Re: HRRI vs IRI in XML

* Norman Walsh wrote:
>I'd be *delighted* to be proved wrong. Get consensus from the
>community that it would be ok to change the definition of system
>literal or the value of href attributes so that "this {string}.html"
>no longer accesses the resource we'd identify with the URI
>"this%20%7bstring%7d.html" but is instead an error and I'll be the
>first in line to fix the specs.

You are denying the perfectly valid option of graceful error recovery if
such values include disallowed characters. XML processors, for example,
may continue normal processing if a system literal contains '#' even
though such a document is in error. I note that this particular part of
the XML specification is not well-implemented in many processors, it is
common for them to treat a literal '\' differently from '%5C' while the
specification requires them to treat them identically.

>If the IRI spec can be extended to cover the characters we need, then
>I think we could say that these things are IRIs. That means we need
>to answer two questions:

That is impossible as you'd want to include the space character while a
number of deployed formats use space-separated lists of resource identi-
fiers, like the xsi:schemaLocation attribute. At the least you would
need to have IRIs-with-white-space and IRIs-without-white-space where
the ones with white space would have a host of problems.
-- 
Björn Höhrmann · mailto:bjoern@hoehrmann.de · http://bjoern.hoehrmann.de
Weinh. Str. 22 · Telefon: +49(0)621/4309674 · http://www.bjoernsworld.de
68309 Mannheim · PGP Pub. KeyID: 0xA4357E78 · http://www.websitedev.de/

Received on Friday, 10 August 2007 06:39:37 UTC