- From: David Brownell <david-b@pacbell.net>
- Date: Mon, 11 Jun 2001 08:39:44 -0700
- To: Rob Lugt <roblugt@elcel.com>, xml-editor@w3.org
- Cc: mary.brady@nist.gov, xml-dev@lists.xml.org
> We downloaded the second edition of the OASIS/NIST Conformance Test Suite > [1]. As has been reported before, several files were missing from the > archive so we applied the patch [2] form the XML Testing Subproject [3] on > sourceforge.net. This appears to fix the missing file problems, all the > other failures seem (to us) to be errors in the individual tests. I'd be interested in seeing at least a second patch (apply after the first), perhaps addressing the issues raised in Rob's note if not the larger set which Richard noted. It's dangerous to have the conformance test suite have conformance bugs; it encourages proliferation of nonconformant XML-based applications, which are hard to un-deploy. Getting agreement on bugs in the tests is important. I think it's clear that there are some UTF-8 issues in IBM's contribution, some of the canonical output cases, and (reported against 1st draft of tests) some content models that should be moved to the "you don't really need to report this error" category. A patch addressing those cases ought to be non-controversial. Likewise with many of the other issues that Richard reportad, though I've not had time to look at all of them. - Dave
Received on Monday, 11 June 2001 11:40:59 UTC