Re: [xri] PLEASE REVIEW: Draft of XRI-As-Relative-URI page

[Nick Nicholas <opoudjis@optushome.com.au>]

The proposal looks fine to me. The xri.* binding looks to me too  
spoofable to run with, so I agree with the W3C's feedback. People who  
want to us their own resolver can get their own domain under xri.net,  
I presume --- which allows coordination with whoever's running XRI  
proper without full centralisation. Unless you have already snagged  
the TLD .xri, I'd treat it as a distraction, and leave it out.

The "schema that dare not speak its name" appraoch for abstract xri's  
is a realistic solution; but abstract xri's do still need to be  
disambiguated on occasion (even if the syntactic sugar of XRI usually  
gives them away). Can a subdomain urn still be used to refer to xri's  
in the abstract? If so, I'd include an example --- and the prominent  
disclaimer that there is no expectation of resolvability of XRI URNs,  
resolution occurs through one of the other bindings.

--
Dr Nick Nicholas;       Business Analyst, Link Affiliates, Melbourne
http://www.opoudjis.net   skype:opoudjis   opoudjis@optushome.com.au
"Most Byzantine historians felt they knew enough to use the optatives
  correctly; some of them were right." --- Harry Turtledove.