[07:14] <mlankhorst> can someone bump the repository size of https://launchpad.net/~canonical-x/+archive/x-staging ?
[07:48] <wgrant> mlankhorst: It's usually best to ask at https://answers.launchpad.net/launchpad. But done.
[07:48] <mlankhorst> thanks
[09:56] <jml> I'm getting an ssl cert error when I try to fetch a branch from Launchpad.
[09:58] <lifeless> is bzr using ssh or http[s] ?
[10:03] <wgrant> All our SSL certs are good. Do you have ca-certificates installed, and how ancient is your bzr?
[10:03] <wgrant> Modern bzrs don't use XML-RPC for HTTP at all, do they?
[10:14] <maxb> I helped someone recently whose ValiCert root cert had disappeared and in the process discovered that Ubuntu One had turned my own into broken symlinks
[10:14] <maxb> It could be something like that
[10:15] <maxb> jml: Try a 'ls -l /etc/ssl/certs/ | grep -Fi valicert' and pastebin?
[10:16] <jml> wgrant: I'm using whatever's on precise
[10:18] <jml> maxb: no broken symlinks then.
[10:19] <maxb> jml: I'd still like to see the paste to check for missing ones
[10:20] <maxb> I'd tell you what to look for but I'm ty[ing from an Android tablet so don't have a reference easily to hand
[10:22] <jml> maxb: https://gist.github.com/jml/6435231
[10:24] <maxb> jml: Yup, you're missing a hex-hash symlink to the valicert class2 root
[10:25] <maxb> First, run 'debsums -s ca-certificates' to check if some files are missing
[10:26] <maxb> The .dpkg-new file being in /etc/ssl/certs is rather bogus too
[10:27] <maxb> I would not expect dpkg to be installing into that directory at all (the files go into /usr/share)
[10:27] <jml> maxb: a file is missing.
[10:27] <maxb> 'apt-get --reinstall install ca-certificates' is probably what you want, then
[10:28] <maxb> after which, 'update-ca-certificates' ought to re-make the missing symlinks
[10:46] <jml> maxb: thanks.
[10:51] <jml> now I'm getting timeouts branching.
[10:52] <wgrant> Timeouts?
[10:52] <wgrant> Branches are served over HTTP directly by Apache
[10:52] <wgrant> Unless the XML-RPC call at the start is failing, they cannot time out except by a network issue.
[10:53] <jml> hmm.
[10:53]  * jml bumps up the logging
[10:53] <wgrant> How did the timeout appear?
[10:54] <jml> https://gist.github.com/jml/6435480
[10:54] <wgrant> Ah, ssh
[10:54] <jml> yeah.
[10:55] <jml> (I'm planning on pushing shortly after I branch.)
[10:55] <wgrant> I think that's recursively stacked...
[10:55] <wgrant> Oh no
[10:55] <wgrant> Just got opened a few times
[10:55] <wgrant> It branched fine from here
[10:55] <wgrant> Are you sure your SSH connectivity to Launchpad is OK?
[10:56] <wgrant> What does -Dhpss think about the whole thing?
[10:57] <wgrant> Though that open_2.1 sounds like the first call.
[10:57] <jml> https://gist.github.com/jml/6435514
[10:57] <jml> yeah, it is.
[10:58] <jml> trying to ssh into bazaar.lp
[10:58] <jml> (seriously, who wrote this crap?)
[10:58] <wgrant> Oh, you're on a Google network?
[10:59] <jml> wgrant: yes. how could you tell?
[11:00] <wgrant> The $HOME that is wrong on quite a number of levels.
[11:00] <jml> ah yes.
[11:00] <jml> you'd be surprised :)
[11:01] <jml> hmm. this might be an auth thing.
[11:03] <wgrant> Speaking of Twisted
[11:03] <wgrant> I hate whoever devised its default log rotation.
[11:03] <jml> yeah, sorry about that.
[11:03] <jml> that wasn't me, but it's been a problem for a while.
[11:03] <wgrant> I've exterminated bits of it, but some stuff still uses it
[11:04] <wgrant> For example, bzr-sftp...
[11:04] <jml> I think the idea of twistd is a really good one.
[11:04] <jml> but the execution hasn't been uniformly great
[11:06] <wgrant> Oh yes, certainly.
[11:06] <jml> wgrant: I think it was something to do with my local auth
[11:06] <wgrant> It is lovely in a lot of ways. Just not the log rotation policy...
[11:07] <wgrant> jml: Manual ssh doesn't work either?
[11:07] <jml> wgrant: it didn't. does now.
