Re: SVG vs. WebCGM

On Tuesday, September 12, 2006, 4:51:52 PM, Jeff wrote:

JS> Chris,

JS> Thanks for the response, I will try to look at some of the links
JS> you've provided.

In fact there is a more up to date comparison:
http://www.cgmopen.org/technical/webcgm_svg.htm


JS>   Maybe these follow-up questions are easy to answer
JS> though:

JS> 1) Is there any technical benefit of WebCGM over SVG?  Just breezing
JS> over the spec I saw that "NURBs" are mentioned...

Yes,  WebCGM 2.0 has NURBS
http://en.wikipedia.org/wiki/NURBS

there was an attempt to add NURBS to SVG1.0, but that did not succeed at
the time. I would like to see NURBS in SVG Full 1.2, but that discussion
has not taken place yet (so nopromises)

JS> 2) Since this is surely to come up again, could some mention of "the
JS> other spec" be given in both specs (maybe it's there and I missed it)?

SVG does mention WebCGM (in the context of the target attribute on
links) but does not have an explicit comparison. Probably because those
comparisons had been made at the XML conference series and were widely
available (if you already know they exist). Linking to them from the
public pages is probably a good idea.


JS>  Someone clueless like me

I don't see you as clueless:-)
 
JS>  will see both specifications and not have a
JS> clear picture of which technology should be pursued for which purposes
JS> and which has a brighter future.

Right. I agree that direction needs to be given in this area.

JS>  To me, they still look like
JS> competing specifications.  Hmm, maybe W3C is pursuing WebCGM now
JS> because SVG adoption is lagging or ... That's just an example thought
JS> process, I'm not trying to be provocative, hopefully you know I'm a
JS> staunch promoter of SVG ;)

They have grown up in parallel, they are both being used, and they have
cross-pollinated to a certain extent technically.

JS> Thanks again for your responses.

JS> Jeff Schiller

JS> On 9/12/06, Chris Lilley <chris@w3.org> wrote:
>> Hi Jeff,

>> On Tuesday, September 12, 2006, 4:08:25 AM, you wrote:

>> JS> You knew someone would ask this eventually, I'm surprised the
>> JS> charter doesn't mention it.

>> (Agreed the WebCGM WG charter does not mention it).

>> It has come up many years ago, in fact. The comparison was discussed at
>> several XML conferences, for example XML Europe 2001:

>> SVG and WebCGM – A Comparison
>> Chris Lilley, Graphics Activity Lead, W3C, France;
>> Dieter Weidenbrück, CEO, ITEDO Software, Germany

>> http://www.gca.org/attend/2001_conferences/europe_2001/graphics.htm
>> http://www.gca.org/papers/xmleurope2001/papers/html/s12-1.html

>> a later, more up to date comparison
>> http://www.cgmopen.org/technical/cgm-svg-20040419.html


>> JS> What is the differences in SVG and WebCGM?  Is SVG intended as
>> JS> general/all-purpose while WebCGM is only for technical/industrial
>> JS> drawings?  This seems like a rather arbitrary distinction.

>> The main difference is field of use. The industrial technical graphics
>> community picked CGM many years ago, its is today very widely used in a
>> particular market segment (primarily defence, aerospace, and
>> automotive).

>> Those users wanted an evolutionary improvement to add reliable,
>> vendor-neutral web linking; this requirement was met by WebCGM 1.0.
>> http://www.cgmopen.org/webcgm/w3c_rpt.html

>> At the same time, CGM has some limitations. Its not easily stylable with
>> either CSS or XSLT; it is not in XML; it lacks the graphical richness
>> needed for design intensive graphics; it has no animation capability.

>> This is why SVG was started, after W3C had grappled with CGM (over the
>> period 1996 to 1998). http://www.w3.org/TR/1998/WD-SVGReq-19981029

>> In general, WebCGM can be converted to SVG 1.1 without loss[1]. Even in
>> the technical graphics community, SVG is also used for training
>> materials etc (which need animation and more graphical richness) but the
>> long lifecycle technical documentation still needs to use CGM. WebCGM 1
>> and WebCGM 2 are still valid ISO CGM and thus can be used in systems
>> whose requirements were drawn upin the mid 1980s.

>> Note that tools for generating WebCGM, such as ISODraw, often export to
>> SVG as well.

>> Some features from SVG, such as having a DOM, are now being added to
>> WebCGM 2.0; but the main driver for WebCGM 2.0 is five years of
>> industrial experience with WebCGM 1.0.

>> JS> Why do we need two standards for scalable vector web graphics?  Can
>> JS> someone outline the purposes, distinctions, directions of these two
>> JS> seemingly competing standards within the W3C ?

>> Hopefully the above clarifies this to some extent. Happy to answer
>> follow-on questions.


>> [1] 99.5%, anyway - the CGM name attribute behaves like a non-unique ID
>> and XML does not have such a construct.

>> --
>>  Chris Lilley                    mailto:chris@w3.org
>>  Interaction Domain Leader
>>  Co-Chair, W3C SVG Working Group
>>  W3C Graphics Activity Lead
>>  Co-Chair, W3C Hypertext CG






-- 
 Chris Lilley                    mailto:chris@w3.org
 Interaction Domain Leader
 Co-Chair, W3C SVG Working Group
 W3C Graphics Activity Lead
 Co-Chair, W3C Hypertext CG

Received on Tuesday, 12 September 2006 15:15:20 UTC