Bill Mills wmills@invisibleworlds.com
Mon, 18 Jun 2001 13:50:02 -0700

I was in fact thinking of a super-handler type thing.  My thought was that 
it would be worthwhile to reduce the required number of handlers, especially

when we have the addidtional vector of transport type (tcp/sctp/...).


-----Original Message-----
From: Eamon O'Tuathail [mailto:eamon.otuathail@clipcode.com]
Sent: Monday, June 18, 2001 1:22 PM
To: Bill Mills; bxxpwg@lists.invisibleworlds.com
Subject: RE: [BXXPwg] SOAP over BEEP


RFC 2396 - section 3, first paragraph - The scheme name is up to the first
':'. Most pieces of deployed software only look at that and then branch off
to a handler - and these handlers are usually configurable based on the
scheme name. They will give you the <scheme specific part> (from just past
to the first ':' to the end).  If you plan to use a ':' separated scheme
name, then you will need some sort of super-handler that is smart enough to
work out the second part of the scheme name. If you use the '.' notation
(e.g. "soap.beep"), then you can directly activate a suitable soap handler.

>> does this indicate that the algorithms
>> currently in use do not conform to the RFC per se?

Which algorithms are you talking about?
To which specs / drafts do they belong?


BXXPwg mailing list