W3C home > Mailing lists > Public > w3c-wai-gl@w3.org > October to December 2005

RE: Validity

From: Paul Walsh <paul.walsh@segalamtest.com>
Date: Sat, 5 Nov 2005 17:55:36 -0000
To: "'Roberto Scano (IWA/HWG)'" <rscano@iwa-italy.org>, <r.castaldo@iol.it>, <w3c-wai-gl@w3.org>
Message-ID: <000501c5e232$207f8f40$0200a8c0@PaulLaptop>

Ok, let me make myself clear for the sake of the record.

Today, certification is given upon the successful completion of audits
based on the guidelines. These all pass, including validity. I'm in
favour of validity for all the reasons stated by others, I just don't
think it's necessary where a site is accessible to all, but still
contains 'some' invalid code that's too costly or too time consuming to
put right.

The first thing we check for is code validity. All tests end up passing
by the time certification is given. However, 'editors' with little to no
coding experience, using a CMS to make updates, usually introduce some
invalid code, yet everything remains entirely accessible and usable to
assistive technologies. 

(That's not to say that some sites don't end up inaccessible at some

This very debate is one of the reasons we are moving away from the three
categories.  We are more interested in providing personalise search
capabilities through the use of machine readable content labels using
RDF-CL. This will enable users to choose the type of accessibility
implementations that are important to them. I encourage the WAI to be
used as the basis for best practise design, but from this week, will no
longer use it as a commercial means of justifying accessibility. The
self-regulated content labelling scheme will be officially launched next

I remember a recent email asking if there was such a scheme available,
so I'll give more information when replying to that email. To give you a
taste, we're also due to launch a new browser in December with built-in
search annotation and rating capability, allowing users to rate a
trustmark. Alas, self-regulation that works to prove the concept of
personalised search!

To summarise, I'm in favour of validity, but I wouldn't like it to be
seen as a defacto checkpoint. I think it's great to help with
accessibility but it ends up on the bottom of my list if all else equals
accessibility anyway.

Hope that helps to clarify my position :)


      -----Original Message-----
      From: Roberto Scano (IWA/HWG) [mailto:rscano@iwa-italy.org]
      Sent: 05 November 2005 17:30
      To: paul.walsh@segalamtest.com; r.castaldo@iol.it;
      Subject: RE: Validity
      How can they conform to double A if code is not valid? (wcag 1.0
checkpoint 3.2).
      If you certificate these web sites for wcag conformance and don't
check for
      validity, you grant a certification for level AA that is not true.
      ----- Messaggio originale -----
          Da: "Paul Walsh"<paul.walsh@segalamtest.com>
          Inviato: 05/11/05 18.11.49
          A: "'Roberto Castaldo'"<r.castaldo@iol.it>,
          Oggetto: RE: Validity
          Firstly, why do you amongst others, feel that the only
          guidelines are those that are done using tools?
          We certify sites every week and if we were to failed them as a
result of
          invalid code, the vast majority of them would fail - even
though they
          meet all Double-A and two checkpoints in Treble-AAA. I hope
you're not
          trying to tell me that these sites should fail the basic level
          accessibility just because they contain invalid code? If you
are, then
          you don't live in the real world. I will reiterate,
introducing validity
          to the lowest level of conformance (whilst ignore the fact
that a site
          can be accessible) will alienate people from using the WAI, me
          What's important in all of this is that machine readable
labels can
          cover guidelines that aren't necessarily categorised, so users
          choose the most important ones for their requirements.
          Please refrain from telling me to stop using a specific reason
for my
          side of the argument. I've seen this debate go on for long
enough and am
          aware that it has been used. Common sense tells me that if a
site meets
          Double-A standards and is very user friendly, you can't fail
it just
          because the code doesn't pass a test using a tool.
          Kind regards,
                -----Original Message-----
                From: w3c-wai-gl-request@w3.org
                Behalf Of Roberto Castaldo
                Sent: 05 November 2005 16:37
                To: w3c-wai-gl@w3.org
                Subject: Re: Validity
                Hi group,
                i do not like any action in favour of tag soup, and I'm
afraid we
          risk to do
                that; the  problem, the real problem about tag soup, is
that none
          of us is
                really able to define what an hypothetical "good tag
soup" may be;
          so it's
                not possible to write a single guideline, or a tecnique
that says
          how to
                write good code without validating it.
                Maurizio and Paul, we all know that a valid page can
have terrible
          code (but
                none has never said it, so please stop using this
argument), so
          valid code
                by itself is not enough, but it is one of the few
anchors, one of
          the few
                impartial and objective milestones which is testable by
          and that
                gives the best interoperability guarantee. That's why
valid code
          is the
                best starting point for any web based project, and
cannot be other
          than L1.
                And what about the W3C compliant authoring tools
shortage? That's
          the actual
      [Messaggio troncato. Toccare Modifica->Segna per il download per
recuperare la
      restante parte.]
Received on Saturday, 5 November 2005 17:55:32 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 16 January 2018 15:33:57 UTC