Inspirational Technology

LDAP Exhaustion and XQuery Lamenting

Spent the day today in a data center installing a couple LDAP servers for a client. It's been a while since I've spent any amount of time inside a data center and I'd forgotten how exhausting it can be. The noise and the cold air blasting through the floor really takes it out of you.

LDAP is an interesting technology, it's what sparked my interest in semi-structured data which led me to working on native XML databases. I hadn't worked with it in quite a few years, but it's a good example of a stable standard. In fact the LDAP protocol it self hasn't changed at all in that time period. The products have of course matured, but it's all still the exact same concepts and at the low levels the details are the same. A refreshing change compared to the spec a week mess that XML has become.

Fortunately the core specs in XML (XML 1.0, Namespaces, XPath 1.0, XSL-T 1.0) have now proven to be stable and I suspect we'll start to see the weaker (and much more complex) later specs beginning to drop off the radar. It's too bad, but in a lot of ways just about everything after the release of XSL-T 1.0 seems pretty irrelevant. This isn't altogether a good thing. The current XPath 1.0 and XSL-T 1.0 specs definitely have room for improvement. I just wonder if XPath 2.0 and XSL-T 2.0 are going to provide that improvement without drowning under the added complexity that being associated with XQuery has introduced.

I've gotten to the point where I don't even pay much attention to XQuery anymore, maybe that's not a good thing, but I just don't see any real world interest in it. In particular, it's pretty much non-existent in the Open Source world. None of the big three Open Source XML databases (eXist, Xindice, Sleepycat DbXML) support it, and I kind of doubt that they ever will. The sad reality is that It's just not asked for all that often.

Once upon a time there was a real need for XQuery(or at least there was for XPath with joins and updates), but in the pursuit for academic perfection the complexity has mounted, the number of associated specs has mushroomed and the spec has delayed it self into irrelevance.

Why am I writing about this? I don't know, maybe I just wish that XML databases actually mattered anymore. There was an awful lot of waiting brought out by the presence of XQuery. It's provided a big cloud to hang over the whole XML database arena. Instead of focusing on doing profitable work with XPath and just adding the missing pieces, innovation stopped and everyone delayed all their plans around the development of XQuery. Now, several years later XQuery still isn't finished and products are finally shipping with incomplete XQuery implementations with proprietary extensions for things like updates. This stuff should have been added years ago and now we have delayed products shipping with implementations that aren't going to interoperate anyway. So what have we gained? In my opinion, not much. In fact I think XQuery may end up killing the entire XML database market.