- From: Liam R E Quin <liam@w3.org>
- Date: Tue, 04 Sep 2012 20:34:23 -0400
- To: John Cowan <cowan@mercury.ccil.org>
- Cc: James Clark <jjc@jclark.com>, public-microxml@w3.org
On Tue, 2012-09-04 at 18:25 -0400, John Cowan wrote: > 2) Versioning this kind of formats just doesn't work, as I have > discovered the hard way. I still believe that it *can* work if defined from the start and not screwed up. What I think of as the BSD shared library versioning system can work fine - 1. all m.n processors must accept x.y documents with x <= m 2. point releases (incrementing n and y) can't change semantics except to define things that used to be errors (so a 3.1 processor must try to process a 3.6 document, but might throw an error on a new feature) 3. if an incompatibility is introduced, the major number must be increased in the spec. XML 1.1 would have been called XML 2 by this rule. But, we didn't get consensus on it soon enough (1997...). We already know XML 1.2 and XML 2 will both be rejected by most XML processors today. But we could do better for µXML, perhaps. Liam -- Liam Quin - XML Activity Lead, W3C, http://www.w3.org/People/Quin/ Pictures from old books: http://fromoldbooks.org/
Received on Thursday, 6 September 2012 00:14:33 UTC