Late January XML-related Links Roundup

One of the things we're trying to do is help pull together interesting bits of information gleaned from XML-related weblogs and mailing lists that may be useful (especially within the WebData XML team) but might escape notice.  Here are a few links that are at least tangentially related to XML, with as little editorial commentary as I can manage:

Oleg Tkachenko discusses the scenario of an XPath injection attack and puts in a plug for the DynamicContext class in the Mvp.Xml libary as a way to use XPath expressions efficiently and securely.

Don Demsak thinks through the implications of the XSLT compiler in the next release of Visual Studio.NET and offers some suggestions for additional tools that others might want to build to exploit the new capabilities.

This is more from the SOA/web services domain than about XML per se, but I was intrigued by Jeff Schneider's report from a few weeks out in the real world.

There was a huge gap in knowledge between architects and developers. The architects knew the buzzwords, had read the white papers and most of the developers were completely clueless on next-gen stuff. Developers were soooo busy learning new J-stuff (maven, hybernate, canoe, etc.) that they just didn't have time to worry about the WS-stuff

I wonder whether things are different in the .NET world; perhaps developers are more in touch with the architectural concepts considering all that MS does with TechEd, PDC, etc.?  I also wonder whether the architects in either the .NET or Java worlds know the realities developers face as well as the buzzwords?  Anyway, this post certainly sets out a challenge for those of us who try to bridge the divide between architects and developers!

Speaking of architecture, the software world's favorite "real" architect, Christopher Alexander, was interviewed on NPR over the weekend.

Alexander has issued the final volume of his four-volume tome, The Nature of Order: An Essay on the Art of Building and the Nature of the Universe, which expands on his basic question: how do we build places and structures that are filled with life? 

[from the interview, answering a question about how this would apply to a home improvement project] If you think about the kitchen that is to be, and you try to imagine what is going to make you feel alive, there will be a struggle. You'll be beset by images, instructions, notions of what is fashionable. You've got voices speaking in your head about all those things, and. the most important thing you must do is get rid of those voices, really and truly clean them out and only pay attention to whether you are actually feeling better. ... even if the answer surprises you.

On the xml-dev mailing list, the big discussion recently is about general principles for XML design.  The rough consensus seems to be that there aren't any -- "Bottom line: Depends on the situation." [Sounds a bit like "get rid of those voices.. only pay attention to whether you are actually feeling better", maybe?] I personally am interested in the convergence of thinking about general data modeling / database design and XML schema design.  XPath 2.0's ability to join across XML documents/collections much as one can join across RDBMS tables, and efficient implementations (via XQuery) in DBMS, should allow us to re-think some assumptions about XML and normalization.

Bob DuCharme writes about the possibility of XLink being cleaned up  by W3C (although I like Micah Dubinko's imagery of the "spectre" of XLink trying to rise from the grave).   I dunno ... XLink is the poster child for a design-by-committee job gone wrong, and a W3C Recommendation that is nothing remotely resembling a "standard.".  R.I.P if you ask me :-)

The W3C Technical Plenary in Boston on March 2 will include a discussion of "Where XML is Going, and Where it Should (or Shouldn't) Go."  At least a couple of us from Microsoft plan to be in the audience and would love to hear our customers/partners' thoughts on this subject in advance.

The W3C Technical Architecture Group has (regrettably IMHO) agreed to wade back into the conceptual swamp surrounding  the relationship between the web services technologies and the Representation State Transfer (REST)  theory that some proclaim to explain the success of the World Wide Web.  Stefan Tilkov links to some of the more  interesting strands in this incarnation of the permathread. Speaking for myself, I found the REST critique of SOAP 1.0 useful a few years ago, and several key ideas  ideas were quickly accepted into SOAP 1.2.  Since then, I have  found the the ever-increasing abstractions of the REST advocacy against all things WS-* reminding  me of  architecture astronauts when I'm feeling charitable, and of Orwell's tirade against language used for "concealing or preventing thought" when I'm not.

In the shameless self promotion department (we are part of the Server and Tools Division), a newsitem related to one of the most interesting things I think have learned about Microsoft in my short time here -- forget the supposed advantages of monopoly;  the divisions  with the most competition are increasingly successful.

And finally, another "idea that the world is least ready for",  the underlying unity of XML and LISP.