- From: Julian Reschke <julian.reschke@gmx.de>
- Date: Wed, 09 Apr 2008 19:33:11 +0200
- To: Mark Nottingham <mnot@mnot.net>
- CC: Henrik Nordstrom <henrik@henriknordstrom.net>, 'HTTP Working Group' <ietf-http-wg@w3.org>
Mark Nottingham wrote:
>
> Not clear. My take (as a native American speaker who has been corrupted
> to speak en-au sometimes) is that 'the' implies that there is only one
> possible set.
>
> However, I'm not that fussed about it; I'm happy to give it to the
> editors to decide (and I suspect they'll change it to 'the').
>
> Unless someone has a good technical argument otherwise, and is
> passionate enough about it to hold us up, we'll let them decide.
I just re-read the thread, and it seems Henrik's proposal was the most
convincing:
[...] lists the set of methods advertised as supported ...
So we keep the "the" (meaning it's not totally random), but we clarify
that the server may have reasons not to include all.
So that's the change (*) I'll apply in a few minutes, in the hope that
we can finally close this one.
That being said, would anybody object if I changed the example from
Allow: GET, HEAD, PUT
to
Allow: GET, HEAD, OPTIONS, PUT
-- I really can't think of a reason not to advertise OPTIONS, and we
don't want people encourage not to support it, right?
BR, Julian
(*) Plus the change for the client requirements.
Received on Wednesday, 9 April 2008 17:34:02 UTC