RE: XHTML

Hi George,

	Apparently there were several points in my talk (on CSS no less)
that were gravely misunderstood. Hopefully 
this message will correct this. I am speaking for myself here, not W3C or
CNET, as usual. :)

	I can't comment on the social value of HTML and its ease of use;
while I personally believe that this is a Good Thing (TM) I certainly can't
establish it as a main priority for W3C! I have heard W3C people speak many
times on this and they generally seem to share your opinion also. I can
vouch for the fact that the HTML WG has certainly taken this into account in
our work.

	As far as attribute values in XHTML; somehow my words on this have
been brutally mangled, for which I accept all responsibility. One is not
required to list all possible attributes for every tag; I think the OBJECT
tag has 36 attributes, some of which are mutually exclusive! What I was
trying to get at was that along with tag minimization being forbidden,
attribute minimization is forbidden also. I think I used this example:

the current state of affairs allows users to use the COMPACT attribute to
the UL element like so:

<UL COMPACT>

in XML we cannot use this sort of attribute minimization; we must say:

<ul compact="compact"> 

I was using this as an example of the sort of changes in XHTML that will
drive users mad.

Whew! Hopefully that is cleared up. 

	However, having said that, I will certainly vouch for my comments re
HTML being similar to Postscript. As you noted, HTML is getting more
complicated all the time, and harder to write. I am convinced that document
authors cannot achieve the sort of power and flexibility they desire in a
browser markup language and still maintain the simplicity of pervious
versions of HTML. This tradeoff between power and simplicity is common to
all computing systems. 
	My belief is that it is better to create the kind of markup language
people want, with the features they need to accomplish their goals, even if
it adds complexity. Vendors will provide tools that will hide the
unnecessary complexity from users. This is where the similarity to
Postscript comes in; users do not need to edit their Postscript documents at
the directive level, and HTML authors should not have to edit their
documents at the tag level; tools should be created that hide the messy
details, of interest only to machinery, from the user. Here is an
interesting paper I wrote on just this subject:

http://www.cnet.com/Papers/W3C/kungpao.html

Hope this clears things up, and let me know if you have other questions!

Regards,

D-

> -----Original Message-----
> From: George Woolley [mailto:george@metaart.org]
> Sent: Tuesday, March 16, 1999 12:32 PM
> To: www-html@w3.org
> Subject: Re: XHTML
> 
> 
> Currently the basics of HTML
> (sufficient to create an information oriented site),
> can be taught in less than an hour.
> There's not a lot of clutter
> and contrast can be used to increase clarity.
> And there is no need to buy a special editor to get started.
> This ease of entry for those who want it
> is a major social value of HTML as it currently exists.
> 
> XHTML 1.0 as specified significantly undermines
> the ease of learning and reading HTML.
> We'll have a lot more tags in our code
> and a lot less possibility of contrasts.
> I was at the talk by Daniel Austin
> (that Dan Meriwether referred to)
> where he told us of the requirement
> for including every attribute in an element.
> If he is correct, HTML code would (in my opinion)
> become a sort of Dilbert like sick joke.
> 
> Some people don't get why anybody cares.
> Indeed, at the same talk Daniel Austin said
> that we should think of HTML
> as having a similar role to PostScript.
> And since most of us don't expect PostScript
> to be easily readable,
> we should not expect HTML to be either.
> Silly me, I thought HTML purported to be
> a mark up language for humans (among other things).
> Why am I concerned?
> Because a major social value of HTML
> is in the process of being destroyed.
> 
> QUESTIONS
> 1) Could someone let Daniel Austin know
>    that his talk has been referred to here twice now?
>    Or alternatively, could someone send me
>    his e-mail address so I can let him know?
> 2) Could someone from W3C
>    comment on whether they are committed
>    to the social value referred to above?
> 3) Does anyone know of a specification
>    (actual or under consideration)
>    that involves anything like the proliferation of attributes
>    that Daniel Austin talked about?
> 
> =====================================================
> >At 17:29 15/03/99 -0500, Dan Meriwether wrote:
> >>For the sake of readability, by humans, let the XHTML 
> elements be all caps.
> >>
> >
> >There are text editors that can present markup in an 
> highlightned way,
> >usually using a different color.
> >
> >>It is also somewhat bizarre to specify that the attribute 
> values must also
> >>be in lower case. What about SRC and HREF attributes? Are 
> you saying that
> >>my entire directory structure now needs to be in lower case?
> >
> >No. this applies to attribute values that are case 
> insensitive in SGML DTD,
> >and not to those those that are case sensitive.
> >In SGMLish HTML you can have Align="Center" but in XHTML
> >it must be align="center"
> >
> >URIs are case sensitive, so is the alt attribute of <img> etc. and
> >they may have mixed case. BUT they also have different meaning when
> >they differ only by case.
> >
> >>
> >>Recently Daniel Austin mentioned, quite strenuously, that 
> very element now
> >>must have every attribute available to it declared within 
> it. In other
> >>words, when I specify a <TABLE> all 9 attributes must be present,...
> >
> >I think that this is not true.
> >
> >Regards,
> >
> >
> >
> >Nir Dagan
> >
> >http://www.nirdagan.com
> >mailto:nir@nirdagan.com
> >tel:+972-2-588-3143
> >
> >"There is nothing quite so practical as a good theory."
> >-- A. Einstein
> 
> 
>

Received on Tuesday, 16 March 1999 17:02:30 UTC