[LEAPSECS] email headers

[Tony Finch]

Another place to look is the IMAP sort extension. Section 2.2 of http://tools.ietf.org/html/rfc5256 says to treat invalid times as 00:00 which might be triggered by a leap second. Also since mailbox order is the tiebreaker (the sort must be stable) you can find out if your IMAP server treats xx:59:60 the same as xx:59:59 or xy:00:00.

Tony.
--
f.anthony.n.finch  <dot at dotat.at>  http://dotat.at/

On 1 Jul 2012, at 01:00, Zefram <zefram at fysh.org> wrote:


> According to RFC 5322 (and its predecessor RFC 2822), the Date: header

> that I'm putting on this message is valid:

> 

>    Date: Sun, 01 Jul 2012 00:59:60 +0100

> 

> I wonder how many MUAs will be unable to date-sort it.  The header

> syntax is specified rather loosely: the BNF allows "99:99:99" as a

> time-of-day, it being merely semantically forbidden, and the usage of

> ":60" is only weakly implied to be limited to leap seconds.  This is

> typical of email header definitions, particularly the older ones, and

> thus the tradition is for email headers to be handled very forgivingly.

> So I presume this message will pass through the mail transport system

> unhindered, and be readable in anyone's MUA, even where software thinks

> the header is bogus.  So I think date sorting is all that's likely to

> go wrong with this message.

> 

> -zefram

> _______________________________________________

> LEAPSECS mailing list

> LEAPSECS at leapsecond.com

> http://six.pairlist.net/mailman/listinfo/leapsecs