Re: [Emailcore] WGLC: draft-ietf-emailcore-rfc5322bis-03

[Steffen Nurpmeso <[e-mail]>]

John C Klensin wrote in
 <969063F2609AB90C285E5B77@PSB>:
 ...
 |Representing seconds should be a non-problem 

The problem is about the time conversion that happens.

  . The mail sender places Date:, according to the standard this
    SHOULD happen in local time.

  . The mail receiver interpretes Date: including the given +HHMM
    offset, and converts this to time zone of the receiver.

    This causes a faulty date representation to the end user on
    the receiver side.

  ...
 |First of all, your relationship with MUA users is not, at least
 |so far, an IETF problem.  It is certainly not a problem for
 |5322bis.
 ...
 |So I don't understand the problem.

Past experience may have guided you wrong.

  ...
 |> The same user made an identical request to the exim MTA
 |> in 2020, and they seem to follow it, at least in parts.  They
 |> simply switch to UTC, whereas i thought the following would
 |> apply better:
 |> 
 |>   Though "-0000" also indicates Universal Time, it is used to
 |>   indicate that the time was generated on a system that may be
 |>   in a local time zone other than Universal Time and that the
 |>   date-time contains no information about the local time zone.
 |> 
 |> Maybe the standard should explicitly mention the above approach
 |> for all timezones which require second precision not to loose
 |> the correct timestamp?
 |
 |I think there are two separate issues in the above.  The first

exim of the University of Cambridge (originally) choose to simply
use UTC in the given case, leaving no indication of the actually
used timezone of the sender.  This change was committed in 2020
and is thus likely (i am not an exim user nor did i look that much
into their source repository) in use on a large number of boxes.

As the issue is new to me i do not know how many programs actually
circumvent the problem in at least the way exim does it, at all.
All which do not produce mail messages that will show up with
a wrong time (potentially, date) in the MUAs on the receiver side.
(I think we are talking about the majority of software.)

And doing the trick and using UTC on the other hand contradicts
the SHOULD on the indication of local timezones.
The standard does offer the -0000 hint, but that seems not to have
been heard yet.

--steffen
|
|Der Kragenbaer,                The moon bear,
|der holt sich munter           he cheerfully and one by one
|einen nach dem anderen runter  wa.ks himself off
|(By Robert Gernhardt)