Re: [xsl] XSLT 2 processors

[Michael Kay <mike@xxxxxxxxxxxx>]

>I'm genuinely surprised that Microsoft has dropped the ball on this. 
They put in a massive effort when they introduced .NET which was after 
their browser played a pioneering role in the genenis of XSL. Not 
affording their .NET library XSLT2 is tantermount to sabotaging the 
progress of XML related development. It just doesn't make sense and is 
pretty lame IMO.

I don't know any detail of what went on inside Microsoft other than a 
few rumours that leaked out, but I do know a little bit about the 
process, or lack of it, that results in product development decisions 
inside large companies. There will always be champions (perhaps in this 
case even a Champion...) advocating investing in a particular 
technology; there will be others with competing proposals, and there 
will be managers who aren't intimately acquainted with the technical 
detail who end up having to make the decisions. There will also be 
customers pushing for one course of action or another, and some of those 
customers will have clout and others won't: not necessarily because of 
pure revenue considerations, but perhaps because keeping them "on side" 
is important for unrelated reasons.

In making these decisions, the biggest problem is that so much of this 
kind of software is free, which means that the value the customers get 
from a product doesn't flow back into investment in that product. When 
the champions for developing a particular piece of software can't put 
together a business case based on the direct revenue from that product, 
the arguments for and against become very woolly and subjective.

XSLT has always been a technology that some people love and others hate. 
The language has deep beauty but superficial ugliness, so this range of 
reactions is not surprising. No doubt this polarisation of views exists 
within Microsoft as much as anywhere else; and no doubt the negative 
views will have been amplified by those advocating other investments 
that were competing for the same budget, such as XQuery and Linq. 
Perhaps some of these voices were even arguing that Microsoft could 
safely leave this area to third parties.

So you shouldn't be surprised. These decisions are complex, there is a 
lot of emotion involved, and very little hard data about the impact of 
deciding one way or another.  Probably at least half the product 
development decisions made in a big company turn out to be wrong, so 
expecting a big company to make the decision you consider right is naive.

Michael Kay
Saxonica