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

Re: Validator Concerns

From: Terje Bless <link@tss.no>
Date: Tue, 28 Sep 1999 20:39:36 +0200
Message-Id: <199909281859.UAA03292@vals.intramed.rito.no>
To: W3C Validator <www-validator@w3.org>
cc: Christian Ziebarth <ChristianZ@netscape.net>
On 27.09.99 at 20:57, Christian Ziebarth <ChristianZ@netscape.net> wrote:

>It seems like the validator is outdated.

Well, yes, it is somewhat dated, but not for the reasons you think.

>It tells me that there is no such thing as a NORESIZE element or a LOWSRC

Which DTD are you validating against? You /did/ place an appropriate
DOCTYPE definition at the beginning of your HTML file?

BTW, those are "attributes" to the FRAME and IMG "elements".

>yet I use both of those and they both work.

How do you define "work"? That some dubious browser implementation renders
your markup in some seemingly reasonable way is by no means an indication
of the validity of the markup; merely that the HTML is broken in a way that
is agreeable to the browser's rendering engine.

>Also, it wouldn't allow me to specify all the text as being bold at the
>beginning of the document and then close the "bold-ness" at the very end
>of the document.

I can't say for sure without looking at the document in question, but that
sounds perfectly reasonable. It would be a strange document indeed where
that would not be a format violation. Since there are rules for what
context elements can be used in, what you describe is exceedingly likely to
be broken.

>I also got marked down for ads that GeoCities puts on my page when I have
>no control over their HTML coding.

Then maybe you shouldn't be using GeoCities? I, for one, wouldn't trust an
ISP that can't even get that little snipped of HTML correctly. This ain't
rocket science, you know?

>These are a few of my concerns with the validator.

If you have more concerns (especially after reading below), please do bring
them to this list. Feedback is always good!

>It seems there were other elements it told me didn't even exist that I use
>and that work for me. I didn't see any of these concerns addressed in any
>of the links provided. Thanks for your time.

You were looking for the wrong thing. You are basing your definition of
"correct" and "working" HTML on what you see in a particular browser or
browsers. This is, at best, inaccurate. A particular variant of HTML is
formally specified in a DTD. It defines each element, what it can contain,
and where it can appear. The validator takes your document and compares it
to the formal definition in the DTD. It then reports any deviations.

You'll note that at no point was "display" or "look" used in the above
explanation. How things look in a browser is incidental and irelevant.

If you use the MSIE "extension" MARQUEE, the validator will flag that as an
error, even if it works just fine in MSIE, because there is no DTD that
contains that element. It's just something that Microsoft dreamed up on
their own.

Similarly, if you don't specify in your HTML file which DTD this file is
supposed to conform to (the DOCTYPE declaration at the top of the file), or
if you specify the wrong DTD, then any element you use that is not
specified in the referenced DTD will be flagged as an error.

>From your mention of the LOWSRC attribute above I'm guessing that you've
given no DOCTYPE or the wrong one. Most of your errors will be fixed by
using the HTML 4.0 Transitional DOCTYPE (this is explained on the validator
web site). The remaining ones (notably, your <B> problem above) should be
far easier to fix using the validator afterwards.
Received on Tuesday, 28 September 1999 15:00:09 UTC

This archive was generated by hypermail 2.3.1 : Tuesday, 1 March 2016 14:17:25 UTC