- From: Chris Lilley <chris@w3.org>
- Date: Mon, 8 Jul 2002 14:11:47 +0200
- To: eric@w3.org, public-ietf-w3c-request@w3.org, ned.freed@mrochek.com
- CC: Mark Baker <distobj@acm.org>, "Ian B. Jacobs" <ij@w3.org>, Lloyd Wood <L.Wood@eim.surrey.ac.uk>, John Stracke <JStracke@incentivesystems.com>, Martin Duerst <duerst@w3.org>, <ietf@ietf.org>, <public-ietf-w3c@w3.org>
On Sunday, July 7, 2002, 8:51:45 PM, ned wrote: >> On Tue, Jul 02, 2002 at 03:39:50PM -0400, Ian B. Jacobs wrote: >> > [1] http://www.w3.org/2002/06/registering-mediatype >> I've asked around about whether this document represents an agreement >> between the relevant parties, but so far all I've heard second hand is >> "I think so". >> Can anybody confirm, for the public record, that this is the Way? nfmc> It certainly is A Way. I'm not sure it rises to the level of The Way, but I see nfmc> nothing procedurally wrong here. >> (FWIW, I'm very happy with it - it's simpler than even I was hoping for) It is fairly simple. I note that, although the discussion period and possibly IETF last call happens in sync with teh W3C last call (which is good) the actual registration still only happens "Once the W3C specification has become a Recommendation". Thus, the catch-22 or chicken-and-egg situation is still not resolved: - mime type registration requires a stable document - stable document requires meeting W3C implementation criteria such as Candidate Recommendation - getting implementation experience for CR requires using the mime type So, shouldn't section 3.2 be renumbered 2.3? In other words, register it before using it, use it to gain implementation experience, and update the registration to take into account implementor feedback, if needed. Since W3C considers a CR stable enough to ask non-bleeding-edge implementors to go implement it, it should also be stable enough to do the registration, too? -- Chris mailto:chris@w3.org
Received on Monday, 8 July 2002 08:13:20 UTC