- From: David Brownell <david-b@pacbell.net>
- Date: Fri, 26 May 2000 12:22:02 +0100
- To: "John Cowan" <jcowan@reutershealth.com>
- Cc: <xml-editor@w3.org>, <xml-dev@xml.org>
> Rightly or wrongly, XML processors are licensed to signal an error ... and the fine print for such "error" permits absolutely any behavior, from ignoring the case to making demons adopt strange flight patterns. I've taken to calling them "optional errors" (from the perspective of a parser); such restrictions appear to be directed to document providers, not parser writers. Just like all those "for compatibility" constraints. > if > they detect a fragment id in a system identifier, ... > > We believe that many parsers silently ignore fragment ids, a > behavior which the XML Rec does not license. The question is: Double check that fine print. Silently ignoring is absolutely permitted for any "error" that's not in the category of "validity error", "well-formedness error" or "fatal error" (the encoding issue is characterized as a "fatal error" as I recall). - Dave
Received on Friday, 26 May 2000 07:21:38 UTC