On Tue, Apr 13, 2010 at 03:58:22PM +0200, Ben Campbell wrote:
-- date-and-time, pattern and description:
Which is the normative description for date-and-time? The ABNF in
the description, or the pattern attribute? I assume the second, but
fear the presence of ABNF will make others assume the first.
Ideally, they should be consistent - and I hope they are. The ABNF is
more detailed - if you read the comments - and copied from RFC 3339.
If we make a change, we should completely remove the ABNF from the
description and simply leave the pointer to RFC 3339, e.g.
For a more detailed description, see section 5.6 of RFC 3336.
Since the ABNF is copied, this does not really change much unless RFC
3336 gets updated perhaps. For now, I have left things as they are but
I am open to be convinced to remove the ABNF if someone feels strongly
about this.
I don't feel strongly--it was just a mild general concern that duplicate
_normative_ text can lead to future errors if, as you say, the RFC gets
updated. But if you see value in having the ABNF in the description, that's
okay with me. At the most, it might be worth putting a comment in the
description to see the RFC for the full normative definition.
Since there is an explicit reference to RFC 3336, I think this is
covered. Many of the data types document formats that are described in
other documents and there is a general trade off what to include and
what to leave out and to include by reference. I guess it is at the
end a judgement call and at this stage of the process, I simply prefer
to minimize changes.
/js
--
Juergen Schoenwaelder Jacobs University Bremen gGmbH
Phone: +49 421 200 3587 Campus Ring 1, 28759 Bremen, Germany
Fax: +49 421 200 3103 <http://www.jacobs-university.de/>
_______________________________________________
Ietf mailing list
Ietf(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/ietf