Dangers of blogging

So in a blog post from last October which was mostly devoted to mocking Stephen Wolfram, I casually insulted HistCite, a citation network navigation program developed by Eugene Garfield, who basically invented the very idea of citation networks. So half a year goes by and then I get mail from Dr. Garfield complaining about this slander. I was rather taken aback that someone of his prestige would take the trouble to respond to an offhand remark by a mostly-anonymous blogger, but I backpedalled fast and responded apologetically:

Please forgive my offhanded nasty remarks; I have the greatest respect for your work, and I'm honored that you'd take the time to respond.

Google and similar web services have spoiled us (or spoiled me at least) -- we expect everything to be free and instantaneous and very easy to use. I was just wishing out loud for something that would let me see citation networks without fuss, bother, or cost. I apologize for calling HistCite "ugly", what it really is (in my opinion) is a complex tool for a professional information analyst, rather than a simple tool for casual browsers, which is what I was idly wishing for. So much academic work is now available casually to non-specialists now (through Google and elsewhere); it would be nice if the citation relationships were equally available.

Since professionally I work in IT for life science and other areas where information and software costs money, I don't really believe everything can come for free.

Well, that's why I blog with a pseduonym, so I don't have to worry too much about saying nasty things and having it come back and bite me. What else is this for? Still waiting to hear from Stephen Wolfram.

The lesson that things on the internet stay around forever and can come back and bite you is of course not new. When Google Groups made usenet postings from the early 90s part of the permanent public record, it caused me some personal embarassment that persists today. For some reason my dumbest posts still seem to come out on top, I suppose because they produced the largest number of dumb responses.

Microsoft ignores semantic web?

The Semantic Web has been the fad of the moment for some years now, and if I was a major platform vendor I would at least make sure I had my toes in the water to support development of applications that wanted to interoperate with it. However, Microsoft seems to be mostly ignoring it, in terms of providing tools and libraries.

If you want to use the latest and greatest semantic web standards for representing ontologies (OWL and relatives), you need a package that can deal with the file formats and more importantly implement the reasoning and inference capabilities implied by this kind of representation. There are a few serious open-source packages that can do this, but they are all in Java or some other non-.NET language. If you want to work on Microsoft's platform, there is a nice package called SemWeb, but it operates only at the RDF-triple level, and is pretty much a one-man hack rather than a major research effort.

When I queried Microsoft about this, they basically said they don't care and I should try to use IKVM to run the Java packages on .NET. IKVM is a very cool hack but that doesn't sound very realistic given the bleeding-edginess of this technology and the need to debug across the API boundaries.

C'mon guys, aren't you about embrace-and-extend? Maybe the problem is they have their own splufty new language-integrated-data-access thingy (LINQ) and see the semantic web as something of a competitor/distraction. Maybe the whole flavor of Semantic Web is too cross-platform for Microsoft (but they took up the non-semantic Web, and XML, this is just the next phase).

So this entry isn't all kvetch, here's a useful compilation of semantic web toolkits for a variety of languages/platforms.