[13:27] <bjf> when following the links on an email validation email i am getting "A server error occurred"
[13:29] <qengho> bjf: Huh. Request another? It's possible for an account to move into an impossible state that the validation link dislikes.
[13:37] <bjf> a second request results in the same thing ... "A server error occurred"
[13:39] <cjwatson> bjf: Is there an OOPS ID or anything like that?
[13:44] <bjf> cjwatson, no
[13:45] <bjf> cjwatson, "Unauthorized  A server error occurred."
[13:45] <bjf> cjwatson, we are setting up a role account for our LP bot. the email address IS set up for us is ubuntu-kernel-bot@canonical.com
[13:46] <cjwatson> bjf: What's the URL, with any tokens removed?
[13:46] <bjf> https://launchpad.net/token/<token>/+validateemail
[13:47] <cjwatson> I'll see if I can find any traces in logs.
[14:04] <cjwatson> bjf: can you tell me what IP address you would have been coming from?
[14:05] <bjf> cjwatson, 184.9.145.80
[14:07] <cjwatson> bjf: Huh, no matches; are you on the VPN?
[14:07] <bjf> cjwatson, yes
[14:08] <cjwatson> OK, found some plausible entries, let's see
[14:31] <cjwatson> bjf: You're hitting a crash while attempting to render a "forbidden" page (d'oh).  This suggests to me that perhaps you're trying to follow that link when logged in as some different user.  Were you perhaps logged in using your ordinary LP account?
[14:41] <bjf> cjwatson, let me try something
[14:42] <cjwatson> You may need to pass the validation link to IS and get them to follow it while logged in as your LP bot.
[14:42] <bjf> cjwatson, you are correct.
[14:43] <bjf> cjwatson, i managed to copy/paste the link and use it with the correct, new account, and it just worked
[14:43] <cjwatson> Excellent
[14:44] <cjwatson> Sorry for the confusion there
[14:44] <bjf> cjwatson, np, just one of those things... should have thought about it myself
[16:23] <rbasak> In a bug list, there's a "Date last updated" column. Is this date in UTC, or local time?
[16:23] <rbasak> We're working out a process based on this column, so need to know if we have a race condition :)
[16:24] <cjwatson> rbasak: You mean in browser pages?
[16:24] <rbasak> Yes
[16:25] <rbasak> Empirically we think it's UTC.
[16:27] <cjwatson> rbasak: Times displayed in the web UI should always be formatted using the requesting user's local time.
[16:27] <cjwatson> Let me double-check.
[16:28] <cjwatson> Yep, it's local time.
[16:28] <cjwatson> Moved LP's idea of my timezone nine hours east, and the top entry on https://bugs.launchpad.net/launchpad/+bugs?orderby=-datecreated changed from "Last updated on 2016-08-16" to "Last updated on 2016-08-17".
[16:28] <cjwatson> Not sure what tests you performed that made it seem like UTC.
[16:29] <nacc> the "last updated 23 hours ago" value being the same for rbasak and myself, I think
[16:29] <cjwatson> Note that it's LP's *stored* idea of your timezone that matters, which you can change from https://launchpad.net/~/+editlocation - not e.g. your system time
[16:29] <nacc> ah
[16:29] <cjwatson> Well, if it was 23 hours ago then it was 23 hours ago.  Timezone doesn't matter for that.
[16:29] <rbasak> And mine is set to UTC :)
[16:30] <cjwatson> You need to look at one that renders as an actual date to tell the difference.
[16:30] <rbasak> OK, we're all set on UTC. So we'll document that and go with it for now. Thanks :)
[16:30] <rbasak> Yeah we were comparing and they were the same. Hence our confusion.
[16:30] <rbasak> It all makes sense now ;)
[16:30] <cjwatson> OK.  (Perhaps you should use the API instead?)
[16:30] <cjwatson> In [1]: lp.bugs[1].date_last_updated
[16:30] <cjwatson> Out[1]: datetime.datetime(2016, 6, 27, 21, 19, 10, 577872, tzinfo=TimeZone(0))
[16:31] <rbasak> Yes, we'll switch to the API if we have to.
[16:31] <cjwatson> Unfortunate that bug listing don't have the usual hover text on the dates with the exact time.
[16:31] <rbasak> We're using the date to divide work up, so just wanted to sure there were no gaps/overlap.
[16:32]  * cjwatson nodes
[16:32] <cjwatson> Er, nods
[18:23] <Odd_Bloke> Don't node.  (Node even once.)