plan for test clean up

I see the brief for my action to fix up some tests as limited to those 
with obvious problems: i.e. they don't work for some reason or other, or 
do not conform to the spec.

I believe these tests to be:
- testlist2.html
- sq1ns.xml
- sq1.xml
- loop.xml
- the grddonrdf test with xmlns="??"

Also the two content negotiation tests:
   conneg
and
  multipleRepresentations

should both have three output files:
   one for each represetation
and
   one being the merge of the other two

Any one of these outputs conforms to the spec.

I will create these output files, and add them to the site.

===

I will start by e-mailing current copies of affected files to 
www-archive, to keep a record more easily accessible than the cvs 
history, permitting better review of my work.

I will change the files, and then report back to the WG as to what I 
have done.
I expect to do this tomorrow.

===

Other things that I think should be done, but I don't think we yet have 
consensus on are:
a) copy any document from the web (profile, transform, etc) into the 
test directory, except for:
    - the library functions
    - the GRDDL namespace document
    - the XHTML namespace document
b) update all tests to refer to the local copies not the originals not 
under our direct control.
c) whereever possible have XHTML doctype declarations and check html 
documents with the validator
d) add comments to all test files indicating license terms and 
copyright. I assume the license terms are the W3C software license terms 
for xsl files, and W3C document license terms for the other files

I am happy to do any of these, given the go ahead by the chair.



Jeremy

Received on Thursday, 15 March 2007 14:24:49 UTC