Mark Watson's Blog
I am the author of 13 published computer books and a consultant specializing in Java, C++, and Smalltalk development. Please check out my two Free Web Books at my main site www.markwatson.com

 



Subscribe to "Mark Watson's Blog" in Radio UserLand.

Click to see the XML version of this web page.

Click here to send an email to the editor of this weblog.

 

 

  Monday, February 24, 2003


Querying RDF and Semantic Web Data

Recently, I installed the Open Source Sesame RDF server on one of my development machines. Sesame is an RDF data store system that has an SQL like query language that allows one to search for patterns in RDF data.

While I think that Sesame is very cool, and in general I think that RDF, topic maps, and the Semantic Web in general are cool and will hopefully prove useful, I had a thought today:

Would it be simpler to store RDF, topic maps, and any other sorts of information spidered from the web in relational database tables? This would probably work better with RDF than topic maps because RDF data are strictly tuples while topic maps are acyclic graphs. Still, I think that the combination of RDF with RDF Schema is equivalent in capabilities to topic maps (no flame wars, please!).

Anyway, I am working on Semantic Web portal software and I am still working out storage and access issues. RDF Schema are fairly static, so leave them in XML, and store RDF data directly in a relational database.
1:33:01 PM    


Programming language independence

I must admit to a little over enthusiasm when it comes to using many programming languages. It is great to be able to switch around from the awesome VisualWorks Smalltalk to my old work horse Common Lisp to the language I hate to love: Java.

I used to hope for something like CORBA that would allow me to use whatever language I wanted and interop with everything else. In my work flow, I have achieved programming independence fairly easily by:

  • Use (simple) XML files to archive all processed data (for me, this tends to be NLP data, results of web spidering for research purposes, etc.)
  • Use of web services (I like the simple XML over HTTP but I am also happy using SOAP and XML-RPC)
  • Project based configuration, not language based configuration
Anyway, I love to design and program and being able to use the most effective language for any task makes work fun and continually refreshing.
9:19:15 AM    

64bit CPUs

There is an interesting discusion on Slashdot this morning on Intel's announcement that they are in no hurry to support the 64bit desktop market. It is true that the Itanium is ill-suited for desktop use - that may be part of the issue.

Apple and IBM will have a real winner, I think, with the 64 bit Power chips. As some posters on Slashdot correctly point out, a 64 bit flat address space allows for creative memory mapping of large files, object oriented memory-mapped file systems, etc.

For AI work, a huge address space is a help, for sure. A few weeks ago, I spent (wasted?) several hours partitioning a machine learning experiment to run on my old and trusty dual P-III Linux box. A huge flat address space with RAM to back it up would have saved me some programming time and reduced 20 hour runs.

AI will be the real 'killer application' for 64 bit archiitectures

Software agents, real NLP capability, the Semantic Web, etc. will change the way that most people use computers. The 64 bit bit architectures will be an enabling technology for "real AI".
8:48:13 AM    


Click here to visit the Radio UserLand website. © Copyright 2003 Mark Watson.
Last update: 9/16/03; 7:35:02 AM.

February 2003
Sun Mon Tue Wed Thu Fri Sat
            1
2 3 4 5 6 7 8
9 10 11 12 13 14 15
16 17 18 19 20 21 22
23 24 25 26 27 28  
Jan   Mar