The title of this post is a quote from Ted Nelson—from his piece I DON'T BUY IN about the web not being hypertext. I thought this title might be a little bit provocative!

If you have been following my blog, you know that I am thinking about XML right and left with regards to how I will build-out the iCite net prototype. I first started working with SGML / XML around 1992, so I am not easily surprised about what XML offers or doesn't.

But, I like to get outside of XML-centric view of data regularly, at least in the current climate where all software is, or wants to be, associated with some XML vocabulary, because I find I lose perspective on non-XML ways to do things.

Of course, in the Java programming world, and I am sure in the Object Oriented programming world in general, there are plenty of folks who are critical of trying to do everything with XML. But, for myself, I like to re-visit some of the "true relational database" (TRDBMS) writings, especially those of Chris Date and Hugh Darwen, and also those of Fabian Pascal. Some of these are on accessible on the web at Database Debunkings.

I should note here that I think it is unfortunate that people get into religious wars between TRDBMS and OO and XMLs and RDFs, etc. There are some important points made for each and I am glad that those are being made, but I think there is a lot of unnecessary antagonism and academic brawling in bad spirit that I eschew and do not condone.

So, with that in mind, first of all, I used the "hierarchy hamburger" title in good spirit and humour—I hope you think it is a funny phrase too!

Next, with a warning about its religious war overtones, I wanted to point out Fabian Pascal's Setting Matters Straight - April 2003 (from The Data Administration Newsletter, TDAN.com). Among Pascal's comments on XML databases (in response to Tom Bradford), he says:

Note, first, that Bradford does not give any thought to why "hierarchic and object databases never gained enough momentum to present a serious threat to SQL (not relational!) databases". Had he bothered to check, he would have found that decades ago hierarchic databases almost ground to a halt due to their complexity, inflexibility, and ad-hoc nature. . .

So, in my own work on the iCite net project right now, I am actually considering something like this: is the "extensibility" of XML (at least when thinking about XML being used as a database structure for data) going to land me in a situation that is actually overly complex and inflexible?

Coincidentally, today I was telling someone I know about the iCite net for the first time, and he happens to be someone who told me about Ted Nelson's "Literary Machines" about 15 years ago. So, we were wondering what had happened with Nelson's Project Xanadu, and found Nelson's website.

So, I read up on this and Nelson created this thing called the ZigZag structure, which is really interesting. If you are interested in data structures / data models, either from the TRDBMS, or the OO, or the XML, or the RDF perspective, definitely check-out ZigZag structures. (Also, if you are into APL data structures, ZigZag borrows the term "rank"!)

What has my attention at the moment is that both the TRDBMS writings and Ted Nelson's criticize the hierarchic databases model and claim that XML (databases) are that model and should be criticized as such.

I think I just need to be sure that I don't fall into the trap of physical database dependency on hierarchic structures, i.e., data buried deep in XML tags. So, I think my task at hand is to consider if any hierarchic view of iCite data (which includes some of the primary views of iCites) is too tightly coupled to an underlying physical hierarchy of data storage.

I don't know if this is correct, but I think this is the kind of consideration to be had about RDF and its relationship to XML. Maybe the graph structure of RDF is too often tied too tightly to the hierarchic structure of XML trees.