] <Kamion> there's a reason even debbugs thought that was poor UI :-P
[10:27] <Kamion> BjornT: good, thanks
[10:28] <mpt_> Kamion, or Malone circa 2005 :-)
[10:28] <Kamion> oh yes I'd forgotten it used to do that
[10:29] <jamesh> Kamion: got an example of a backwards bug?
[10:30] <jamesh> Kamion: there was a branch merged that broke comment ordering which was fixed yesterday
[10:30] <lifeless> needs a prod cherry pick
[10:30] <lifeless> https://launchpad.net/distros/ubuntu/+source/evolution/+bug/9870
[10:30] <Ubugtu> Malone bug 9870 in evolution "Better spam filtering for evolution" [High,Confirmed]  
[10:30] <Kamion> jamesh: 47848 is the one I'm on at the moment
[10:30] <jamesh> Kamion: could be that today's rollout doesn't include that patch, so needs a cherrypick
[10:30] <lifeless> is an example
[10:30] <Spads> anchors are a good way to implement message pointers to keep track of old vs new if you really want to skip to unread stuff
[10:30] <Spads> and they work in just about everything
[10:30] <lifeless> jamesh: does the fix include tests ?
[10:31] <jamesh> lifeless: it included some tests, but I didn't do a full review of it.
[10:34] <jamesh> lifeless: the problem was that the tests were passing before because they were relying on the natural result set ordering for a query.  The branch fixing the problem added the correct order by clause and reordered a bit of the sample data, but it is a bit difficult to tell if it would catch similar issues
[10:34] <lifeless> hmm
[10:36] <jamesh> I only did a quick read through the patch kiko posted and pointed out one bug that stood out to me
[10:36] <lifeless> If we had a normal object model, I would be suggesting layered tests
[10:37] <jamesh> I wonder if you could get postgres to randomise the ordering of a result set before sorting it ...
[10:37] <lifeless> heh, I was just wondering that
[10:40] <spiv> Can we somehow just append an "ORDER BY random()" to every query? :)
[10:40] <lifeless> sqlobject to the rescue
[10:41] <spiv> And there really is a random() in postgres too!
[10:42] <jamesh> I wonder how many of our tests would fail if the test suite did that
[10:42] <jamesh> spiv: and it works in an order by clause!
[10:42] <spiv> Yeah, I think we could stick that in SQLObject...
[10:42] <spiv> jamesh: excellent!
[10:42] <spiv> jamesh: Let's stop wondering, and find out! :)
[10:42] <jamesh> I don't think we'd want it in production use -- just for the test runner
[10:42] <spiv> Right.
[10:48] <jamesh> spiv: so "select id from table where id < 20 order by id % 2, random();" gives the expected results (even IDs in random order, followed by odd IDs in random order)
[10:48] <jamesh> and different results each time
[10:49] <spiv> jamesh: sounds good
[10:49] <spiv> So it should just be a matter of finding the right spot in SQLObject to add that to the ORDER BY clause.
[10:51] <spiv> Probably a "self.ops['dbOrderBy']  = list(self.ops['dbOrderBy'] ) + ['random()'] " in SelectResults.__init__
[10:52] <spiv> Which ought to be sane even after calling results.clone(), because ops['dbOrderBy']  is always reset.
[10:52] <lifeless> rs=lifeless for a test-only patch to do that :)
[10:53] <mpt_> random-order comments would be nifty
[10:54] <spiv> Well, I'm off to dinner right now, but I'll look at it later if no-one else gets there first.
[10:59] <sivang> morning
[11:18] <lifeless> stub: dude, that sucks
[11:18] <lifeless> Halve karma given to Canonical staff
[11:19] <stub> lifeless: I'm sure Mark would be happy for you to trade your paycheck for karma ;) Hell, you can trade your paycheck to me directly for more karma ;)
[11:19] <lifeless> lol
[11:19] <lifeless> whats the reason though ?
[11:19] <lifeless> (and for that matter, I can update the db for more karma :))
[11:20] <stub> Karma's primary goal is to promote community activity. When 4 our of 5 of the top 5 karma earners on the Launchpad front page are Canonical employees, it doesn't really promote that.
[11:20] <lifeless> another way to address that would be to not show staff in the top 5 report
[11:20] <stub> Although we need a special scaling factor for Carlos too since he is a walking karma bug magnet
[11:21] <stub> lifeless: This is a hack to do just that
[11:21] <stub> lifeless: When we have private teams, we can do it that way.
