On Thu, Nov 07, 2019 at 04:43:17PM +0000, Ben Hutchings wrote: > On Wed, 2019-11-06 at 21:18 +0100, Greg Kroah-Hartman wrote: > > On Wed, Nov 06, 2019 at 07:59:32PM +0100, Pavel Machek wrote: > [...] > > > I'm confused. You said "by Tue ... 08:27:02 PM UTC". That 8 PM is 20h, > > > but did the release on 10h GMT+1, or 9h UTC -- 9 AM.... so like 11 > > > hours early, if I got the timezones right. > > > > > > Does PM mean something else in the above context? > > > > Ugh, no, you are right, I was ignoring the PM thing, I thought the -u > > option to date would give me a 24 hour date string, and so I thought > > that was 8:27 in the morning. > > > > Let me mess around with 'date' to see if I can come up with a better > > string to use here. I guess: > > date --rfc-3339=seconds -u > > would probably be best? > > The --rfc-822 option should give you something close to the current > format, but with 24-hour numbering. Ah, missed that. I'll go use that. Also, I just remembered that if you look in the email headers, there's a "X-KernelTest-Deadline:" field that uses the --iso-8601=minutes format. I think kernelci uses that to parse for when it needs to finish things by, but I don't know if that actually happens or not. I'll go fix the string up for the next round. thanks, greg k-h
