Last Call Problems

July 26, 2000

Leigh Dodds

This week the XML Deviant dips into the SVG developer lists to find developers frustrated with the specification, which is still at Last Call status.

SVG Frustration

A recent posting to the W3C www-svg mailing list demonstrated that frustration is surfacing among some developers over the apparent lack of progress on the Simple Vector Graphics (SVG) specification. Lawrence Fry described the slow progress as disheartening:

When is svg expected to be a W3C recommendation? This standard has been moving at a snails pace recently, and the release of two successive "last call" specs is disheartening for those of us relying on the new standard.

Fry went as far as to suggest that the delays had been "orchestrated" by Working Group members with vested interests in seeing the specification delayed further. While these claims are not supported by the available evidence, they appear symptomatic of wider frustration with the standards process and the role that vendors play.

In response to a related discussion concerning vendor buy-in to the standard on the svg-developer list, Wade Harrell urged developers to be cautious before leaping to condemn particular vendors:

I applaud Adobe for making this plugin so early in the process of SVG's growth. Most companies would not be willing to risk that kind of development time on something that is still in the 1.0 working draft phase, I can not say that I blame them. SVG is evolving and fortunately Adobe has given developers a chance to experiment with, and give feedback on, the standard. I am hesitant to condemn any company at this point for what appears to be lack of participation in SVG's growth. Let's wait until there is a half dozen commercial sites out there using the technology and then we can say "wake up and smell the coffee".

Other SVG developers had singled out Macromedia as showing little sign of support for the emerging standard. These comments were prompted by the announcement of tighter integration between Macromedia Flash and Internet Explorer, causing dismay from developers hoping for support for SVG in the Macromedia toolset. The integration takes the form of proprietary XML vocabularies that can be embedded into an HTML page. SVG could easily have been leveraged to achieve the same results.

John McKeown, highlighting the efforts of other vendors, invited Macromedia to comment on their intentions regarding the specification:

Other working group members such as Adobe and Corel have already provided useful support for SVG in their editing tools. IBM and Sun have also been active in supporting the language. I assume Macromedia play an active part on the SVG working group, and I'm sure they have a lot of knowledge to offer. It would be nice to hear from somebody in Macromedia about how they intend to support SVG (if at all).

Wade Harrell, in summing up a lengthy posting, observed that it falls upon the developer to increase awareness of a standard by demonstrating its capabilities:

I am anxiously awaiting the future of this standard. In the meantime I will work with what we have and hopefully turn some heads. Awareness will bring about support, so it is our job as developers to see how much we can do.

This point was echoed by Alejandro Fernandez who believed that real-time, database-driven SVG applications would demonstrate its power:

The more powerful and well-drawn/designed svg there is in the world [...] the more valuable svg will be.

Open Processes

In comparison to other fledgling standards, SVG is lucky in the number of implementations that are available. Add to this the availability of a conformance test suite, and progress begins to seem very healthy.

However it's worth considering what might be the root cause of the uncertainty and frustration that developers are feeling. One obvious cause is lack of transparency in the W3C process. SVG is not alone in its Last Call status; indeed many specifications are well beyond their Last Call dates with little sign of additional activity. There is an opportunity here for the W3C to provide greater feedback on its progress. Continuing updates would at least keep developers informed, if not actively curb the wilder conspiracy theories. Ongoing public feedback from vendors beyond initial testimonials would also be of benefit.

Of course there have already been some improvements in this area, notably with the publication of the XML Schemas Last Call comments and the recent developer poll over XML Schema concerns (see "Schemas Revisited" and the published poll results for more information).

However the opening of the xml-uri debate (see "News From the Trenches") did little more than move the discussion into a public forum. Lack of process meant that little was really achieved. A proposal has been made to initiate an official W3C Activity to address the xml-uri issues. Interestingly the proposal notes

...that public discussion of activity proposals is non-traditional, though not prohibited by W3C process

So there is obviously some leeway within the W3C process for greater communication with the developer community. Let's hope this Activity can set a precedent for future efforts.