[11:24] <stub> Although I guess joining with the EmailAddress table wouldn't be too bad... I think this approach is better though ;) Karma is an odd term for something you are being paid to earn ;)
[11:24] <jamesh> stub: can we also cut the karma of people we don't like too? :)
[11:25] <stub> jamesh: I already do
[11:25] <lifeless> stub: well, dunno about you, but most of the lp activity I do is not paid work
[11:25] <LarstiQ> ah, so _that_ is why my karma is dropping so rapidly!
[11:27] <jamesh> LarstiQ: nah.  It's because youre a member of a fake language team :)
[11:28] <stub> Your lucky - Carlos lost over 700,000 points today ;)
[11:28] <carlos> stub: remove all my karma related with translations and I will warn you if I need to do any fix by hand
[11:28] <carlos> so you don't need to scale anything due my karma
[11:29] <stub> carlos: Just joking. It always seems to be your account we notice Karma bugs on ;)
[11:29] <carlos> stub: not really
[11:29] <carlos> rosetta-admins was the first giving problems ;-)
[11:29] <stub> Ahh... true ;)
[11:33] <stub> Anyway - it is just an experiment. Easily reversible and tunable.
[11:38] <stub> So what is the patch that needs cherry picking?
[11:41] <stub> r3796? (Not many to choose from - we are dangerously close to running HEAD as usual :-/ )
[11:43] <jamesh> yeah
[11:43] <jamesh> that's the one
[11:46] <carlos> stub: hmm I guess you didn't run again the karma update script, right?
[11:46] <stub> carlos: nope
[11:46] <carlos> I'm still on the top list of contributors
[11:46] <carlos> ok
[12:03] <SteveA> stub: ping
[01:07] <jordi> carlos: hey dude
[01:09] <cprov> morning
[01:10] <cprov> stub: ping
[01:19] <carlos> jordi: hi
[01:21] <stub> cprov: pong
[01:22] <cprov> stub: did you run the deletions for soyuz ?
[01:22] <stub> cprov: Not yet. I'll do that right now.
[01:22] <cprov> stub: perfect, thanks 
[01:24] <stub> cprov: Done.
[01:25] <cprov> stub: less oops, thank you 
[01:34] <stub> cprov, Kinnison, malcc: https://chinstrap.ubuntu.com/~dsilvers/paste/filebR8O0P.html
[01:34] <stub> Does that seem reasonable (as discussed with cprov on the mailing list)?
[01:36] <cprov> stub: looks fine, you can land it, although they doesn't protect us from the last bug, as I said
[01:37] <cprov> stub: but they are, indeed,  necessary to keep those join-tables sane.
[01:40] <Kinnison> I'm not so convinced about the new constraints
[01:40] <Kinnison> In particular, if the queue is to be used to model packages getting into derivatives (which makes sense) then we can't have the constraint
[01:41] <stub> Can we keep it until then?
[01:42] <cprov> Kinnison: we will need a new DRQ for derivative, the constraints just organize DRQB/DRQS ...
[01:42] <Kinnison> stub: sure, constraints can always be relaxed
[01:57] <ddaa> jamesh: replace symlink by file is a baz bug
[01:58] <jamesh> ddaa: hmm?
[01:58] <ddaa> "A test for symlink -> file replace and vice versa.
[01:58] <ddaa> Doesn't currently work, maybe due to a pybaz bug."
[01:58] <ddaa> on cscvs/svn-symlinks
[01:59] <jamesh> ddaa: no.  this was a pybaz bug -- I fixed it and landed the change in the copy of pybaz on chinstrap
[01:59] <ddaa> oh?
[02:00] <jamesh> ddaa: see r181 of pybaz in rocketfuel
[02:00] <jamesh> pybaz's WorkingTree.delete() did not function correctly for symlinks
[02:01] <ddaa> oh, right
[02:02] <ddaa> it should not be used anyway, because it does the wrong thing for directories
[02:02] <ddaa> well, it does the wrong thing for cscvs, at least
[02:02] <carlos> later
[02:03] <ddaa> because in cscvs we want to avoid deleting a directory that contains versioned files, to help ensure that the import is consistent
[02:03] <ddaa> jamesh: thanks for fixing that
[02:05] <jamesh> ddaa: no problem.  It took me a while to track the problem down to pybaz, since the function would return without error if the given file was neither a regular file or directory
[02:07] <ddaa> jamesh: apologies, I wrote that code in 2003, I didn't have a clue.
