[00:44] veebers: got a few minutes? [00:44] thumper: otp with babbageclunk [00:44] if it is dependency related, I have information [00:44] and suggestions [00:45] veebers, babbageclunk ^^ [00:56] thumper: well, was related to the move to macaroon/bakery v2 (failing test). [00:57] I'm considering a somewhat more agressive stance on imports [00:57] which impacts all this work [00:58] thumper: did you want to chat now, or should I have lunch first? I'm needing to pop out for an errand shortly [00:58] perhaps quick now? [00:58] 5 min max [00:59] https://hangouts.google.com/hangouts/_/canonical.com/dep-hell [00:59] thumper: ack, 5 min timebox :-) [01:10] thumper: could u plz come into babbageclunk's office (aka a-team standup)? [01:10] thumper: https://hangouts.google.com/hangouts/_/canonical.com/a-team-standup?authuser=0 [01:10] ack [03:30] jam: how goes the removal of api server machines? [03:30] babbageclunk: got a few minutes? [03:35] thumper: yup [03:35] thumper: in 1:1? [03:36] ack [03:37] hmm, we don't appear to have a 1:1 anymore? I'm in the hangout called christian-tim === frankban|afk is now known as frankban [08:00] externalreality: gave some feedback on 8555 === fnordahl_ is now known as fnordahl [08:04] jam, Ack on the feedback. === tinwood_ is now known as tinwood [08:41] babbageclunk, veebers: if you're around, i replied to that mail thread about the macaroon error response [08:52] rogpeppe: hey o/ I'm a little bit around :-) [08:52] veebers: hiya [08:52] veebers: so ISTM that this probably isn't a macaroon size issue at all [08:52] rogpeppe: what's the best way to get you the comparison data? Print out the headers and body from somewhere? [08:52] rogpeppe: oh? [08:53] veebers: did you see the email i just sent? [08:53] rogpeppe: just reading now, oh are you saying introduced in juju/juju (that rev?) [08:53] veebers: i don't know exactly when this has been changed, but pretty recently [08:55] rogpeppe: interesting. There was a large branch landeed by babbageclunk but that might be a red herring [08:55] veebers: looks like it was done in eab1c2550ffc0695a2dfa1ab9583d62863efc41b [08:55] ah, that would be the large one I mentioned :-) [08:55] veebers: BTW isn't that a backwardly incompatible change? [08:56] rogpeppe: which, eab1c25? [08:57] veebers: yeah [08:57] rogpeppe: I'm not familiar with the changes, only aware it went in [08:57] veebers: ok [08:58] veebers: well, given that that commit seems to have been working, perhaps it's worth looking at the size issue too [08:58] rogpeppe: what gives you concern? (I have faith in babbageclunks and wallyworlds reviews) [08:58] rogpeppe: aye, the test pass on that commit [08:59] veebers: if an old client connects to the new server, i'd be concerned that it can still see the macaroon error response [08:59] veebers: maybe it decides what to do based on something in the request though [09:01] rogpeppe: I can mention that tomorrow if you like, get some closure :-) [09:01] veebers: i'm just trying to work out what goes on in the code [09:02] veebers: in the meantime, it would be great to get an example of the old and the new macaroons that are being returned from the api server [09:03] veebers: to see why the size is changing [09:03] veebers: because that may be indicative of some other issue [09:07] rogpeppe: posted you some logs, let me know if that's useful or not, I'll grab from tip of develop now (i.e. v1 macaroons) [09:07] veebers: sorry, my network stack just died [09:08] veebers: what was the last thing you saw me say? [09:08] veebers: last thing i saw from you was "I can mention that tomorrow if you like, get some closure :-)" [09:08] rogpeppe1: ugh sucks. I saw:veebers: because that may be indicative of some other issue [09:08] oh [09:09] rogpeppe1: I chucked you some logs, let me resend, let me know if it's the details you need === Spads_ is now known as Spads === aluria` is now known as aluria === Berg is now known as Guest57443 === frankban is now known as frankban|afk === mwhudson_ is now known as mwhudson