W3C home > Mailing lists > Public > www-svg@w3.org > March 2003

Re: Binary-coded version of SVG?

From: Chris Lilley <chris@w3.org>
Date: Wed, 12 Mar 2003 21:12:53 +0100
Message-ID: <15816139343.20030312211253@w3.org>
To: www-svg@w3.org, "Tayeb Lemlouma" <Tayeb.Lemlouma@inrialpes.fr>
CC: robin.berjon@expway.fr, "Double Ye" <iamdoubleye@yahoo.com.cn>

On Tuesday, March 11, 2003, 6:36:14 PM, Tayeb wrote:


TL> Hi,

>> That's one use case.

TL> Right, I specified that use case to answer to the message of  Double
TL> Y.(http://lists.w3.org/Archives/Public/www-svg/2003Feb/0055.html).

Yes, but how do you know that this is the use case that Double Ye was
thinking of? My original question was to find out which particular
reason for a binary form prompted the original comment.

There can be many reasons, see for example
http://lists.w3.org/Archives/Public/www-tag/2003Feb/0224.html
for a partial list. Reduction of parse time should be added to that
list.

>>Are you working specifically in this area?


TL> This is a part of our work but not the only one. We are working in all which
TL> is related to the content adaptation and negotiation problem in
TL> heterogeneous environments
TL> (http://opera.inrialpes.fr/people/Tayeb.Lemlouma/NAC.htm,
TL> http://opera.inrialpes.fr/people/Tayeb.Lemlouma/publication.html) and to the
TL> Device Independence problem (http://www.w3.org/2001/di/).

So you would see the binarization as happening in response to an
explicit client request, perhaps by a proxy?


>> I think that anything *completely* specific to SVG would be a
>> mistake. It would have some utility for SVG Tiny content, but it
>> would soon encounter arbitrary XML. And I'm not getting into
>> "details" such as interoperability. Imho the best approach is to
>> use something generic, optimised for SVG in an open manner (but
>> then I'm biased).

TL> You got the point.

I would agree with this point also, something that is entirely SVG
specific would not be useful.

TL> Personally I don't know any effort proper to SVG,

I am aware of three or four, see the reference above to the TAG
summary for some of them.

TL> this is an interesting problem to see if a specific binarization
TL> of SVG can be more optimized than existing XML binarizations.

Yes, approaches such as XMill can, with sufficiently careful creation
of an XML Schema and sufficiently careful optimization of type values,
give a more space-efficient delivery if that is the goal.



-- 
 Chris                            mailto:chris@w3.org
Received on Wednesday, 12 March 2003 15:13:18 GMT

This archive was generated by hypermail 2.3.1 : Friday, 8 March 2013 15:54:24 GMT