[02:08] <jamesh> ddaa: no problem.
[02:09] <ddaa> jamesh: btw, the revno you gave to stub earlier, that's the fix for the branch scanner?
[02:10] <jamesh> ddaa: earlier today?  No, that was to fix comment ordering in Malone
[02:11] <jamesh> ddaa: I gave him the revno for the branch scanner fix last week, but I don't think he got round to cherry picking it.  It should have gone out with today's rollout though
[02:11] <ddaa> okay, I'll be looking out for cron emails
[02:12] <jamesh> ddaa: the branch puller parts are definitely working: the last_mirrored_id is set for many branch rows.
[02:13] <jamesh> if last_scanned_id starts getting set then it is definitely working
[02:14] <ddaa> jamesh: I got the ForbiddenAttribute errors from launchpad-error-reports
[02:14] <ddaa> if it stops crashing and burning, I'll know it
[02:14] <jamesh> ddaa: okay.  It'd be worth checking if the code base used by the branch scanner has been updated then.
[02:15] <ddaa> stub usually does it :)
[02:15] <jamesh> when fixing things I tweaked the way the branch scanner test suites were run to trigger the ForbiddenAttribute error and then fixed it
[02:16] <ddaa> yeah, I know that the tests running in lib/importd did not have a proper zope testing environment
[02:16] <stub> I  updated the branch scanner installation with everything else.
[02:16] <ddaa> the branch scanner code being there is a combination of "historical reasons" and "it looked like a good idea at the time" (see bzrscan.py, that you deleted)
[02:19] <ddaa> stub: thank you
[02:19] <ddaa> stub: can I make you braids?
[02:21] <jamesh> ddaa: so you are sure the ForbiddenAttribute errors you got weren't for a run prior to the rollout today?
[02:22] <ddaa> jamesh: yup, the run was between 0:30 and 8:00 BST
[02:23] <jamesh> ddaa: that sounds like before the rollout to me -- 0:30 would be 6:30am for stub, wouldn't it?
[02:24] <ddaa> something like that
[02:24] <ddaa> hu
[02:25] <ddaa> okay I read you badly
[02:25] <ddaa> I'm sure the error I got was for a run prior to the rollout today.
[02:25] <ddaa> need... lunch...
[02:30] <ddaa> I did not mean to say that it was still broken, I'll just wanted to know what was the status, and to say that I'll keep an eye out for it since it was just rolled out.
[04:21] <carlos> spiv: hi, around?
[04:21] <spiv> carlos: yep
[04:21] <carlos> spiv: I'm having problems with a test
[04:22] <carlos> it's supposed to fail, but it doesn't
[04:22] <carlos> and I think it's related to cached values
[04:22] <carlos> and flush or commit is not fixing it
[04:23] <carlos> hmm, wait, last try shows me something even weird, the change is there but my test is not detecting it.... 
[04:23] <carlos> spiv: sorry, give me 10 minutes to check something and I will ping you again...
[04:24] <spiv> carlos: Sure :)
[04:27] <carlos> spiv: btw, aren't you supposed to be sleeping?
[04:28] <elmo> stub: ping
[04:28] <spiv> carlos: very soon :)
[04:28] <jamesh> spiv: I did up a basic patch for the sqlobject result order randomisation: shows up lots of bugs
[04:29] <spiv> jamesh: wonderful! :)
[04:29] <spiv> Ah, I see you already added it as a w-i-p branch.
[04:30] <elmo> asuka/staging is going down for a HW upgrade
[04:32] <carlos> spiv: ok, seems like the problem is with the diff operation I do, for some reason it doesn't detect the differences... or my code ignores such detection...
[04:32] <jamesh> spiv: the "make check" output was 10962 lines long ...
[04:34] <spiv> jamesh: Whee!
[04:35] <spiv> jamesh: If you try just a small subset of the test suite, say just one pagetest story, is there some obvious culprits causing the bulk of the problems?
[04:35] <spiv> jamesh: Or do you think we'll be fixing this for months to come? ;)
[04:37] <jamesh> spiv: there are some cases of one error causing followon errors
[04:38] <jamesh> ah.  I broke count queries
[04:38] <spiv> Ah.
[04:39] <jamesh> count doesn't like orderBy
[04:39] <spiv> Yep.
[04:47] <jamesh> actually, it is a problem with set ops
[04:47] <jamesh> just happened to be a count query I was looking at
[04:49] <spiv> Ah, ok.
