Posted by cwdjrxyz on 11/30/07 17:19

On Nov 30, 10:34 am, Mark <u...@example.net> wrote:
> Check out this page:
>
> http://www.google.co.uk/intl/en/options/
>
> No doctype, table layout, presentation attributes, font tags...
>
> If somebody posted this page here for a critique it would be ripped to
> shreds.
>
> Run it through the W3C validator:
>
> ---8<---
> No DOCTYPE found! Attempting validation with HTML 4.01 Transitional.
> Failed validation, 120 Errors
> --->8---
>
> HTML has to be pretty bad to fail validation against HTML 4.01
> Transitional, and with no less than 120 errors.
>
> Is this also a clue that Google doesn't really care about the quality of
> HTML when ranking pages?

You are using the UK Google site. The page structure and code on the
US Google site is somewhat different. However, validate
http://groups.google.com/grphp?tab=wg . It has 558 validation errors.
The home page for Microsoft has over 30 validation errors. The home
pages for Firefox and Opera validate. Many sites of large corporations
have been up many years and have just been patched from time to time.
As a result they are a tag soup of many versions of html and sometimes
xhtml and will validate as nothing. They tend to get fixed only to
correct a problem produced by an upgrade of some browser, and the fix
is often the minimum that will get rid of the problem - not a complete
rewrite of the site in valid code. One could neary say, the larger the
corporation and the older their web site, the more validation errors
it has. I suspect that because the code on many sites is extremely
long and would require much time to update, the powers-that-be in the
corporation see no need to spend a lot of money to update the site to
modern code and take the attitude that the site should just be kept
working with the minimum cost for writing code.

[Back to original message]