Re: [mile] Artart last call review of draft-ietf-mile-rolie-10

[Peter Saint-Andre <[e-mail]>]

On 10/23/17 12:33 PM, Waltermire, David A. (Fed) wrote:
> Ben, Alexey, and Adam,
> 
> Working through Martin's ARTART review of ROLIE, we had the following exchange with Martin.
> 
>>>> From martin:
>>>> Question: it's fairly widely accepted that use of IRIs in Atom has
>>>> been less than successful.  Do you want to mandate use of URIs
>>>> instead?  This would apply to both link relations and the "src" attribute.
>>>
>>> From Stephen:
>>> We wanted to stay in line with the requirements in Atom, if this becomes
>> an issue it may be worth re-examining.
>>
>> From Martin:
>> This is your chance.  You don't get to go back or re-examine once your
>> protocol is widely deployed.
> 
> I understand Martin's point above. It is basically that URIs are more narrowly scoped and interoperable than IRIs are. From my experience I have seen that URIs are broadly used, and IRIs have not seen much use. Given this feedback, it looks like the best thing to do might be to break with ATOM and use the more restrictive URI in ROLIE to ensure the best possible interoperability within resulting solutions.
> 
> What do you think about such a change? Any advice?

The Atom RFCs were published at a time when IRIs were at the height of
their (limited) popularity. Although I don't recall the details, it
might even have been the case that an Area Director said "you really
ought to be using IRIs because they are more general". I don't see a
significant issue with scoping MILE/ROLIE to use URIs, because they are
a subset of IRIs.

Peter