[01:02] <cjohnston> wgrant: can you look at what I have and explain to me what your wanting things to look like please?
[01:02] <cjohnston> with all your changes, I'm not sure that I know
[01:06] <wgrant> cjohnston: My branch doesn't impact publish-ic-link at all
[01:06] <wgrant> I only changes stuff inside the diff table.
[01:06] <cjohnston> correct...
[01:06] <cjohnston> I'm talking about your branch
[01:07] <wgrant> Oops, missed the fact that the MP you linked was more than two hours ago, not relevant to this convo.
[01:07] <wgrant> Looking
[01:08] <wgrant> cjohnston: What do you have?
[01:08] <wgrant> I don't see a branch.
[01:08] <cjohnston> wgrant: I have it up (plus one small change) at 162.213.35.40
[01:08] <cjohnston> https://code.launchpad.dev/~name16/myproject/team/+merge/1
[01:10] <wgrant> cjohnston: The main issue I see is the width.
[01:10] <wgrant> cprov's comment blows things up, and opening a draft comment (at least in Firefox) has ridiculous padding on the right.
[01:11] <cprov> what ? who ? me ?
[01:11] <wgrant> Oh
[01:11] <cjohnston> yes, you
[01:11] <wgrant> cprov's comment is actually fine, until the draft super-padding appears.
[01:11] <cjohnston> wgrant: so, the IC's, should they be 100% of the table.diff width?
[01:14] <wgrant> cjohnston: I deliberately added some horizontal margin to make the comments not just look like another line of the diff.
[01:14] <wgrant> My intent (feel free to do something less terrible) was to have them always fit inside the normal diff width
[01:15] <wgrant> That all works well until you invoke a draft editor
[01:15] <wgrant> The editor isn't meant to change the width at all.
[01:15] <cjohnston> ok.. so you like the width as it is..
[01:15] <cjohnston> (except the edit)
[01:15] <wgrant> It's marginally the least bad so far, IMO.
[01:16] <wgrant> But I'm very much open to suggestions.
[01:16] <wgrant> And it's easy to change now we're not tied to the table-based layout.
[01:16] <cjohnston> personally, I liked it the way it was.. but as we aren't designers, I don't care to play for hours on this
[01:17] <wgrant> It was really hard to see the comments in the old one.
[01:21] <cjohnston> I want a new icon for adding a comment
[01:21] <cjohnston> I don't like that one
[01:21] <cjohnston> (not your fault, just my opinion)
[01:28] <wgrant> Oh yeah, that's by no means even vaguely final, I just needed one to throw in the CSS for experimentation.
[01:28] <wgrant> And I remembered the name of that one :)
[01:29] <cjohnston> lol
[01:36] <cprov> wgrant: can you check qastaging mbox ? there should be a MP email with non-ascii diff content.
[01:37] <wgrant> cprov: There are errors, rather
[01:37] <wgrant> See #launchpad-ops; DB perms not there yet.
[01:37] <cprov> putz
[01:41] <wgrant> cprov: Fixed now, can you try again?
[01:41] <wgrant> I just ran an ASCII one through, and the email worked fine.
[01:42] <wgrant> brb
[01:43] <cprov> done
[02:02] <wgrant> > +Now let's mess with encoding áéíóúçã
[02:02] <wgrant> >
[02:02] <wgrant> zoing!
[02:02] <wgrant> >
[02:03] <wgrant> cprov: Looks good
[02:03] <cprov> wgrant: great!
[13:44] <cjohnston> wgrant: any chance you can look at adding a draft comment on my instance?
[13:49] <wgrant> cjohnston: !
[13:49] <wgrant> It looks like not totally terrible
[13:49] <wgrant> Well done
[13:50] <cjohnston> I just s/edit/add for the button, if you refresh you should be able to see that
[13:50] <cjohnston> I think that's slightly better than edit
[13:50] <wgrant> Indeed
[13:51] <cjohnston> what other changes are needed for this branch?
[13:57] <wgrant> cjohnston: We should monospace and fix the word breaking in the comments if possible.
[13:58] <wgrant> cjohnston: word-wrap: break-word; works if the div is width-constrained. Somehow convincing it to not extend outside the table would work.
