xml-dev - RE: XML for 3D geometry and objects

[Tagore Smith <tagore@internexus.net>]

On Fri, 21 Jan 2000, Bill dehOra wrote:

> 
> :  I forgot to mention that Macromedia were part of the VML 
> :  proposal, so I
> :  guess we'll see VML in their products too.  I wonder what 
> :  this'll mean for
> :  Shockwave & Flash.  
> 
> Certain death I hope. Which changes the 80/90% figure greatly. All the
> companies involved in VML are also involved in authoring the SVG drafts. 3
> of the 9 VML authors are also authoring SVG, if my count is correct. Sounds
> like spread betting. Can you animate in VML?
> 

   This week one of my clients pointed me at www.cartier.com and told me
he wanted a site like that. I understand why he likes the site it's Cool
with a capital C. It's also very poorly designed IMHO, and while part of
this is the fault of the designer, some of it is caused by inherent
problems with Flash. Since the site is designed as one big movie the back
button on your browser is broken, you can't bookmark within the site, and
you can't link to part of the site. As well, no search engine can index
it.

   It seems like having a markup based vector language could solve a lot
of these problems; for that matter making Flash a HyTime application would
be a good start. But there are some subtle but very important problems
that don't go away easily.

   What are the advantages of using xml (there had better be some,
because there are some serious drawbacks in terms of complexity). Correct
me if I'm wrong but it seems to me that some of the big ones are:

1. xml can become a lingua france for defining the syntax of languages
exchanged across the internet, allowing for a (hopefully high) degree of
interoperability

2. Because xml allows you to embed a lot of information about the
structure of a document it is easier to define the semantics of specific
xml applications in a way that is recognizable by computers. This is only
possible because of a separation of presentation and content.

3. given 1 and 2 above, the web (really the tools that present, index,
etc. the web) can become much smarter. The standard example seems to be
"find me the best price out there on a pair of blue jeans".

   OK, so this is basic stuff- we all know this. But the question is how
can you define a Flash-like mechanism that preserves these advantages. An
example- I want to build a site as one big animated vector movie. (Or for
that matter I want my entire site to be one big navigable piece of
streaming video, which presents even worse difficulties). You can tell me
it's a bad idea until you're blue in the face, but if I'm like most of my
clients it's not going to do any good. 

   So, no problem here, right- since this is xml based, we can link to
specific frames in my movie. And all that cool scalable vector text will
be encoded as unicode, right? We can use nested elements or attributes to
encode both the semantics and the presentation of the visual elements. We
could use a separate document to describe the presentation of this
document. This is difficult, but not impossible. But of course, this all
depends on the makers of authoring tools making good decisions as
well. Hmm.. Adobe GoLive... every other "WYSIWIG" html authoring
program... pardon me if I'm not so hopeful. But still it seems that we
could define a scalable vector language in a way that preserved the
advantages of xml. Particularly important would be addressing- I want an
element <product>Blue Jeans</product> in my xml document but I also want
to be able to animate the letters independently. Again HyTime has some
useful things to say about this.

   But SVG doesn't even make an attempt to do this. I haven't read the
whole draft, but it's pretty clear that there are going to be real
problems indexing content. Mainly because my previous example is now
<text>Blue Jeans</text> (some attributes snipped). And if I want to put
the two words on different lines it looks to me like I'm going to have to
do <text>Blue</text> <text>Jeans</text>. And if I want to animate each
letter separately?

   Of course I could use XSLT and maybe some other processing to generate
my SVG. Do you think it likely that that's how authoring programs are
going to work? I'm pretty sure that if SVG becomes mainstream it's going
to be created in WYSIWIG authoring tools, and the chances that
<text>Blue</text> and <text>Jeans</text> are even going to wind up near
each other is pretty slim. 

   OK, so for the advantages of xml to really be seen you need good
authoring practices, and there are probably ways around some of these
problems- maybe using RDF and content negotiation. But I still have  
qualms about it.

   I'm sorry if this is too off-topic. I think that these issues are going
to be very important to the development of xml and the web, and in more
contexts than just this one. 

						Tagore Smith

    

 




xml-dev: A list for W3C XML Developers. To post, mailto:xml-dev@ic.ac.uk
Archived as: http://www.lists.ic.ac.uk/hypermail/xml-dev/ or CD-ROM/ISBN 981-02-3594-1
Unsubscribe by posting to majordom@ic.ac.uk the message
unsubscribe xml-dev  (or)
unsubscribe xml-dev your-subscribed-email@your-subscribed-address

Please note: New list subscriptions now closed in preparation for transfer to OASIS.