[04:04] <nacc> rbasak: if you happen to be around, ping -- or in the next few hours, have an importer question for you
[05:16] <pitti> xnox_: DNSSEC> that smells like an SRU then? it should mostly be alright, but some devices still have trouble with it; and we don't do portal detection; so I'd disable it again
[07:44] <ackk> join #server
[07:59] <xnox_> pitti, yes, there is inflight bileto, hoping for unapproved queue SRU today/soon (need to fix up missmerged cherrypick first though)
[08:28] <rbasak> nacc: o/
[08:41] <rbasak> slashd: ~ubuntu-sru-developers should work now (thanks slangasek!). Could you perhaps review and upload some suitable SRU from the sponsorship queue to test?
[08:46] <LocutusOfBorg> hello jbicha I suppose I can close lp: #1657924
[09:04] <xnox> ok, systemd sru refix; and thrown into bileto - coffee time
[11:42] <jbicha> LocutusOfBorg: that's fine, my config file was broken so maybe some day I'll look into it more
[11:45] <LocutusOfBorg> thanks
[11:46] <LocutusOfBorg> will keep it as expired
[13:58] <slashd> rbasak, morning, I think my ACL is now set by the TB --> https://lists.ubuntu.com/archives/technical-board/2017-April/002302.html but I still can't approved series nomination, do you think it is something we can arrange ?
[14:03] <rbasak> slashd: can we check that uploading SRUs works first please?
[14:04] <slashd> rbasak, that was my second question, is there a way I can easily check before doing an real upload ?
[14:06] <rbasak> slashd: use ubuntu-upload-permission
[14:07] <slashd> rbasak, so far I checked with this : http://people.canonical.com/~ubuntu-archive/archive-permissions/individuals and I don't see my name, will try with ubuntu-upload-permission
[14:17] <nacc> rbasak: heya, I think I figured it out last night, but will see if i remember today
[14:32] <slashd> rbasak, seems like ubuntu-archive-tools indicate :
[14:32] <slashd> == All rights for slashd ==
[14:32] <slashd> Archive Upload Rights for ubuntu-sru-developers: archive 'primary', pocket 'Proposed
[14:33] <slashd> but ubuntu-uploader-permission doesn't reveal much
[14:35] <rbasak> slashd: it looks like ubuntu-uploader-permission can't check for a specific pocket.
[14:35] <rbasak> slashd: can you perhaps check the sponsorship queue and try and review and upload an SRU from there?
[14:35] <jbicha> slashd: quick, fix some bugs! :)
[14:37] <slashd> jbicha, rbasak, sure will have a look this week, and once having 1 upload I'll ping back for the series nominations ?
[14:37] <rbasak> Yes
[14:37] <slashd> rbasak, sound like a good plan, thanks
[14:37] <rbasak> I don't want to chase up series nominations until we know the upload actually works. Since otherwise it could just be that.
[14:38] <slashd> rbasak, totally make sense thanks
[14:42] <nacc> bdmurray: slangasek: looking at LP: #1654008, it seems like the SRU of update-manager isn't actually available (but was at some point?). rmadison indicates it's a source only package now in t-u.
[14:42] <bdmurray> nacc: cjwatson mentioned fixing in #ubuntu-release
[14:43] <nacc> bdmurray: oh ok, someone was asking in #ubuntu
[14:43]  * nacc will go look at logs
[14:43] <cjwatson> I forgot to accept my copy apparently
[14:44] <nacc> cjwatson: :)
[14:48] <nacc> rbasak: re: LP: #1680064, can you change the default branch back so i can debug?
[14:51] <rbasak> nacc: it won't let me do it as it doesn't exist :-/
[14:51] <rbasak> I wonder if I should delete the repo and then re-push it.
[14:51] <nacc> rbasak: i'm testing with my own repo, let me see
[14:52] <nacc> rbasak: that's strnage as i was able to change it on the existing repos (when i was testing this)
[14:52] <rbasak> Oh.
[14:52] <rbasak> I have a hunch as to what failed.
[14:55] <rbasak> nacc: sorry, my fault. I didn't actually check if "usd import" had claimed success, only that the repo appeared pushed (as I'd switched machine)
[14:55] <rbasak> Looks like it did actually fail, so never got beyond the push.
[14:55] <rbasak> Sorry for the trouble.
[14:55] <nacc> rbasak: ah!
[14:55] <nacc> rbasak: yeah :) now why did it fail?
[14:55] <rbasak> lazr.restfulclient.errors.PreconditionFailed: HTTP Error 412: Precondition Failed
[14:55]  * rbasak shrugs
[14:55] <rbasak> Never seen that before.
[14:55] <nacc> ah ... lp something something?
[14:55] <rbasak> Usually it hangs at that stage.
[14:56] <nacc> hrm, i've also not seen that
[14:57] <nacc> rbasak: fyi, i'm working on or retooled patches-applied algorithm (LP: #1662325)
[15:05] <cjwatson> "precondition failed" is normally code for mid-air collision
[15:05] <nacc> cjwatson: ah ok, thanks
[15:05] <cjwatson> the etag sent in your request to represent the state of the object you're modifying doesn't match the current state
[15:14] <rbasak> nacc: so that was from:
[15:14] <rbasak>   File "/home/robie/Code/Canonical/usd-importer/usd/importer.py", line 1092, in main
[15:14] <rbasak>     lp_git_repo.lp_save()
[15:14] <rbasak> If that matters.
