Re: [Tzdist] draft-ietf-tzdist-caldav-ref-02

[Cyrus Daboo <[e-mail]>]

Hi Ken,
Following up on this now that WG last call is over.

--On May 4, 2015 at 12:13:46 PM -0400 Ken Murchison <[e-mail]> 
wrote:

> My only comment/concern on this document is the new text in Section 3.1.3
> regarding use of the Prefer header field for clients to require a
> 'calendar-no-timezone' capable server to include VTIMEZONE components in
> the iCalendar data.  My reading of RFC 7240 (Section 2, para 1) leads me
> to believe that a server is free to ignore any/all preferences in the
> Prefer header, therefore a client would have no guaranteed way to opt-out
> of the tz-by-ref behavior.
>
> I could be reading RFC 7240 incorrectly, or maybe we can force
> 'calendar-no-timezone' capable servers to implement and obey the
> 'vtimezone' preference as this document leans towards doing.  Either way,
> we may want to reach out to the HTTPbis folks to get some input.
>
> Other alternatives OTH (perhaps not good ones) might be to use the Expect
> header field, or doing something with the Accept header field.

I do think what we are doing is OK, particularly given that we use the DAV 
header field to advertise the fact that the server will obey the Prefer 
option behaviors the spec defines. In any case, the Prefer header item will 
need to be registered and go through the designated expert review process - 
so let's get that done and see if they object to this usage.

-- 
Cyrus Daboo