What an amazing week! Web 2.0 Summit The week included the Web 2.0 Summit. I crashed the event to talk with Jeff Nolan and also managed to have a fast conversation with Anil Dash. I also bumped into John Furrier…….

Coming from SAP and having some understanding of what the other big enterprises are doing in this area (I led a competitive strategy team at SAP focused exclusively on Oracle) I can offer that much of what you see in semantic web research in these companies is directed at BI solutions, enterprise search, and integration of structured and unstructured data. All are important but none are to the extent that you (or I for that matter) would prefer.

The one company that really is pushing the envelope of a database in the cloud is salesforce.com with both appexchange and the recently announced Apex language. However, close inspection of that two things reveals that they are anything but semantic… in fact you could almost just call them “PL/SQL in the cloud” and still be accurate.

More to my comment about applications and process, both of which are dependent on a well defined data schema if for no other reason than this is what the culture of these companies demands because they believe it is in their competitive interests to have it that way – control. The recently launched Workday does offer a glimpse of a more declarative ideology but I’ll have to learn more about that before commenting.

I’ll end here because this dialogue just gave me a great idea for a full post that I need to think about so more before writing.

Thanks again.

There is absolutely no conflict between what MÃ¥rten’s company delivers and what I believe in, and MySQL is a great backend for Semantic Web applications. I’d also note that MÃ¥rten wasn’t talking about products in the enterprise, he was talking about the “Great Database in the Sky” (and I didn’t actually call him regressive).

You mention Oracle, they recently built support for RDF (the core Semantic Web data language) into their flagship product. Other companies in the enterprise arena are also active around Semantic Web technologies, notably HP and IBM, and I believe SAP are busy around Semantic Web Services. Even so the enterprise market isn’t necessarily the first target for languages like RDF/SPARQL (however the W3C are trying to promote these technologies in those circles). This stuff is designed to make the Web better. But given the events of the past decade or so that’s liable to impact the enterprise.

On the question of working from applications and process – right now I disagree about these being so significant in this context, but am prepared to be convinced. The Web started as a fairly minimal, fairly local info-sharing application, but had features like uniformity of interface and ease of use that allowed it to benefit from the network effect. I rather like your phrase the “notion of the data just being there and the relationships figured out later”, because to a great extent it reflects what happens in the world, outside the confines of the firewall.

Anyhow, though I largely disagree with your impression, thanks for taking the trouble to comment on this.