W3C home > Mailing lists > Public > www-validator@w3.org > October 1999

Re: W3C supposed to have something to do with STANDARDS

From: Terje Bless <link@tss.no>
Date: Tue, 5 Oct 1999 15:17:09 +0200
Message-Id: <199910051322.PAA06997@vals.intramed.rito.no>
To: W3C Validator <www-validator@w3.org>
On 04.10.99 at 13:36, Uriel Wittenberg <uw@urielw.com> wrote:

>Terje Bless wrote:
>
>>On 29.09.99 at 06:36, Uriel Wittenberg <uw@urielw.com> wrote:
>>
>>>The HTML standards body should not be approving HTML that's been run
>>>through any arbitrary validator.
>>
>>They aren't! If /you/ use the badge on a page then /you/ are asserting
>>compliance with the relevant DTD (which, for HTML 4.0, happens to come
>>from the W3C). What it is that makes you think you are justified in your
>>assertion is largely irrelevant. The W3C makes no such assertion.
>
>Where do you get this information? The distinct impression given by the
>W3C site is that use of the logo reflects _their_ approval.

Ok. We are both operating from "impression given by the W3C site". I'm
applying logic that says it would be meaningless for the badge to assert
anything but compliance with a standard. It seems you feel it must mean
explicit endorsement by the W3C.

I'm not sure these two positions are reconcilable.


>The definition of what's approvable is not made clear by the site. (I
>haven't made a research project out of this.) But I believe you're not
>correct -- I think it's more than adherence to the DTD, since e.g.
>!DOCTYPE is required. (Correct me if I'm mistaken.)

I'm not sure I understand this paragraph fully.

The DOCTYPE needs to be present because that is what specifies which DTD
should be used to interpret this document. The DOCTYPE can reference any
DTD; including several that have not been specified by the W3C. The DOCTYPE
can even reference your own DTD using a URL and the SGML parser that the
Validator uses can fetch it from your site and validate the pages against
that DTD.

If you use the "Netscape-HTML" or "Hotjava-HTML" (from Netscape and Sun;
DTD by WebTechs IIRC), the badges will not indicate W3C anymore because
those standards were not created by the W3C. If you use another DTD, the
validator will tell you that it's valid, but it doesn't have a badge for
that type yet.


>>If you want the W3C to give absolute assurances as to the validity of a
>>page,
>
>Please. Whoever said such a thing? This is a caricature of my position.

I'm sorry. My mistake.

If the badge indicates approval by the W3C, then the W3C needs to make sure
that the page is in fact valid for the badge to have any meaning. Since
this is a practical impossibility, the only reasonable assumption is that
the W3C asserts nothing but the owner of the page asserts compliance. This
places the burden of proof on the page developer; where it belongs IMO.



>>Do not hire a web design shop that do not produce valid HTML. Do not
>>produce anything but valid HTML even if it means the Fortune 500 company
>>that wants them will go to a different web design shop.
>
>Really, appealing to altruism and asking people to forgo money is not the
>way to promote standards.

I'm not appealing to altruism; I'm appealing to professionalism. If you
work in construction, you can and will be criminally prosecuted if you
knowingly use poor materials and/or do shoddy work. The fact that you did
it because your client asked you to is not even considered extenuating
circumstance.

If a client asks for web development work that not only violates standards
but is also a violation of good practice; you are in a moral equivalent of
the situation in the construction business. The consequences are far less
severe (I feel like I'm invoking Godwin's Law here ;D), but the principle
is the same.

A more pragmatic attitude is that some day you'll run up against a a client
that knows what s/he's about; that will look at the non-standardscompliant
work you've done before and decide that you aren't good enough.


>>>The W3C logo should only be permitted -- and is only permitted, given a
>>>sensible reading of the site -- after W3C-endorsed validation.
>>
>>Well, the "W3C" has spoken on this so I guess there isn't much point in
>>you telling them that they are wrong, is there?
>
>I think there's been some mistake and some clarification or discussion is
>still needed -- for the sake of our common objective of promoting
>standards.

Yes, of course. Pardon the sloppy way I expressed myself.

I meant that your statement "is only permitted" above was contrary to what
Gerald wrote in a message on this list. Disagreement of how things /should/
work is one thing, but saying that when the W3C says A they really mean B,
is pointless.


>>Please have the common courtesy to subscribe to the lists you post to!
>
>For you, Terje, I've resubscribed (for a while).

Thank you. It's frustrating to converse with someone off-list because they
miss a lot of context that I see.


>I think it'd be reasonable to say the badge is permitted if the author,
>based on reasonable experience, is confident it *would* pass the W3C
>Validator.

And you want the text on validator.w3.org to spell out this intention to
make it more clear?


>>Do you want the W3C to maintain a list of "approved" validators? That
>>would be a bit much to ask, IMO, as the work involved would be
>>significant.
>
>I don't agree. They're not obligated to evaluate each applicant. If all
>the experts on this list feel WDG's ok then based on that W3C could list
>the WDG service on its webpage.

Well, the practicalities of the situation are such that "validator.w3.org"
== "Gerald" AFAIK. He's pressed enough for time without adding additional
work to his load. The "experts" on this list can opine all they want and it
won't change the fact that in the end it's Gerald that is responsible.
Pointing the finger and saying "Liam made me do it" is not good enough. If
the site lists other validators that are ok, then Gerald really would have
to evaluate them first. While not insurmountable, it is a non-trivial
amount of work.


>If the W3C wants to get into it, they could set up a service that pays for
>itself whereby there's a fee and a formal agreement.

Yes, but do you really think such a service would be popular?

There are any number of such schemes that the W3C /could/ implement, but I
can think of none that would be both economically feasible _and_ popular.
Cheap services would be economically infeasible and expensive services
would not be used. How many individuals do you know who are W3C Affiliate
Members ($5K/year)?
Received on Tuesday, 5 October 1999 09:22:35 GMT

This archive was generated by hypermail 2.2.0+W3C-0.50 : Wednesday, 25 April 2012 12:13:52 GMT