[XSL-LIST Mailing List Archive Home] [By Thread] [By Date]

Re: HTML is a formatting/UI language was: RE: Formatting Obj


Subject: Re: HTML is a formatting/UI language was: RE: Formatting Obj
From: Guy_Murphy@xxxxxxxxxx
Date: Tue, 20 Apr 1999 17:54:11 +0100

Hi Paul.

On the first point, all I meant was that people are in the main looking at
the rendered result not the code that specified the rendering, as this was
suggested to be a problem.

On the second point, if one of the objectives of XML and it's related
technologies is human readability then we are concerned with semantics.
HTML strips away the ability to easily deploy semantics suitable for the
domain. With FOs we have a chance to specify formatting for rendering, so
that
presentation can be truly seperated from document/data semantics. And to
repeat this means that the FOs can express semantics relevent to
presentation whether aural or visual.

Lets get real here a minute....

We've had HTML for how many years now? And we have also had text to speak
for a good number of years also. Anybody seen during this time a rush by
the Web community to support aural presentation of Web pages?

At present even if we used something like MS Agent from text to speach,
implimenting anything beyond simple text-reading would be a nightmare. Yes
it could read the text in a literal fashion but the moment you come to
giving expression to the actual HTML tag set your screwed.

At a simple level, HTML is so loose, unclear and unspecific (I was tempted
to simply say crap) that there are 100000000 and 1 mish-mashed pages
pulling all sorts of dirty tricks to render they're page. Think that H2
tags should be delivered in a certian way? It would require the dangerous
assumption that authors use H2 tags correctly.

Hakon then talks about being able to carry over data semantics into
presentation using DIV classes.... just HOW are you going to interpret that
auraly.

I'm sorry, I cannot regard the consideration of HTML for the future of Web
development, and the labeling of FOs as dangerous as anything more than a
red herring. I think the posturing over aural presentation is simply that:
posturing as I believe that HTML is as bad or if not worse for describing
aural presentation as it is page formatting.

Now come up with a namespace to sit alongside FO for aural presentation and
it would take a hacker one week to build a HTC to extend the feature set of
IE5 to parse and present this namespace in an intelligable fashion...
*then* we might actualy see Web content delivered in a meaningful fashion
using XSL stylsheets to transform either on an individual media basis into
this namespace, or this namespace sat along-side visual formatting.

I firmly belive that accessibility will only be advanced in a meaningful
way when we give developers the means to address it meaningfully. The
suggestion that domain specific presentation delivered from XSL is
dangerous and that we're safer in the paper boat that is HTML is laughable.

The best way to specify aural presentation is.... ummm to specify it *not*
infer it.

Go surf the Net. Take a look at the HTML being used out there..... then
come back and tell me what's dangerous and semantically unintelligable.

[I truly find it hard to credit that we're going over HTML again at this
stage of the game]

Cheers
     Guy.





xsl-list@xxxxxxxxxxxxxxxx on 04/20/99 06:04:33 PM

To:   xsl-list@xxxxxxxxxxxxxxxx
cc:    (bcc: Guy Murphy/UK/MAID)
Subject:  Re: HTML is a formatting/UI language was: RE: Formatting Obj




Guy_Murphy@xxxxxxxxxx wrote:
[SNIP]

> As a side note.... most people browsing the Web are most concerned with
the
> appearence of the Web app they are using and the functionality
provided....
> not the source code used to render it, and I beleive that it is not worth
> while sacrificing the functionality provided by clean domain specific FOs
> (or other media specific objects), so that people have a better time of
> viewing source. If one is that concerned about seeing the original data
one
> can go fetch it.
I'm not sure what this all means. The source is supposed to be sent across
and transformed on the client side, right?
> It is also easy to forget that we will have XHTML, and we already have
> XML+CSS, and nobody is suggesting that XSL is a direct competitor for
these
> across all solutions. It is likely that most Web sites will present in
> either XHTML or XML+CSS. If presentation semantic is required as part of
a
> solution there are already means for implimenting this. If however this
is
> either not a concern, or the richist presentational expression possible
is
> a requisite, then I feel that FOs are absolutely the way to go, not a
> watered down formatting semantic.
Can you please be precise on what formatting objects give you that HTML
does not and cannot?
--
 Paul Prescod  - ISOGEN Consulting Engineer speaking for only himself
 http://itrc.uwaterloo.ca/~papresco
[SNIP]




 XSL-List info and archive:  http://www.mulberrytech.com/xsl/xsl-list



Current Thread
Keywords