[dhcwg] one remaining issue with the timezone draft: what happens if a server only implements one of the two options?

[Eliot Lear <lear at cisco.com>]

Stig has raised a concern that a client might have to play guessing
games with a server to determine which of the TZ options (if any) it
implements.  There are several approaches that could be taken:

    * The nascent proposal to do something with option sets and
      preferences.  While elegant, I would rather not gate this draft on
      that draft's progress.  Also, it could limit deployment of this
      option until that option is implemented.  However, once available
      there is no reason why the tz options could avail themselves of
      such an option in the future.
    * A preference of one option over the other such that if both are
      requested only one is returned.  This would be a special case of
      the first option.  Sounds like additional code useful only to the
      timezone option.  I do not recommend this approach.
    * Stating that if a server implements one of these options it SHOULD
      implement both to reduce the risk of a mismatch between client and
      server.  There does exist very good heuristic code to go from one
      to the other, and we reference that code in the draft (it's EMACS
      elisp).  I recommend this option.

And so without objection I will act on the third approach.  I would like
to point out that Stig is the originator of this idea, and thank him for
catching this problem and providing a reasonable alternative to move
forward.

Eliot

_______________________________________________
dhcwg mailing list
dhcwg at ietf.org
https://www1.ietf.org/mailman/listinfo/dhcwg