November 11, 2003
@ 11:10 PM

I noticed the followingRDF Interest Group IRC chat log discussing my recent post More on RDF, The Semantic Web and Perpetual Motion Machines in my referrer logs. I found the following excerpts quite illuminating

15:43:42 <f8dy> is owl rich enough to be able to say that my <pubDate>Tue, Nov 11, 2003</pubDate> is the same as your <dc:date>2003-11-11</dc:date>

15:44:35 <swh> shellac: I believe that XML datatypes are...


16:08:15 <f8dy> that vocabulary also uses dates, but it stores them in rfc822 format

16:08:51 <f8dy> 1. how do i programmatically determine this?

16:08:58 <JHendler> ah, but you cannot merge graphs on things without the same URI, unless you have some other way to do it

16:09:02 <f8dy> 2. how do i programmatically convert them to a format i understand?


16:09:40 <shellac> 1. use


16:10:13 <shellac> 1. use a xsd library

16:10:32 <shellac> 2. use an xsd library


16:11:08 <JHendler> n. use an xsd library :->

16:11:30 <shellac> the graph merge won't magically merge that information, true

16:11:34 <JHendler> F: one of my old advisors used to say the only thing better than a strong advocate is a weak critic

This argument cements my suspicions that the using RDF and Semantic Web technologies are a losing proposition when compared to using XML-centric technologies for information interchange on the World Wide Web. It is quite telling that none of the participants who tried to counter my arguments gave a cogent response besides "use an xsd library" when in fact anyone with a passing knowledge of XSD would inform them that XSD only supports ISO 8601 dates and would barf on RFC 822 if asked to treat them as dates. In fact, this is a common complaint about them from our customers w.r.t internationalization [that and the fact decimals use a period as a delimiter instead of a comma for fractional digits]. 

Even in this simple case of mapping equivalent elements (dc:date and pubDate) the Semantic Web advocates cannot provide a solution to how their vaunted ontolgies can provide a solution to a problem the average RSS aggregator author solves in about 5 minutes of coding using off-the-shelf XML tools. It is easy to say philosphically that dc:date and pubDate after all, they are both dates, but another thing to write code that knows how to treat them uniformly. I am quite surprised that such a straightforward real-world example cannot be handled by Semantic Web technologies. Clay Shirky's The Semantic Web, Syllogism, and Worldview makes even more sense now.

One of my co-workers recently called RDF an academic plaything, after seeing how many of its advocates ignore the difficult real world problems faced by software developers and computer users today while pretending that obtuse solution to trivial problems are important, I've definitely lost any interest I had left in investigating any further about the Semantic Web.


Friday, November 14, 2003 7:03:20 PM (GMT Standard Time, UTC+00:00)
Does this also mean you're no longer interested in XML schema, if the RFC 822 problem can't be solved with it?

How do you solve this, btw?
(I use XSLT)
Friday, November 14, 2003 8:03:41 PM (GMT Standard Time, UTC+00:00)
I find it puzzling that Dare puts forth such energies into railing against RDF. Especially here with an argument about 822 style vs W3CDTF style timestamps. They've nothing to do with each other. The RDF folk have decided to use the ISO-8601 descendent form largely because the 822 style is such a screw up internationally. Besides which, the 822 style was never intended for serious programmatic parsing. Given the days in which it was invented the ability to strip out a line of text that was already formatted for human readability was a great savings. To think that's anywhere close to being needed in this day and age is ludicrous at best.

As for your lost interest, I'd daresay it will soldier on without you and probably better for it.
Saturday, November 15, 2003 10:10:53 PM (GMT Standard Time, UTC+00:00)
My feelings for W3C XML Schema should be well known to anyone who reads my blog (it is too complex for simple cases and lacks functionality for complex cases so it is the worst of both worlds) however I don't see what this has to do with the transformation problem. It was never intended to solve this problem. I solve the transformation problem using a combination of regexes and the System.DateTime class in the .NET Framework. The System.DateTime class can handle some RFC 822 dates but not others so we make a first pass using regexes to see if the date needs to be massaged in any way or not and if it does it is altered then passed to the Parse() method of the System.Datetime class.

I was interested in RDF because I thought it could solve some problems I'd like to see solved in the area of syndication and blogging especially with the impending arrival of WinFS. However I've realized that RDF is a waste of my time. I have no interest in debating the pros and cons of RDF, the only reason I'm even bothering to blog about this is so that in future if someone asks my opinion about RDF I can just page slap them with links to the related entries in my blog.
Monday, November 17, 2003 10:20:26 AM (GMT Standard Time, UTC+00:00)
"It was never intended to solve this problem." Same goes for Semantic Web technologies.
Thanks for the code tip. fwiw, the situation's pretty similar with java.text.SimpleDateFormat.

btw, I think there's likely to be a lot of potential for using RDF with WinFS. The RDBMS model is fine locally, but interop between machines is likely to be difficult without communicating something of the DB structure/semantics. The partial understanding RDF offers could be very useful here. I'll give you a shout when it starts happening ;-)
Friday, November 21, 2003 10:55:40 PM (GMT Standard Time, UTC+00:00)
I never cease to be disappointed by how much energy people are willing to put behind deriding that which they do not understand. That /you/ could not find use for it hardly indicates that others share your perspective.
Comments are closed.