Re: <A> content model in Cougar

Walter Ian Kaye (
Sat, 27 Jul 1996 19:40:29 -0700

Message-Id: <v03007803ae20821ce07a@[]>
Date: Sat, 27 Jul 1996 19:40:29 -0700
From: Walter Ian Kaye <>
Subject: Re: <A> content model in Cougar

At 9:10p +0200 07/27/96, Arnoud "Galactus" Engelfriet wrote:
>In article <>,
>Arne Knudson <> wrote:
>>      I've noticed that the Cougar DTD declares %A.content (the entity
>> describing the allowed content for the A element) as %text. The HTML 2.0
>> DTD, though, has the following declaration:
>Funny, I never noticed that until you pointed this out. And yet, all
>the times where I wrapped <A NAME> around headers, I got complaints
>from validators. Hm.
>> instances of HTML 2.0, but fails under HTML 3.2. Was it a conscious decision
>> to remove %heading from %A.content, or would it be wiser to make HTML 3.2
>> backwards compatible by using the same defininition (with %HTML.Recommended
>> and a default %A.content) as in the HTML 2.0 DTD?
>I think that since A is regarded as text-level markup, the heading would
>terminate the current block element, and the anchor inside it as well.
>IOW, A *can't* contain headers.

Hmm. I would think that *any* displayed text/heading/image is a
candidate for a hyperlink target or anchor. And since <A> is wrapped
around non-containers, it should be wrappable around containers (such
as headings) as well. Is there some reason someone decided it shouldn't
be? I'd like to know the reason...

    Walter Ian Kaye <>     Programmer - Excel, AppleScript,
          Mountain View, CA                         ProTERM, FoxPro, HTML     Musician - Guitarist, Songwriter