Peter van Dijk wrote:
>
> > After doing some experimentation, I discovered that the IMAP server is
> > getting its time from the "From " header in the mbox, not the "Received:" or
> > "Date:" headers, and passing this on to Outlook. When using POP, Outlook
> > evidently gets the time itself from a "Received:" header.
>
> a From: header, I think. There is no common standard for Received-headers.
> Also, in qmail, these will be UTC as well :)
Umm, there isn't a timestamp in the "From:" header, just the mailbox
"From " header (I assume that's what you meant). Anyway, Outlook seems
to be able to cope with UTC in "Received:" headers, as in the POP case;
just not in the "From " header that IMAP passes, because there's no
indication that "From " is in UTC.
> > This 'fixes' the time seen in Outlook, I guess because either IMAP or
> > Outlook can now figure out the local time. My question is this: does this
> > solution violate any standards or will it break anything obvious? Why
> > doesn't qmail give some indication that the date/time given in this header
> > is not local time?
>
> I can't find anything about this in RFC822, probably because RFC822 doesn't
> describe mailbox formats, only message formats :)
Yeah, this was my experience.
> I don't know. I don't think it will break anything. I do think IMAPd is
> broken for supplying the user with that info. But UW IMAP is broken anyway.
> You do know that _another_ buffer overflow in UW IMAP was found somewhere
> in the last few days?
I agree completely. I'd much rather not use mailbox or UW IMAP, but
circumstances require them at this point. Perhaps sometime in the
future...
Thanks,
Eric
