W3C home > Mailing lists > Public > www-validator@w3.org > February 2001

Re: Table Validation

From: Terje Bless <link@tss.no>
Date: Tue, 27 Feb 2001 05:50:50 +0100
To: Marty Tillinger <marty@tillinger.com>
cc: www-validator@w3.org
Message-ID: <20010227082742-r01010600-6a19ad45@>
On 21.02.01 at 19:56, Marty Tillinger <marty@tillinger.com> wrote:

>I've used http://validator.w3.org/ to help me get web pages right, but 
>found some cases where there were display problems, but the validator 
>showed no problems.

The Validator doesn't care about "display"; it has no concept of the term.
What it does do is check the syntax strictly and accurately against the
defined syntax that is valid for (X)HTML.

>I then found  www.doctor-html.com. It checks for tables to have matching
>elements with proper nesting. td /td, tr, /tr, etc.It reports on structure
>much better as well as browser compatibility.

A particular browser vendor's implementation has nothing to do with
syntactic valididy. In this particular case, what's going on is that TD
elements are getting implicitly closed when another element is seen, which
may or may not be where you would have placed the TD had you done so

>However, I'm not sure it is as standards-based as your validator.

It's not. It merely checks against some random programmers idea of "good
HTML" (possibly taking into account known deficiencies in browser
implementations) and not the actual valid syntax of the language. You'll
want to start by making sure somethig is Valid -- by checking with the W3C
Validator -- then running it through something like Dr.HTML (including
looking at it in various browsers etc.), and finally going back to the W3C
Validator to make sure you haven't introduced any syntax errors in the

>I'd like to see you add the checking of matching end tags and the proper
>nesting of tags as part of the validator.

Sorry. We can't do that. It's perfectly legitimate to leave off closing
tags in some cases -- in which case a conforming parser wil infer the
position of the closing tags and sometimes get it wrong -- and trying to
add such a feature would impose _our_ arbitrary standards of "good" HTML
without a solid, objective, measure to compare it to.

>It also checks for missing images.

This, OTOH, could be added in some future version (it's on the TODO).
Received on Tuesday, 27 February 2001 02:27:51 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:29 UTC