- From: Robin Berjon <robin@w3.org>
- Date: Wed, 03 Apr 2013 18:03:38 +0200
- To: "HTML WG (public-html@w3.org)" <public-html@w3.org>
Hi all, I've been poring over the way in which the HTML specification source is organised in order to split out Microdata (which as you know isn't split in the WHAT version) and I have come to the conclusion that it is problematic. Without boring you with extensive diffs, the way it's done is basically this: • Anywhere that Microdata integrates with other parts of HTML, some small sections, sometimes just a few words, are fenced off to ensure they don't get generated as part of the HTML draft. • That content is then duplicated in a special section that is only included in the W3C Microdata draft. I'm sure that the fine bunch of hackers that you all are can immediately spot the problem here. That content easily goes out of sync. What's more, changes fencing off (or, worse, failing to fence off) a few words here and there in a 120K lines document are easily missed. So I don't think that that's a viable way forward, and am proposing a change. Microdata remains defined as a separate specification (I don't mind merging it if people prefer, but I don't think that that will be acceptable). However the integration points where it modified HTML are in HTML. Before anyone climbs on any manner or form of high horse, here are the aspects that this has an impact on: • Validation constraints (if you have itemprop, you must have href/src/etc.; some global attributes are added; under such and such condition some elements may becomes flow or sectioning content) • A small change to DnD • Appendix listings • Acknowledgements None of this in any way mandates MD, or grants it special status. It's just providing integration information, mostly for validators. If this is a concern for RDFa people, I would be more than happy to entertain a similar set up for RDFa if you think it makes sense. As a final note, please consider that while this is obviously open to discussion, maintaining the current system involves a lot of dull, error-prone make-work and painstaking bug and regression finding for the editors. We would therefore require that you kindly take that into account before objecting, and that you make sure that any objection are based on the type of solid concerns that justify sending sweet, innocent, funny, charming, and by and large beloved people into the salt mines of Mordor. -- Robin Berjon - http://berjon.com/ - @robinberjon
Received on Wednesday, 3 April 2013 16:03:48 UTC