RE: [xri] URI Templates syntax

[Eran Hammer-Lahav <eran@hueniverse.com>]

Then the question becomes, what is the right way to do it...

I have been thinking about that, mostly in the context of host-meta.

It is already extensible by allowing each relation type or protocol to define their own syntax and vocabulary. The only use case we have today for templates is in host-meta (though it was originally requested for OpenSocial but not sure of the status of that at this point). 

So we can either define a template attribute 'syntax' with a default URI value that means {%} or remove it altogether and not have a default template syntax, in which case host-meta will define it (probably in the same way...).

EHL

> -----Original Message-----
> From: Scott Cantor [mailto:cantor.2@osu.edu]
> Sent: Tuesday, September 22, 2009 7:24 AM
> To: 'Drummond Reed'; Eran Hammer-Lahav
> Cc: 'XRI TC'
> Subject: RE: [xri] URI Templates syntax
> 
> Drummond Reed wrote on 2009-09-22:
> > Eran, FWIW I agree with you that ideally whatever XRD uses would be
> > compatible with whatever IETF is going to standardize. I don't know
> how
> > long the IETF process is going to take, however, or which way the
> wind
> > is blowing there. So I'm afraid my only feedback is to go with your
> best
> > instincts on this.
> 
> Or make it pluggable/extensible. Probably needless to say, I consider
> such
> pluggability an asset, not a complicating factor.
> 
> -- Scott
>