On Aug 4, 2005, at 3:25 AM, Paul Hoffman wrote:
At 7:37 PM -0400 8/2/05, Robert Sayre wrote:
One way of saying this would be "Atom Processors MAY ignore leading
and trailing whitespace in _____________."
That works for me. Another idea is "Atom Processors MAY ignore
leading and/or trailing whitespace in elements whose content does
not allow leading and/or trailing whitespace, such as IRIs and
________."
The more I think about this family of suggestions, the less I like
them. There is no coherent argument from the published
specifications of XML and IRIs that allows our 4.2.6 language "Its
content MUST be an IRI" to regard
<atom:id>
http://example.com/foo
</atom:id>
as anything but an error. I suggest an alternate formulation:
===========================================================
"Implementors are advised that there is a common class of error in
which, when the content of an element is specified as being an IRI,
software which generates Atom documents introduces spurious white
space before and after the IRI, for example:
<atom:id>
urn:uuid:60a76c80-d399-11d9-b93C-0003939e0af6
</atom:id>
In this case, since [RFC3987] makes it clear that the leading and
trailing white space cannot be part of the IRI, implementors MAY
choose to be forgiving in regards to this error, by ignoring the
leading and trailing white space."
===========================================================
I'm strongly -1 on treating this as anything but an error. We may at
our discretion make it forgiveable. This formulation would also
allow the feed validator to complain.
I suppose if we did this, we should generalize it slightly to apply
to date constructs as well. -Tim
