Comments on James Clark's Random Thoughts: What's allowed in a URI?

Thanks for working through this. Maybe the Java f...

2008-11-21T01:38:00.000+07:00

Thanks for working through this. Maybe the Java folks can be persuaded to take notice.

BTW, RFC3987bis has been 'delayed'. In the interim, the XML Core WG, with permission from the RFC3987bis editors, has published the LEIRI section as a W3C Working Group Note: Legacy extended IRIs for XML resource identification

I wonder if writing a compliant implementation to ...

2008-11-19T19:09:00.000+07:00

I wonder if writing a compliant implementation to some spec has to be this hard. In particular the areas related to XML schema always seem horribly confusing.

Some specs are certainly better written than others. Still, it might be nice if someone could wrap up all the confusing cross-references every now and then... this would probably greatly increase the number of conforming implementations (but who has the time ...).

James,RFC 2396 (and java.net.URI) are seriously ou...

2008-11-18T04:59:00.000+07:00

James,

RFC 2396 (and java.net.URI) are seriously out of date. Please upgrade to RFC 3986, and consider either borrowing or rolling your own URI handling code.

This is extremely helpful. Thanks.

2008-11-17T21:40:00.000+07:00

This is extremely helpful. Thanks.

You could save some work by looking at the jena IR...

2008-11-17T21:20:00.000+07:00

You could save some work by looking at the jena IRI library. Jeremy Carroll spent some time navigating the numerous RFCs, and I believe the code is partially driven by them (?).

You can find him at: jeremy (at) topquadrant.com