[15:14] <nacc> rbasak: the 412 ?
[15:15] <nacc> rbasak: hrm, well that's what updates the default branch :)
[15:15] <rbasak> Yep :)
[15:16] <nacc> hrm, it's odd, because the lp_git_repo object is obtained from lp just before that
[15:16] <nacc> so i'm not sure why there would be a collision
[15:16] <nacc> rbasak: did you keep the directory? is it reproducible (it should try to adjust the default branch on every push)
[15:16] <rbasak> Yes I have the directory
[15:17] <rbasak> I can try another clone
[15:17] <rbasak> Uh, import
[15:17] <nacc> yeah
[15:17] <rbasak> It reran without error
[15:18] <nacc> hrm, ok -- so i'd presume to say lp something something for now ...
[15:18] <rbasak> But I can't change HEAD back.
[15:18] <rbasak> "This repository does not contain a reference named 'refs/heads/master'."
[15:18] <rbasak> I can try delete and recreate if you want?
[15:19] <nacc> rbasak: ah so it can be already set to an invalid value, but you can't set it to one
[15:19] <rbasak> Right
[15:19] <nacc> rbasak: up to you -- would be good to see, as in my testing it was fine, so if it's reproducible, should be easy to debug and fix
[15:19] <rbasak> OK I'll try delete and recreate now.
[15:20] <nacc> rbasak: thanks
[15:20] <rbasak> It failed again
[15:20] <rbasak> Different error
[15:20] <nacc> heh
[15:21] <rbasak> http://paste.ubuntu.com/24408024/
[15:21] <rbasak> Race between Launchpad getting the push and processing the repo?
[15:21] <nacc> that's what i'm wondering now
[15:21] <nacc> which would makes sense as to why you got the 412 before based upon cjwatson's comment
[15:21] <rbasak> Yeah
[15:21] <nacc> cjwatson: --^ after a push to a repository, is there a way to ensure it's 'settled' on the lp side before we grab a reference to that repository and change state?
[15:22] <cjwatson> You could certainly be racing with the scanning job
[15:22] <cjwatson> What attributes do you need to change exactly?
[15:22] <nacc> the default branch
[15:22] <rbasak> Just the HEAD symbolic ref.
[15:22] <rbasak> Since we don't have a "master".
[15:22] <rbasak> Incidentally, a bunch of other stuff assumes that HEAD exists, such as git-build-recipe. I hit that today.
[15:23] <cjwatson> Can't you just push HEAD?
[15:23] <rbasak> (even when I don't need to to exist)
[15:23] <cjwatson> I forget exactly how that works, I know it's weird in git
[15:23] <rbasak> It used to be that you couldn't do that.
[15:23] <rbasak> But it might have changed more recently.
[15:24] <cjwatson> Ah yes, I had to implement an extension for that for git-to-git imports
[15:24] <cjwatson> OK, so I think you should probably just loop-and-retry-on-412
[15:24] <nacc> got it
[15:24] <nacc> i can do that
[15:24] <rbasak> Or 404
[15:24] <rbasak> (this second error was a 404)
[15:25] <nacc> right, and in our case, presuming push() passed, we know it should exist eventually
[15:30] <nacc> rbasak: something like http://paste.ubuntu.com/24408055/
[15:30] <nacc> rbasak: as i think we need to re-grab the repo object to avoid the 412
[15:31] <rbasak> nacc: looks good. Do we need a timeout?
[15:31] <nacc> rbasak: not sure :)
[15:31] <nacc> rbasak: probably? but i'm not sure how to do that
[15:31] <nacc> rbasak: i think not for 412, but for 404 maybe we should?
[15:32] <rbasak> Probably easiest to just loop until timeout and ignore all 412 and 404.
[15:32] <nacc> rbasak: yeah, but what timeout? :)
[15:32] <nacc> rbasak: 1m?
[15:32] <rbasak> 30s? 60s?
[15:33] <cjwatson> nacc: .lp_refresh() would work too
[15:33] <cjwatson> (well, that doesn't work with 404, only if the .lp_save() fails)
[15:33] <nacc> cjwatson: got it
[15:35] <cjwatson> I'd do something like lazr.restfulclient._browser.Browser._request_and_retry does if I were you
[15:35] <cjwatson> exponential backoff with a limit
[15:35] <nacc> cjwatson: thanks, will look
[16:37] <nacc> rbasak: can you test http://paste.ubuntu.com/24408457/ ?
[16:37] <nacc> i'm not sure if the default is sensible (3 retries with exponential backoff)
[16:40] <rbasak> ack
[16:40] <nacc> rbasak: thanks
[16:47] <rbasak> nacc: it worked
[16:47] <rbasak> I had to try twice because the first failed on the push, which I think is a bug in libgit2 in Xenial.
