W3C home > Mailing lists > Public > www-qa@w3.org > May 2003

Re: Dogfood... (was: MathML 2.0, second edition)

From: Lofton Henderson <lofton@rockynet.com>
Date: Fri, 30 May 2003 11:33:39 -0600
Message-Id: <5.1.0.14.2.20030530112241.02e24270@rockynet.com>
To: Terje Bless <link@pobox.com>, W3C QA IG <www-qa@w3.org>

At 07:17 PM 5/30/03 +0200, Terje Bless wrote:

>Lofton Henderson <lofton@rockynet.com> wrote:
>
> >Ref:  http://www.w3.org/QA/WG/2003/05/SpecGL-MathML20-review.html
>
><http://validator.w3.org/check?uri=http://www.w3.org/QA/WG/2003/05/SpecGL-
>MathML20-review.html>
>
>70 errors? Most of which are banal HTML syntax in XHTML or bad nesting of
>elements... For shame! :-)

Oops!  Our bad.

Explanation (but no excuse):  this got handed down the line through several 
people, and in haste we sent it without giving it a final check.

>What can the Validator do to make it easier to integrate validation into the
>publishing workflow for QA in general? Or perhaps specifically for SpecGL
>reviews?

Validation happens as a consequence of any of our WD publications, by the 
requirements of pubrules.

But it sometimes gets overlooked in hastily processed "internal" 
documents.  (Which, as in this case, become "external".)

What to do to make it easier to avoid such lapses?

Let's see, I was at the end of a chain of several people, and used jigedit 
to post this with Web Commander.  If it were somehow possible to get a 
validation notification at time of posting (by Web Comm'dr, or other 
tools), then that would relieve me of having to remember every time, "oh 
yeah, has anyone validated it yet?".  It is that extra, voluntary step in 
the workflow that is problematic.

-Lofton.
Received on Friday, 30 May 2003 13:33:35 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 6 January 2015 20:40:32 UTC