Re: draft-ietf-impp-im-01 (was: Re: WG Last Call on multiple docu ments (deadline Jan 12))

Graham,

Thursday, January 9, 2003, 2:38:30 AM, you wrote:
Graham> At 08:01 AM 1/7/03 -0800, Dave Crocker wrote:
>> >> So, a megabyte string of random 8-bit values is a valid TransID and we
...
>>Any other syntactic limitations, such as text versus binary?

Graham> As a touchstone, I'd suggest thinking in terms of a URI containing an SHA-1 
Graham> hash.  Strictly, that's probably overkill, but if we ever get to a point of 
Graham> wanting or needing to generate TransIDs randomly, then not by so much.

explicitly specifying a way to ensure uniqueness sounds like quite a good
idea.

an alternative algorithmic approach that is used in email is to have the ID
comprise a domain name and a clock tick from that domain, where the domain
guarantees that it has issued that clock tick (for this purpose) only once.

d/
-- 
 Dave <mailto:[email protected]>
 Brandenburg InternetWorking <http://www.brandenburg.com>
 t +1.408.246.8253; f +1.408.850.1850




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