> James M Snell scripsit: > > > For URI Templates they generally would be. For IRI Templates, however, > > given that the whole point behind IRI's is to allow non-ascii characters > > to be used, it makes very little sense to limit the varname to the ASCII > > alphabet. Allowing this does not make the template any harder to parse. > > I may want to support IRI Templates in a language that supports Unicode > fine, but has a limited set of identifier names. That argues for > limiting the space of variable names, and I don't see a compelling > argument in favor of it. What languages do you have in mind? Python (3000), Ruby, Java, Haskell, C, C#, C++, Visual Basic.NET, ECMAScript, and some/most Lisps all support non-latin characters in identifiers (usually based on Unicode Standard Annex 15). I don't think that it makes sense for new specifications to step back into the Euro-centric past. - BrianReceived on Thursday, 29 November 2007 17:58:38 UTC
This archive was generated by hypermail 2.4.0 : Sunday, 10 October 2021 22:17:50 UTC