[04:49] <jamesh> ERROR:  ORDER BY on a UNION/INTERSECT/EXCEPT result must be on one of the result columns
[04:50] <spiv> Yeah, probably simplest to not try adjust those for now.
[04:51] <jamesh> was getting a syntax error originally.  I added parentheses which changed the error to the above, which is complaining about the outer "order by" clause
[05:10] <Lord_Athur> hi all
[05:11] <carlos> hi
[05:13] <Lord_Athur> I was trying to validate my OpenPGP key, the launchpad page says that I'd have to descrypt a msn with my key in order to validate, how do I do it?
[05:18] <carlos> Lord_Athur: launchpad sent you an email with such message
[05:18] <carlos> Lord_Athur: your email client should do it if you have it configured to use your gpg key
[05:19] <Lord_Athur> :O
[05:20] <Lord_Athur> it didn't do it
[05:20] <jamesh> Lord_Athur: try saving the PGP data block (including the "--- BEGIN PGP ..." lines) to a file
[05:20] <jamesh> Lord_Athur: then run "gpg --decrypt filename"
[05:21] <Lord_Athur> thanks, i'll do it
[05:21] <carlos> jamesh: aren't we sending that by email?
[05:21] <jamesh> carlos: sending what?
[05:22] <jamesh> carlos: sounds like Lord_Athur's email client doesn't have PGP integration, so I was suggesting a way to decrypt the data
[05:22] <Lord_Athur> I did it, thanks all
[05:23] <jamesh> carlos: iirc there is a bug about providing some non-encrypted data in the encrypted validation emails, but it hasn't been implemented
[05:24] <carlos> jamesh: oh, I misunderstood your suggestion and thought that you were talking about something we were rendering in our website
[05:25] <carlos> I see
[05:49] <carlos> danilos: ping
[05:50] <carlos> salgado: ping
[05:50] <salgado> carlos, pong
[05:51] <carlos> salgado: do you have 5 minutes to talk about your review of danilos' branch ?
[05:51] <salgado> carlos, sure
[05:51] <carlos> salgado: danilos told me that you said that we shouldn't use initialize for things like process a form submission
[05:52] <carlos> but, when we switched to current notification system, I was told (I think by Stuart) that in fact, that's the way to do it
[05:52] <carlos> instead of adding the form submission call from the pagetemplate
[05:52] <carlos> so I would want to know if your comment is our current policy
[05:53] <carlos> because I'm using the initialize every time I need to process a form
[05:54] <carlos> the idea behind adding it to initialize was to execute the form submission after the notification message init
[05:54] <salgado> well, I think it's clearer to have only generic initalization stuff in initialize() and have a separate method to process the form submission
[05:55] <salgado> and I've never seen the initialize method used to process the submission
[05:55] <carlos> well, I'm talking about calling a method that process the form submission from initialize
[05:55] <carlos> instead of adding it directly to the page template
[05:55] <mpt__> http://www.skweezer.net/skweeze.aspx?q=https%3A%2F%2Flaunchpad.net%2F&i=1
[05:55] <carlos> so maybe I misunderstood the explanation from danilo ;-)
[05:55] <mpt__> awesome
[05:57] <salgado> carlos, even that I haven't seen and wouldn't do myself
[05:57] <carlos> salgado: https://chinstrap.ubuntu.com/~dsilvers/paste/fileUnhM0G.html
[05:57] <salgado> yeah, I understood what you mean
[05:58] <carlos> so you think we should put it in our page template?
[05:58] <carlos> I don't mind if it works ;-)
[05:59] <carlos> but I just want to know if we could define an standard way to do it so we don't mix different ways to do the same driving crazy to anyone following our code
[05:59] <salgado> another reason for not having it on the initialize() method is that you may have multiple templates sharing a single view, but using different methods on that view to process the submission
[06:01] <salgado> so, I wouldn't like to have form-submission code run directly or indirectly from initialize(), but if we choose that as our policy, I'd have to accept
[06:01] <carlos> salgado: https://chinstrap.ubuntu.com/~dsilvers/paste/filevMJHph.html <-- That's what SteveA suggested me to use
[06:02] <carlos> salgado: I'm not saying it's our policy, I'm just asking for a standard way to do it no matters if it's the way I'm doing it atm or the way you suggested to danilos
[06:02] <carlos> but just a way to do it
[06:02] <salgado> yeah, I'm saying that I don't know what our policy is, or even if we have one
[06:02] <kalosaurusrex> salgado: hey so I submitted a request to have my db cleared of support stuff. how do I make sure that doesn't happen now?
