[00:01] <wgrant> I don't know of any.
[00:01] <wgrant> Which build?
[00:05] <maxb> https://launchpad.net/~mercurial-ppa/+archive/staging-stable-snapshots/+build/2092164
[00:05] <wgrant> I guess it's probably a new directory rename race.
[00:05] <wgrant> I'll get a LOSA to poke around on Monday.
[00:06] <jelmer> yeah, they should all be gone by now
[00:06]  * jelmer has a look
[00:07] <jelmer> wgrant: oh, right.. you're finally allowed to ping losa's on Monday :-)
[00:09] <wgrant> jelmer: More importantly, I can see logs.
[00:17] <jelmer> wgrant: ah, yeah
[00:20] <jelmer> maxb: I'm inclined to close bug 682430 as Won't Fix
[00:20] <_mup_> Bug #682430: Sourceforge http: (no SSL) imports now fail <code-import> <Bazaar Subversion Plugin:New> <Launchpad Bazaar Integration:Triaged> <https://launchpad.net/bugs/682430>
[00:21] <maxb> jelmer: Seems reasonable. The workaround of https is entirely adequate
[00:23] <maxb> if any fix was to be attempted, it seems like it would be best to fix the svn client libs to allow limiting the use of a single persistent connection, than add various retry logic in places
[00:24] <maxb> Or better, tell sf to stop being silly with their http connections :-)
[00:25] <jelmer> maxb: Yeah, the latter seems the most reasonable.
[00:26] <jelmer> Perhaps they have put this into place to kill long expensive interactions with their SVN server, but that seems silly given this will a) make people retry more often an b) use SSL, which is more expensive CPU-wise.
[00:30] <jelmer> hmm, odd, one of my daily builds just failed with a chroot error
[00:39] <jelmer> maxb: perhaps they're just closing idle connections earlier?
[00:40] <jelmer> maxb: other than what you mentioned in that bug report did you see any particular symptons?
[01:25] <maxb> no, nothing else
[02:36] <lifeless> [02:36] <lifeless>     Hard / Soft  Page ID
[02:36] <lifeless>       54 / 4108  Archive:+index
[02:36] <lifeless>       35 /  139  BugTask:+index
[02:36] <lifeless>       22 /  245  POFile:+translate
[02:36] <lifeless>       16 /    0  Person:EntryResource:retractTeamMembership
[02:36] <lifeless>       14 /  175  Distribution:+bugs
[02:36] <lifeless>        8 /    0  BinaryPackageBuild:+retry
[02:36] <lifeless>        7 /  100  ProjectGroupSet:CollectionResource:#project_groups
[02:36] <lifeless>        7 /   13  DistroSeriesLanguage:+index
[02:36] <lifeless>        5 /   39  Archive:+packages
[02:36] <lifeless>        5 /   16  Distribution:+archivemirrors
[19:18] <pcjc2> Hi, any particular reason I can't see the SSO blueprint specs?
[19:18] <pcjc2> https://wiki.canonical.com/LaunchpadLoginService  -> Not allowed to view this page
[19:19] <pcjc2> https://launchpad.canonical.com/Launchpad-SSO/infrastructurespecs#login -> Not allowed to view this page
[19:19] <lifeless> wiki.canonical.com is staff only.
[19:19] <pcjc2> I was trying to get to grips to how things were working
[19:19] <lifeless> I don't know if those specs /need/ to be private
[19:19] <pcjc2> Have been exporting my local dev instance over SSH to some testers.. One couldn't log on
[19:19] <lifeless> #canonical-isd has some of those developers.
[19:19] <pcjc2> Turned out to be a cookie problem though.. he was using a text based browser, W3M
[19:20] <pcjc2> With a a patch to stop it rejecting some LP cookies, it worked for me when I tried. LP has a related bug: Bug#385702
[19:20] <_mup_> Bug #385702: statistics dont work in replay mode <Return To The Roots:Fix Released> <https://launchpad.net/bugs/385702>
[19:20] <pcjc2> Urm.. no
[19:21] <pcjc2> Bug #59510 (The other was the Debian Bug # for the same issue)
[19:21] <_mup_> Bug #59510: Can't log in with w3m due to bad cookies <Launchpad Foundations:Triaged> <https://launchpad.net/bugs/59510>
[19:22] <lifeless> also bug 523229
[19:22] <_mup_> Bug #523229: The Continue button isn't selectable in w3m for sso login <iso-testing> <Canonical SSO provider:Won't Fix> <Launchpad Foundations:Triaged by gary> <https://launchpad.net/bugs/523229>
[19:26] <pcjc2> I'm amazed just how nice the LP code is
[19:26] <lifeless> why thanks :)
[19:26] <pcjc2> Or rather, how detached from the nasties of writing HTML it is.. and how awesome storm seems
[19:27] <pcjc2> I might even try doing some coding ;) - Any mileage in a drop-down expand for viewing patch attachments to bugs inline with the comments?