[11:07] <wgrant> Aha
[11:07] <wgrant> Dodgy agent?
[11:07] <jml> wgrant: I guess I'm surprised the error was timeout.
[11:07] <jml> wgrant: interesting thought.
[11:07] <jml> maybe
[11:07] <jml> I'm running i3wm, so normal desktop sessiony things sometimes don't work as I'd expect.
[11:08] <wgrant> That 'timed out while waiting to read' doesn't seem to have come from bzr
[11:08] <jml> no, that'll be ssh
[11:08] <wgrant> Indeed, that's from OpenSSH itself
[11:10] <jml> wgrant: btw, congrats on the recent buildd work.
[11:11] <wgrant> buildd-manager's currently getting a pretty serious makeover.
[11:11] <jml> about time :)
[11:11] <wgrant> Making the Twisted slightly less insane.
[11:11] <jml> interesting.
[11:12] <wgrant> Disentangling the Twisted and DB code
[11:12] <jml> if I'd done it, I would have clarified the interface to LP st it didn't talk directly to the db
[11:12] <wgrant> That's the eventual plan
[11:12] <jml> cool.
[11:12] <jml> wgrant: could I persuade you to split lp's bugtracker into a separate product?
[11:13] <wgrant> But before I can do that I need the Twisted to run as a layer on top of the model access code, rather than a horrible intertangled web
[11:13] <wgrant> Heh
[11:13] <wgrant> That would be lovely.
[11:13] <jml> wgrant: yeah. I'd really like it for my FLOSS work.
[11:14] <jml> but I'm probably not going to do the work because a) it's work, b) that much work, I almost might as well write my own
[11:15] <wgrant> Indeed, we have one of the better bugtrackers around, I think, and it's a bit unfortunate that it's so heavy to run.
[11:16] <wgrant> And due to lazr.restful it is some challenge to split out the rest.
[11:17] <jml> well, if I went at it, I'd have the liberty of forking
[11:17] <jml> so I'd just clone LP and start deleting the non-bugtracker bits
[11:17] <wgrant> Yeah.
[11:17] <jml> shedding a silent tear that I can't use the trademarked edit icon.
[11:18] <wgrant> But the drunken exclamation mark will feel left out.
[11:18] <wgrant> :'(
[11:18] <jml> wgrant: blame the lawyers.
[11:18] <wgrant> Perhaps we can do a Kickstarter for a freely licensed equivalent of this essential resource.
[11:18] <jml> the thing is, that's still a non-trivial amount of work
[11:19] <cjwatson> Even more if you want to actually keep it up to date
[11:19] <jml> cjwatson: well, yes. I wouldn't bother about keeping it up-to-date with LP development though.
[11:20] <jml> because there's not a great deal going on, and because it'd probably change enough that it'd be an exercise in futility
[11:20] <jml> mostly it'd just be security bits
[11:20] <jml> e.g. I'm not sure that anyone outside of Ubuntu cares for the distro model in LP.
[11:21] <jml> that said, I always underestimate the cost of maintaining software. even undistract-me I have to spend time every week or so reviewing & merging patches.
[11:21] <cjwatson> LP is part of Ubuntu Engineering nowadays, so yeah, a lot of the recent work has been soyuz/buildd kind of area
[11:21] <cjwatson> jml: ironic
[11:21] <jml> & it's tiny
[11:22] <jml> cjwatson: I strive to live a poetic life
[11:23] <jml> cjwatson: makes sense.
[11:23] <jml> also, soyuz was always the area to me that I would have found most fun to improve
[11:24] <wgrant> I tried...
[11:25] <jml> wgrant: heh. I'd rather not get started on my "I should haves" just now. I'll be here all day, and the whisky doesn't get served for another 4 hours.
[11:25] <wgrant> Heh
[11:27] <cjwatson> Isn't that what a hip-flask is for?
[11:29] <jml> I think if I take that particular road to perdition, I'd just spike my coffee
[11:29] <jml> problem is, I'd have to pay for my whisky  then
[11:30] <jml> lunch
[17:29] <cave> hi all
[17:30] <cave> is there a RSS Feed available for Projects on Launchpad?
[17:30] <cave> https://launchpad.net/~csoler-users/+archive/retroshare-snapshots/+packages  <-- i want to read this feed for updates.
 is there a RSS Feed available for Projects on Launchpad?
 https://launchpad.net/~csoler-users/+archive/retroshare-snapshots/+packages  <-- i want to read this feed for updates.
[20:38] <Ampelbein> cave: As far as I see it isn't supported, bug 207333
[21:01] <cave> thanks Ampelbein