[06:03] <kalosaurusrex> if it hasn't happened do I assume it just won't anyway? no one emailed me back or anything /shrug
[06:04] <salgado> carlos, but initialize()s docstring doesn't mention anything about using it to process form submission --it only mentions initialization
[06:05] <carlos> mpt__: so that's a tool that gets a webpage and 'fixes' it to be used from a mobile device?
[06:05] <carlos> salgado: ok, I will note that in my email. Thanks
[06:06] <salgado> kalosaurusrex, this is usually not a trivial task, so it may take some time. where did you submit the request?
[06:06] <BjornT> salgado, carlos: fwiw, i'd consider form processing being part of the initialization process. by processing the form you set the widgets in their desired states before rendering the page. calling a method first thing in a page template is kind of a hack.
[06:09] <carlos> BjornT:  just a comment, if you execute it as the first thing in a page, the notification system will not work, so I don't think salgado suggest to put it as the first line (or at least it was not working when we implemented it)
[06:09] <salgado> carlos, if you put it first thing in a page it does works
[06:09] <salgado> (the notifications, I mean)
[06:10] <carlos> then it's fixed now, it was not working when stuart implemented it
[06:11] <salgado> by first thing in a page I mean before the <html> line... I thought that always worked
[06:16] <flacoste> i have the following error:     *  Module sqlobject.dbconnection, line 446, in accumulateSelect
[06:16] <flacoste>       assert select.ops.get('selectAlso') is None, (
[06:16] <flacoste> AssertionError: selectAlso isn't implemented with accumulations like SUM and COUNT
[06:16] <flacoste> any idea of what I should do to work around that?
[06:16] <kalosaurusrex> salgado: I emailed the launchpad list.
[06:17] <salgado> flacoste, are you trying to do a sum() or count() on the result of a union/intersect/except?
[06:17] <flacoste> indirectly
[06:17] <flacoste> salgado: no, not a set operation, i'm using selectAlso to have the rank() of the full text search
[06:18] <flacoste> salgado: to sort by relevancy
[06:18] <flacoste> salgado: i'm not calling count on the results set though, BatchNavigator is
[06:18] <salgado> ahh, right. that's selectAlso. I misread it
[06:24] <carlos> salgado: no, it didn't. In fact, I think danilo had problems doing it that way and then I had to tell him the initialize thing...
[06:25] <salgado> carlos, the problem on that template existed because the call to the process_form() method was inside the <div fill-slot="body">
[06:25] <danilos> carlos: well, we haven't tried adding another <tal:do-this-first> in the template
[06:25] <danilos> carlos: and... (salgado already got into those details... ;)
[06:26] <carlos> salgado: I see
[06:27] <danilos> carlos: it was simply *after* the notification rendering stuff, so it was enough to move it in front, and salgado suggested moving it all the way to the top
[07:34] <flacoste> salgado: regarding the AssertionError related to selectAlso, after looking at the problematic method I removed the assertion and things work fine
[07:35] <flacoste> flacoste: should I file a bug about that? who can modify the Launchpad SQLObject version for that?
[07:35] <salgado> flacoste, wasn't that assert in an sqlobject method? wasn't there any comment explaining the assert?
[07:36] <flacoste> salgado: the only comment is the assertion message: selectAlso isn't implemented with accumulations like SUM and COUNT
[07:36] <salgado> flacoste, I don't remember who can commit to our sqlobject branch, but spiv can, for sure
[07:38] <salgado> maybe somebody implemented it later and didn't remove the assert
[07:38] <flacoste> there is nothing to implement
[07:38] <salgado> or maybe it's still not implemented and some things will break. :/
[07:39] <flacoste> selectAlso adds other things in the selection list
[07:39] <flacoste> that part isn't used in the count() or sum() query
[07:40] <salgado> well, then I guess it should be fine to remove the assert
[07:40] <flacoste> and as far as I can remember, the WHERE clause cannot reference these extra fields
[07:41] <flacoste> so, I don't see how it could break
[07:41] <flacoste> i'll file a bug and mail the list
[07:42] <flacoste> plus, selectAlso is a canonical addition and it's currently used only for the rank() function
[07:43] <salgado> yeah, it was added by Mark
[07:58] <DSG-01> I'm away