[
Lists Home |
Date Index |
Thread Index
]
Jonathan Borden wrote:
>
>...
>
> Much of these issues were the essence of the xml-uri@w3.org debacle. What
> you say is true. The problem, from an XML Namespaces POV, is that XML
> parsers ought not be burdened with an algorithm that depends on the URI
> scheme, particularly when new schemes, with new comparison rules, may be
> registered with the IETF in the future. From a practical point of view,
> simple string comparison is the best solution for a simple XML parser, on
> the other hand this doesn't take into account important features of URIs
> (such as relative URIs, domain case insensitivity etc.). So there you have
> it.
Practically speaking it would be quite feasible to implement equivalence
just for HTTP URIs and not other ones. The Web's record of deploying
non-HTTP URIs is quite pitiful. Other than the handful (ftp, mailto,
etc.) that were standardized In The Beginning, no more have gained
currency. For instance rather than having RealAudio URIs, there is a
tiny bootstrap file that you get through HTTP. It has within it
addressing information for the media player software.
Paul Prescod
|