[19:28] <lifeless> that would be lovely
[19:29] <lifeless> assuming something sensible xss wise
[19:31] <pcjc2> xss?
[19:31] <lifeless> cross site scripting attacks
[19:32] <pcjc2> ok, will have to get someone to review it if / when I've done it ;)
[19:32] <lifeless> I suggest discussing the broad approach here or on the launchpad-dev mailing list - we
[19:37] <lifeless> we've got some very clueful guys who can help
[19:52] <pcjc2> thanks
[19:59] <kees> hi! where can I find the code that manages the ppa.launchpad.net port 22 listener?
[20:00] <lifeless> its called 'poppy' in the code base
[20:00] <lifeless> whats up?
[20:02] <elmo> no, it's not
[20:02] <elmo> that's the ftp one
[20:02] <elmo> he means the ssh based one
[20:02] <lifeless> elmo: I thought they had common code
[20:02] <elmo> I hope not
[20:02] <elmo> the ftp one is zope based
[20:02] <elmo> the ssh one is twisted
[20:02] <lifeless> oh yeah
[20:02] <lifeless> we should fix that
[20:03] <lifeless> anyhow, find the tests for one, you'll find the tests for the other
[20:05] <lifeless> kees: whats up?
[20:13] <jelmer> kees: lib/lp/poppy/twistedsftp.py ?
[20:43] <wgrant> They do share some stuff.
[20:46] <wgrant> Thanks elmo.
[20:49] <jelmer> 'morning wgrant
[20:49] <wgrant> Morning jelmer.
[20:50] <lifeless> wgrant: welcome
[20:50] <thumper> \o/ finished with email
[20:50] <thumper> and a formal welcome to wgrant
[20:50] <thumper> :)
[20:50] <thumper> embraced fully at last
[20:51] <wgrant> Heh, thanks lifeless, thumper.
[20:51] <jelmer> Ah, right.. forgot it's already Monday over there. :-)
[20:51] <jelmer> wgrant: Welcome!
[20:51] <thumper> hi jelmer
[20:51] <jelmer> 'morning Tim
[20:55] <wgrant> jelmer: Were you able to work out which package supposedly has broken shlibs?
[20:55] <wgrant> jelmer: synapse is the only one I can see, and it depends on (>= 0.2.14)...
[20:56] <jelmer> wgrant: no, I didn't look into that issue further
[21:01] <thumper> bug fixed for bugjam \o/
[21:02] <thumper> well, was fixed months ago
[21:02] <thumper> but closed is closed
[21:34] <wgrant> No ~launchpad admins around at the moment, I suppose?
[21:36] <thumper> wgrant: not until spm starts
[21:42] <kees> lifeless, jelmer, cool, thanks. I was curious after reading https://bugs.launchpad.net/soyuz/+bug/689199
[21:42] <_mup_> Bug #689199: SSHFP DNS record for ppa.launchpad.net <fingerprint> <ssh> <ssl> <Soyuz:Confirmed> <Ubuntu:Invalid> <https://launchpad.net/bugs/689199>
[21:42] <kees> er, wrong one. this one: https://bugs.launchpad.net/soyuz/+bug/689213
[21:43] <jelmer> kees: are you sure?
[21:43] <kees> so is the poppy code the same as the ppa downloader?
[21:43] <kees> jelmer: sure of what?
[21:44] <jelmer> kees: That last URL, it gives me a 401
[21:44] <kees> it's a private bug, that somehow didn't get lp security subscribed.
[21:45] <kees> there, subbed now.
[21:45] <kees> /lib/lp/services/sshserver looks more like the core?
[21:45] <elmo> kees: poppy is the codename for the code behind upload.ubuntu.com and it's PPA equivalent
[21:45] <elmo> kees: whether it's ftp (original) or ssh (newer)
[21:46] <kees> ah righto, now I remember
[21:47] <lifeless> kees: ppa downloader?
[21:47] <kees> lifeless: I can't type. uploader.
[21:47] <jelmer> kees: thanks
[21:47] <lifeless> hah
[21:52] <wallyworld> thumper: wanna chat?
[21:52] <thumper> wallyworld: yeah man
[21:52] <thumper> wallyworld: mumble?
[21:52] <wallyworld> thumper: yep
[21:57] <jelmer> lifeless: I think Jacob's point was that the fact that ssh server offered port forwarding was widening the attack surface.
[22:26] <lifeless> I must have misunderstood things then