Is XML 2.0 doomed?

XML 2.0 seems to be becoming the new buzzword and hot topic among many forums.

While I think that XML 1.0 would deserve a certain amount of factoring, I don’t think that XML 2.0 is likely to ever happen nor even that it is something we should wish.

The reasons of the success of XML 1.0 are not that difficult to analyse:

1. The cost/benefit of developing XML 1.0 applications compared to previous technologies has been generally analysed as highly positive.
2. XML 1.0 is very near to being the greatest common denominator between the needs of a very wide range of applications, including document and data oriented applications.
3. XML 1.0 has been proposed by a normalisation body that had the credentials to push such a specification.

I don’t think that this is likely to happen again for XML 2.0:

1. The unfortunate XML 1.1 recommendation has shown that the cost of the tiniest modification to XML 1.0 is so high that it is difficult to think of a benefit that could compensate this cost. While XML 1.0 is certainly imperfect, the cost of its imperfections isn’t just high enough.
2. A fairly good consensus on the features supported by XML 1.0 has been possible in a small Working Group working reasonably isolated from the pressure of the lobbies that finance the W3C. All the recent specifications developed under more pressure and hype such as W3C XML Schema, SOAP, WDSL, XPath 2.0, XSLT 2.0, XQuery 1.0 and others show that this not likely to happen any longer and that on the contrary, a XML 2.0 specification would most probably loose the balance that has made XML 1.0 successful.
3. During the past six years, the W3C has lost a lot of credibility to the point that its most influent participants are now ignoring its most basic recommendations such as XHTML, CSS, SVG, XForms, XLink and many others. This loss of credibility would greatly compromise the success of a XML 2.0 recommendation published by the W3C.

What is likely to happen with XML 2.0 is either a recommendation that is easier ignored by the community at large or much less generic, lightweight and flexible than XML 1.0.

I think I would prefer the first option!