WebCGM

1
WebCGM SVG Revisited

• Lofton Henderson
• Dieter Weidenbrück

2
Revisited?

• 2003 study to cut hype provide data about
• Why two W3C Recommendations for s.v.g.?
• Lots of claims
• Little real data
• Since then
• Changes in SVG (new stuff in 1.2 WD)
• Better understanding of some issues
• Evolution of implementations
• Results have been disputed

3
Some Key Differences Found

• Technical
• Object linking
• DOM, Event model
• Animation, Styling
• Encoding File sizes (esp. raster images)
• Interoperability framework

4
2003 Overall Conclusion

• Despite vastly greater power of SVG, WebCGM
  actually has an edge in functionality specific to
  Technical Graphics.
• WebCGM has edge in
• Specificity
• Stability Maturity
• Reliability

5
2004 Results Preview

• Technical differences narrowed (not gone).
• WebCGM still has interoperability framework edge.
• SVG is making real progress in other application
  sectors.
• SVG for peripheral applications to core TG, e.g.,
  training animation.

6
Context Technical Graphics Requirements

• Complex geometry with modest graphical rqts
• Precision
• Text
• low typographical requirements
• precision
• Metadata rqts modest but very specific
• Reliability
• Reusability and longevity
• Interoperability

7
Demo

• Dieter do a little demo of TG here.

8
About the Formats

• WebCGM
• profile of an established ISO standard,
• targeted at Web-based technical graphics
• SVG
• New W3C standard,
• targeted at creative graphics design, high-
  quality, dynamic Web pages

9
W3C Says

• W3C Graphics Activity Statement
• There are two different markets for vector
  graphics
• W3C scalable graphics requirements
• WebCGM partial SVG full
• In principle they should coexist complement

10
Questions

• Isnt SVG much more powerful?
• There are lots more than 2 SVG viewers
• Do it with scripting
• Raster comparison is unfair
• Isnt clear-text XML an advantage?

11
More Questions

• Isnt CSS styling a big advantage?
• You confuse formats implementations
• SVG 1.2 will fix that
• Use ltggt (or class) for name linking
• WebCGM is perfect but SVG is not?

12
Updated 2004 -- graphics

• Ref tables in Big Paper
• SVG 1.2 will have
• Non-scaling lines, link title multi-link,
  screentip.
• Out-of-line link solution (actually, 1.1)
• (Text issues clarified)
• (Raster capabilities clarified)
• WebCGM (2.0) is getting
• Simple DOM and limited styling
• Better developed event model

13
Interoperability framework 04

• The paramount consideration for TG
• Measured on several axes
• Extensions
• Resource limits
• Language flavors and profiles
• Predictability of text model ()
• Completeness of implementations ()
• Test suites ()
• Maturity and stability

14
Interoperability framework 04

• WebCGM update
• 1st release of WebCGM product data (ICS)
• Interop problem tracking system
• Incremental development of the 5 plug-ins
• Minus no summary matrix raw ICS
• Minus divergence with ATA, AECMA

15
Interoperability framework 04

• SVG update
• Number of plug-ins increased from 2 to 4
• More tests for 1.1
• Updated implementation status matrix
• a few 1.2 tests (unpublished)
• Functional completeness of implementations
  improving.
• Minus Validator dead?

16
Interoperability framework 04

• SVG and profiles
• Responsibility of application sector (e.g.,
  ATA)
• No Rules for Profiles for now.
• Assessment a TG profile is a MUST.

17
Looking ahead

• WebCGM outlook
• WebCGM 2.0 stuff limited DOM, events
• Go live with Interop Problem Tracking Service
• Develop roadmap
• Re-sync with ATA
• Resource pinch

18
Looking ahead

• SVG outlook
• Finish 1.2 (2004/2005)
• Continued penetration mobile, 3GPP, etc
• Some talk of SVG support group/consortium
• Profiles? (sort out a venue for TG profile
  development)

19
Conclusions 2004

• Technical differences
• In published versions, WebCGM still has some edge
  over SVG
• Will narrow with publication of SVG 1.2 and its
  product implementation
• Wont disappear completely
• But its a second-order difference
• Interoperability framework is the big difference

20
Conclusions 2004

• Early life cycle of standards
• Initial excitement high expectations
• Copious but unfocused implementation
• Partial/wrong implementations too numerous
• Deflation of expectations
• Core group organizes and attacks problems
• Interop confusion slowly mitigates
• Slow improvement of usability/productivity
• SVG is in first half
• WebCGM is in second half

21
All things considered

• WebCGM still has edge for TG because
• Strict profile focused on TG, versus none.
• Stability of the standard itself.
• Maturity of the implementations.
• Support group (CGMO) focused on interoperability
  solutions.

22
Q and Ahttp//www.w3.org/Graphics/svghttp//www
.cgmopen.org