« Content Storage in Alfresco | Main | Katie Couric's Blog »



And now, what is the different between XPath and SQL Query?
I have no idea. I just can find the differences between XML and SQL query.


It's taken way too long, but the XQuery specifications are now Proposed Recommendations. The final Recommendation won't be until after 31 December 2006, but will probably be relatively soon thereafter (unless some unforeseen glitch is discovered).

See http://blogs.datadirect.com/jonathan_robie/2006/11/xquery_is_a_proposed_recommend.html.

Very good article expressing the differences between those three.

However, I'm curious why WebDAV and the DASL basicsearch query mechanism was left out. This has seemed to solved most of the problems related to Document and Content management searches, but what is your opinion?


G'day John,

Have you looked at what's happening in the .NET world, in particular LINQ? My initial impression is that it might be a step in the right direction.

For the non-projection cases I'm also intrigued by the idea of transparent, lazy-initialised object graph traversal via proxies (ala Hibernate, but with the data provider being an app rather than an RDBMS). There's a stagnant SourceForge project called CarrierWave that started to explore some of these ideas.


Hi Jeffery -

This is an argument that several have made in the committee.

In general SQL92 is close enough that we were able to create a fairly portable system with Documentum. It is only when you get into the esoterics of cursors, stored procedures and triggers that you have a real problem. Surprisingly, there has been a lot of concensus on bindings such as ODBC, JDBC and OleDB.

Fortunately, this is a standard and we can define the language how we want. We are definitely not going anywhere near things like cursors and stored procedures, and probably not venturing far into updates.

This really is about querying, how do we discover information in a repository? How do we entice developers into building applications that discover information in a repository?

A nice, simple SQL that adheres to the requirements above could do the trick. So could XQuery. Which would attract the most developers? My guess is something that resembles SQL, even if it were slightly quirky.

The problem with SQL, though is: whose SQL are you talking about? As with (X)HTML SQL has suffered from a pretty big divergence between the textbook standard and vendor implementations.

The comments to this entry are closed.