- From: Henry S. Thompson <ht@inf.ed.ac.uk>
- Date: Fri, 03 Oct 2008 10:31:07 +0100
- To: Elliotte Harold <erharold@gmail.com>
- Cc: xml-editor@w3.org
-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 The question of the actual value of the substantive changes to the name character inventory, first in XML 1.1 and now in the proposed 5th edition of XML 1.0, has been extensively discussed elsewhere (see, for instance, [1] [2]). There is no doubt that the original designers of XML intended it to be a major step forward in making the Web available to users of _all_ languages, not just those that can be written with ASCII, and this change is absolutely in line with that goal. With regard to your process point, the XML Core WG considered this for a long time before going the route we did. We were strongly influenced by the evident reluctance of the community to support XML 1.1, despite near-universal approval of its substance. There's no doubt making these changes is significant, but strictly speaking also allowed by our own rules: the Process Document sets certain requirements for a spec. to be published as a Proposed Edited Recommendation, particularly as regards the kinds of changes it involves and evidence of interoperability, and this change does in our opinion (and of the W3C Director, in that he approved the publication of the PER on this basis) satisfy those requirements. The word 'erratum' is perhaps misleading in this context. This is a change, but an allowed change: the changes in XML 1.0 5th Edition fall into class 3 from section 7.6.2 Classes of Changes to a Recommendation [3], as they affect conformance without introducing new features. As regards the Process requirement for implementation experience, the XML Core WG understands that implementing this relaxation in XML 1.1 parsers has been technically straightforward: it is a matter of replacing a rather large "permitted" table with a much smaller "forbidden" table. The interoperability of those parsers provided preliminary necessary evidence that interoperable implementation of the changes proposed in XML 1.0 5th edition will likewise be straightforward. More recent experience, detailed in the implementation report [4], confirms this expectation: four parsers, - From both vendor and open-source efforts, implement the changes and interoperate with respect to the tests which have been added to the XML Test Suite [5] to cover the new edition. We have brought this forward because we believe the benefits outweigh the costs, based on a number of efforts to sample the likely response. Please let us know if you are satisfied with this response. ht [1] http://lists.xml.org/archives/xml-dev/200802/msg00410.html [2] http://rishida.net/blog/?p=135 [3] http://www.w3.org/2005/10/Process-20051014/tr#correction-classes [4] http://www.w3.org/XML/2008/01/xml10-5e-implementation.html [5] http://www.w3.org/XML/Test/#LatestNews - -- Henry S. Thompson, School of Informatics, University of Edinburgh Half-time member of W3C Team 10 Crichton Street, Edinburgh EH8 9AB, SCOTLAND -- (44) 131 650-4440 Fax: (44) 131 651-1426, e-mail: ht@inf.ed.ac.uk URL: http://www.ltg.ed.ac.uk/~ht/ [mail really from me _always_ has this .sig -- mail without it is forged spam] -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.2.6 (GNU/Linux) iD8DBQFI5eZbkjnJixAXWBoRAhBiAJ4yFjofD2tp39wifGuowRMfRIJKzwCeKYYs TZW4itLaI7pQ9Y5asBLpYdI= =gs+a -----END PGP SIGNATURE-----
Received on Friday, 3 October 2008 09:31:43 UTC