RE: CPIM and URIs (summary)

On 29 May 2002 15:00, John D. Ramsdell wrote:
> Graham Klyne <[email protected]> writes:
> 
> > I'm not sure that there is a compelling reason to base the syntax on
> > the email documents.
> 
> Graham,
> 
> As far as I am concern, your syntax is fine, or a mailto like syntax
> is fine.  What is really important to me is this group reaches a rough
> consensus on the precise syntax of both instant inbox and presence
> URIs, and that that concensus be reflected in the CPIM document in
> clear, consistent text.

If my memory serves me correctly, the group reached a rough consensus on
this issue months (years?) ago. Trying to find a reference to this in
the archives is likely to be a bit of a needle in a haystack search so I
didn't look yet. From what I remember, there was a discussion about the
many intricate ways of encoding e-mail addresses which made parsing
addresses pretty awful and how we wished to avoid that.

On 29 May 2002 14:26, Graham Klyne wrote:
> My inclination would be to define an email-address-like syntax within 
> the
> CPIM specs, allowing UTF-8 in the local part (also the domain part if
IDN 
> looks like going that way), and applying minimal restrictions to make
it 
> URI-friendly.  Something like my previous message.

I think this is pretty much along the lines that I remember - if the
possibility of % encoding and @ character makes it difficult, why don't
we exclude it? Do we really need to allow @ within an address?

Adrian.




  [reminder: [email protected] for non-technical discussions, please]