[16:48] <rbasak> The second time (after deleting the repo from LP again) it worked fine.
[16:49] <nacc> rbasak: ok cool, i'll push it up then
[16:49] <rbasak> Thanks!
[16:50] <nacc> rbasak: thanks for the quick test
[17:29] <hjd> Has it been announced anywhere when the AA-archive will open? :)
[17:37] <cjwatson> hjd: Nobody can give an estimate until we have a name from Mark
[17:38] <hjd> ah, ok
[17:46] <jbicha> and it's usually at least a few days after that
[18:07] <sarnold> nacc: thanks, I'm not sure where 1678349 fell through the cracks.
[19:01] <nacc> sarnold: thanks! yeah, i can't see it, so I can't tell it either, but figured a ping here was the right way to get it to your team
[19:09] <slashd> rbasak, I got added to the ~ubuntu-release-nominators team, I think I'm all set now, thanks again
[19:10] <rbasak> slashd: great!
[20:11] <jbicha> slashd: are you looking for stuff to SRU? are you able to SRU anything in main?
[20:18] <jbicha> if so, I recommend bug 1550210 which was worked around by both Ubuntu GNOME and Ubuntu Studio, but see my comment at the end
[20:21] <slashd> jbicha, haven't tried yet, will give it a try this week
[20:22] <ricotz> jbicha, just released Thunderbird 52 is GTK3 ;)
[20:24] <jbicha> ricotz: great, I encourage you to reply to the list. Do you have an opinion on the email client question?
[20:26] <ricotz> jbicha, I guess I am too old school to have a objective opinion
[20:27] <ricotz> aka using thunderbird for over 15 years
[20:27] <infinity> mutt 4 lyf, yo
[20:27] <infinity> *gang sign*
[20:29] <rbalint> infinity:  :-)
[20:31] <nacc> rbasak: interesting, i think with very few changes, i've got the spph from a ppa now, just need to figure out how to get to the files with the right object (.dsc, etc). Also in doing some digging, ubuntutool has a UbuntuCloudArchiveSourcePackage object (need to look still what it defines)
[20:33] <tsimonq2> infinity: I recently started playing with mutt, and it's impressive, IMHO. I'll still use Thunderbird when I'm on my main system, though.
[20:34] <infinity> tsimonq2: mutt's a complete non-starter for people who need to read a lot of html email, especially with embedded images.  I, however, have the luxury of being able to say "I'm a nerd who won't read your mail if the majority of the content is in images, kthx."
[20:34] <nacc> heh
[20:34] <tsimonq2> Hahaha.
[20:35] <sladen> infinity: this is a feature
[20:35] <tsimonq2> infinity: Well all the mailing lists I'm subscribed to are *explicitly* set to use only plaintext, and when I need to view HTML mail (rare nowadays) I just use Roundcube Webmail (which is an AWFUL program for doing anything but reading non-encrypted, HTML emails)
[20:35] <stgraber> what, e-mails can contain images? :)
[20:37] <tsimonq2> The only things I can think of that send me HTML emails are confirmation emails for signing up to things, and Viagra spam. (the latter of which I have an imapfilter + SpamAssassin setup to deal with, but it's not PERFECT)
[20:38] <tsimonq2> infinity: But if I'
[20:38] <tsimonq2> Grrrrrrrrrrrrrrrrrrrrrrr
[20:38] <tsimonq2> If I'm sending patch mail, mutt hands-down
[20:39] <infinity> stgraber: It can, yeah, but it helps if you're tall enough to read above the signature footer.
[20:40] <tsimonq2> HAH
[20:40] <stgraber> infinity: haha, when I wrote that I was thinking, "how will Adam turn that into a short joke?", that's how :)
[20:40] <infinity> stgraber: Man, it wasn't easy.
[20:40] <infinity> stgraber: But I powered through.
[20:41] <tsimonq2> XD
[20:43] <infinity> I'm going out to buy all the hamburgers in a 2 mile radius.  Anyone need anything?
[20:44] <tsimonq2> infinity: Funny enough, I work at McDonald's, so if you give me 10 minutes I can tell you about all our deals in detail. XD
[21:03] <Zta> Crikey... Both Dropbox (https://www.dropbox.com/sh/l0vzh88wq0lipjv/AABUIgRsUCNFjI2QoC_fiJhma) and Bitbucket (https://bitbucket.org/stephan-henningsen/aseprite-dist) are giving me a hard time with deep-linking from a shared link.  That is, https://www.dropbox.com/sh/123/ works but https://www.dropbox.com/sh/123/dists/xenial/InRelease gives 404 even though the path technically exists.
[21:09] <Zta> Ah, getting there: deb https://bitbucket.org/stephan-henningsen/aseprite-dist/raw/master xenial main
[21:16] <Zta> Yay!  Success!
[22:06] <cjwatson> infinity: ":auto_view text/html" works fine as long as the formatting of the email isn't particularly important ...
[22:07] <infinity> cjwatson: I use sometihng along those lines, yes.
[22:08] <cjwatson> (what I normally do is do nearly all my routine mail handling with mutt, and use K-9 on my phone for anything that's too awkward to read that way)