|
|
Posted by Andy Dingley on 11/21/07 12:47
On 20 Nov, 12:51, The Magpie <use...@pigsinspace.co.uk> wrote:
> >> Personally, I prefer XHTML 1.0 over HTML 4.x because it is much easier
> >> to ensure it is error-free. But that's a personal choice.
>
> > What makes you think that?
>
> I should have thought that was obvious.
Obviously not.
> HTML was designed from the start as a subset of SGML that would permit
> lax coding of tags
There's nothing "lax" about SGML's permitted optimisations. You're
allowed what you're allowed, no more. The fact that it's famously
difficult for humans (~Jukka) to understand the rules relating
optimised tags to elements doesn't mean they're "lax", just complex.
In fact, HTML is considerably _less_ "lax" than SGML might otherwise
have permitted.
If you said that it was harder to manually _author_ correct HTML than
correct XHTML, I'd have agreed with you. However to validate it as
being error-free, you do this to either of them by applying a pre-
existing freely available validator to them. For you as an author,
this is equally difficult for either.
> and allow addition of non-specified tags
HTML doesn't permit this. The fact that many people _chose_ to, and
that some chose to do it in a manner (by specifying a new and extended
DTD) that was valid SGML (although thus no longer HTML at all) doesn't
make it right.
> XHTML is - of course - simply XML.
Is it? All that can be said today with any real confidence about
"XHTML on the web" is that you can't do it with "XHTML as XML" because
IE refuses it.
Now if XHTML is XML, we're allowed XML namespacing and so we can
(directly contradicting your claim) safely add extra elements or
attributes. However in today's world of "XHTML as XHTML" we can't do
this with any more validity then simply dropping them into HTML as in
the past.
Navigation:
[Reply to this message]
|