Next message: Bjoern Hoehrmann: "Re: DTD problem, or validator problem?"
From: Bjoern Hoehrmann <derhoermi@gmx.net>
To: Yuri Generalov <yuri@cactus-books.com>
Cc: www-validator@w3.org
Date: Thu, 16 Aug 2001 23:17:47 +0200
Message-ID: <igdont8a91bvvr1e79q0cd1i1ksq6cqfg7@4ax.com>
Subject: Re: when using validator is not sensible
* Yuri Generalov wrote:
>I am trying to use valid HTML (or XHTML) on all sites I build. However, I more
>and more doubt if I should continue using it.
>
>For me, the process of building valid code is largely a struggle between two
>things which sometimes can not exist together: valid code and interoperable
>web page.
You start with 100% interoperability with an HTML 4 document; even HTML
2.0 user agents are able to render the document in an appropriate manner
since HTML 4 was designed to be downwards compatible with existing
browsers. You _reduce_ interoperability when writing invalid code; you
rely on the error recovering behaivour of the implementations you have
tested your page with. I recommend reading through the archives of
related mailing lists or newgroups, you'll see a lot of people
complaining on what NN6/IE6/Opera5 make of their pages that "worked" in
horribly broken NN4 or alike. That was caused each time by writing
invalid code.
>No, this is not a mistake. In my experience, using validator to check
>for code validity and picking out all of invalid code meant loosing interoperability
>and finally, when validator said "Congratulations..." my webpages showed up
>differently in different browsers.
Where's the problem? If there is a problem, I doubt you've written
"good" HTML, even though it might be valid, but feel free to point
at examples.
--
Björn Höhrmann { mailto:bjoern@hoehrmann.de } http://www.bjoernsworld.de
am Badedeich 7 } Telefon: +49(0)4667/981028 { http://bjoern.hoehrmann.de
25899 Dagebüll { PGP Pub. KeyID: 0xA4357E78 } http://www.learn.to/quote/