I've thought more about this and am convinced that if the spec is to include URIs (for example, in temporal expressions) then they're going to have to be explicitly distinguished. Heuristics to determine that an expression is a URI is not going to work. (I can elaborate on that if anyone wants me to.)
I first suggested that we should enclose them in parentheses when they are used as temporal expressions in intervals. But (a) that won't work, and (b) they need to be distinguished as URIs whether they are temporal expression in intervals or used for some other purpose, e.g. a season qualifier.
So I suggest that URIs within the spec be enclosed in quotes.
(Saašha has said that quotes cause problems but I am not sure exactly what those problems are, or if they are serious enough that we need to avoid them.)
> -----Original Message-----
> From: Discussion of the Developing Date/Time Standards
> [mailto:[log in to unmask]] On Behalf Of Ray Denenberg
> Sent: Wednesday, July 06, 2011 4:24 PM
> To: [log in to unmask]
> Subject: Re: [DATETIME] The EDTF draft specification - maybe a few
> From: Saašha Metsärantala
> > I would suggest to add a space after the first temporal expression to
> > make it clear that the slash between the two temporal expressions is
> > not part of the first temporal expression.
> Good point, the syntax is ambiguous with the slashes since there can be
> any number of slashes in a URI, but I don't think a space is a good
> I have changed this so that temporals (in an interval, only) are
> enclosed in parentheses.
> But that's not a good solution either, because other sorts of
> approximate, etc. expressions can begin with open parentheses.
> So, I think that the temporals within an interval need to be enclosed,
> but I'm not sure what character should enclose them. I think quote
> may be the best solution.