[14:23] <daker> yo
[14:24] <cjohnston> hey daker
[14:25] <daker> wgrant: hi
[14:27] <daker> technically you can't set the width to px since launchpad is using a fluid layout (% & auto so no fixed width)
[14:27] <daker> i those cases, i think we can use "calc"
[14:27] <daker> in*
[14:28] <daker> calc(100% - Xpx)
[14:28] <daker> width : calc(100% - Xpx);
[14:34] <wgrant> daker, cjohnston: Or what about setting table-layout: fixed on table.diff, and then a fixed px or % width on table.diff .line-no?
[14:34] <wgrant> That'll prevent the comment divs from causing the table to expand.
[14:35] <daker> cjohnston: can you test that ?
[14:35] <wgrant> Throw in a font-family: monospace and a word-wrap: break-word on the comment bodies, and it starts to look at bit more sane.
[14:35] <wgrant> The only real problem with that approach is that the line-no column is no longer dynamically sized, but that's a bit weird anyway.
[14:38] <cjohnston> wgrant: how's that look?
[14:45] <wgrant> cjohnston: Not bad! The line-no needs to be able to cope with up to four digits plus space for the icon, but that's a trivial tweak.
[14:47] <cjohnston> do you know how big that would need to be?
[14:48] <cjohnston> and what happens when people give you 10000 + line MPs? ;-)
[14:48] <wgrant> I think we trim it after like 10000 lines
[15:13] <cjohnston> wgrant: would the problems from last night effect branch scanning at all?
[15:14] <wgrant> cjohnston: At the moment? No.
[15:14] <cjohnston> just LP being slow
[16:03] <wgrant> cjohnston: !! it finally scanned
[16:03] <cjohnston> yay!
[16:16] <cjohnston> wgrant: just pushed a new rev.. now the big question... will it scan. lol
[16:16] <wgrant> The button text doesn't show up in other places. Can you find out why?
[16:16] <wgrant> We shouldn't be hiding it with an outrageous text-indent :)
[16:18] <tasdomas> lbox propose gives me this error: error: Failed to update merge proposal log: Server returned 401 and body: (<BranchMergeProposal at 0x2af0e8f83850>, 'description', 'launchpad.Edit') - no idea, what's it about, though..
[16:18] <wgrant> tasdomas: You don't have permission to edit that merge proposal.
[16:19] <tasdomas> wgrant, you mean the merge proposal on launchpad?
[16:20] <wgrant> tasdomas: Correct.
[16:20] <tasdomas> wgrant, how can that be possible? (or, in other words, what do I do to fix this?)
[16:22] <daker> wgrant: and their is a nasty bug in caused by the help modal
[16:22] <wgrant> tasdomas: Should you have permission to edit that merge proposal? Which is it?
[16:23] <tasdomas> wgrant, https://code.launchpad.net/~tasdomas/juju-core/ensure-availability-more-output/+merge/220306
[16:24] <wgrant>         return (user.inTeam(self.obj.registrant) or
[16:24] <wgrant>                 user.inTeam(self.obj.source_branch.owner) or
[16:24] <wgrant>                 self.forwardCheckAuthenticated(
[16:24] <wgrant>                     user, self.obj.target_branch) or
[16:24] <wgrant>                 user.inTeam(self.obj.target_branch.reviewer))
[16:24] <wgrant> Are you sure your client was authenticated as the correct user?
[16:30] <tasdomas> wgrant, sorry to have bothered you - it was my own mistake
[16:31] <tasdomas> but your replies pointed me in the right direction
[16:32] <wgrant> tasdomas: What was the issue?
[16:32] <wgrant> Wrong MP URL, or wrong creds?
[16:32] <tasdomas> wrong creds
[16:32] <wgrant> Ah, great :)
[16:36] <wgrant> cjohnston, cprov: I'll be back in a few hours, please try to poke holes in inline comments.
[16:36] <wgrant> Night
[17:07] <cjwatson> Is there some documentation somewhere of how to upgrade local development Launchpad instances to PostgreSQL 9.3, so that we can stay in sync with production?
[17:41] <stub> cjwatson: The production docs would walk you through it, but I suspect the pg_upgradecluster provided by the Debian package would work a treat.
[17:43] <stub> cjwatson: https://pastebin.canonical.com/110495/ is the current draft for the production upgrades
[17:44] <cjwatson> Thanks.  Will have a try tomorrow, maybe ...
[19:31] <cjohnston> cprov: r17011 qa-ok
[23:53] <wgrant> cjwatson: You can also just pg_dropcluster 9.1 main, install launchpad-database-dependencies-9.3, purge postgresql-9.1, and then rerun utilities/launchpad-database-setup and make schema