Throwing off Ballast (Part II)

One other thing to get rid of is my ISO involvement.

When I got interested in Topic Maps in 2000/2001 I was quickly sucked into this circle of standardization. First only by providing intense feedback as developer on the SC34/WG3 mailing list, then also by participating in meetings all over the world.

Against popular belief these meetings were neither boring nor brutally controversial. Except maybe 2003 when the Ontopian and the Newcombian church were arguing about the virginity of Mary, i.e. the reifiability of an association and its components. But otherwise I have only pleasant memories. And despite my strong CSy (computer sciencey background) I certainly enjoyed the overall likemindedness of the group. And the occasional young blood (read: new contributors, don't read: child sacrifice). And the places to visit.

As long as I could coalesce the ISO work somehow with academic work, and with it help to fund the travel and the time, I was able to invest into Topic Maps technology as a sole developer. That particularily refers to TMQL, the query language, where I tried to develop a reference implementation in lockstep with the standard drafts. It was very expensive to (a) research the underlying theory properly and (b) come up with an extremely flat and simple to implement language. (Yes kids, TMQL is simple to implement.)

All the plans came to naught 2007 when Bond University defrauded me and produced a net damage of several hundred thousand dollars. I could compensate that for the time being and continued to work on TMQL up until the first half of 2008, but now I cannot see any perspective, academically or commercially, in it.

In the past I had approached several companies and suggested the development of a TMQL parser (or virtual engine), but there was not even a spark of interest (with the exception of the Topic Maps Lab in Leipzig). Instead, TM vendors continued to foster their own query languages. Which is their right, of course, but which also shows the desinterest in TMQL.

So all the ideas I have now will flow into another language, TempleScript, a programming language for geosemantic networks. That is one language, for all aspects: writing facts, writing constraints and writing queries. And assignments.

Long live Haskell.

Posted In