W3C home > Mailing lists > Public > www-archive@w3.org > November 2008

Re: markup validator test log

From: olivier Thereaux <ot@w3.org>
Date: Mon, 17 Nov 2008 16:08:39 -0500
Message-Id: <5C38CA13-108D-4C21-9AA1-7983137F8C2C@w3.org>
To: www-archive@w3.org
Cc: Dan Connolly <connolly@w3.org>, Jose Kahan <jose.kahan@w3.org>, Ted Guild <ted@w3.org>, Jean-Guilhem Rouel <jean-gui@w3.org>

Thank you Dan for publishing the result of the markup validator  
automated testing (2008-11-17).

I looked at all the FAIL tests and can confirm that there is no  
regression since the latest release, 0.8.3.

One of the main issues causing a number of the tests to fail was a  
typo in the DTD published with the 2008-02 working draft of XHTML 1.1  
(a very odd case among W3C specifications: XHTML 1.1 used to be a REC,  
but is now obsoleted by... a moving target). I fixed the problem by  
updating the DTD in the validator catalog with the version currently  
served by the XHTML WG.

The other issues are either:
- features not yet enabled (appendix C checking)
- known old edge case bugs (bogus FPI, etc) - some of them should be  
documented in bugzilla
- problems with DTD (SVG 1.1 Basic)
- minor issues with the apache setup on the test server

As a result, I do not consider any of these failures in the automated  
test suite to be a showstopper for a release.

Below are my noted on all the FAILs.

-- 
olivier

* XHTML 1.1 Document Type support test
=> the DTD published with the 2007-02 WD of XHTML 1.1 had a typo.
Replacing it with the latest version from http://www.w3.org/MarkUp/DTD/xhtml11-flat.dtd
and the test passes

* XHTML 1.1 (as application/xhtml+xml) Document Type support test
=> ditto above, fixed

* SVG 1.1 Basic
Not a regression - same problem with 0.8.3 and before.
The problem is with the SVG DTDs, as reported in
http://lists.w3.org/Archives/Public/www-svg/2007May/0010.html
and explained in
http://lists.w3.org/Archives/Public/www-svg/2007May/0011.html
Someday hopefully the SVG WG may fix the DTDs.

* Test for warning about ampersand as data (in SGML)
This is a regression in the SGML parser. I remember our old parser  
used to throw a warning for this legal, but not recommended, practice  
in HTML. The more recent parser we are now using does not.

* Test of warning for non-HTML compatible XHTML document
... checking of HTML compat implemented but not enabled, as requested  
by XHTML WG

* bogus FPI #2: HTML 4.01 "Strict"
See the section "Documents that are invalid, which the validator  
passes as OK" in http://qa-dev.w3.org/wmvs/HEAD/dev/tests/
This is an old, strange bug:
  - document with a unknown FPI
  - the validator falls back to HTML4.01
  - validation succeeds


* Charset mismatch: HTTP Content-Type and XML Declaration (sends  
warning, proceed with HTTP and pass)
Fixed after fixing the XHTML 1.1 issue above

* DOCTYPE with a relative URI for the system identifier. Should  
probably pass if the sgml parser was given the base URI(?)
http://www.w3.org/Bugs/Public/show_bug.cgi?id=1521


* text/html, no charset, fbc set (W02)
Looks like an apache setup issue on the test server: the test case  
should be served with no charset, apache2 forces utf8...
Content-Type: text/html; charset=utf-8

* text/html, no charset, override set (W04)
Ditto above

* text/html, no charset (W04)
Ditto above

* Test for xmlns in HTML content
HTML 4.01 with xmlns attribute on root element. Should trigger a  
validation error (it does) and preparse warning W11 (it doesn't!)
This is a known bug, should be in Bugzilla.

*  charset mismatch HTTP / XML decl (W18)
Fixed after fixing the XHTML 1.1 issue above
Received on Monday, 17 November 2008 21:08:50 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 7 November 2012 14:18:19 GMT