[BXXPwg] proposal for 1:N interactions

james woodyatt jhw@wetware.com
Thu, 31 Aug 2000 12:22:33 -0700


At 22:49 -0700 2000.08.30, Marshall Rose wrote:
>
>hi. i guess i'd prefer to view ANS as 1:N positive responses rather than 1:N
>responses. so, i'd expect that things like
>
>ANS ... ANS ERR
>
>and
>
>ANS ... ANS RPY
>
>would be errors according to the consistency rules.
>
>if i get back an ANS, i'm thinking that things are happening and i'm getting
>back positive stuff.
>
>if i get back an ERR, i'm thinking that nothing happened whatsoever.

So, what do you get if things are happening and you're getting back 
positive stuff, but then something goes wrong before everything is 
done happening?

Seems to me that the consistency rules imply that any indication 
something is going wrong has to be squirreled down in the payload of 
one of those ANS messages.  If that's so, then 'ANS' doesn't really 
indicate a positive response-- just *a* response, which may be 
positive or negative.

This is why I'm asking if I'm perhaps looking for symmetry where it 
isn't appropriate.  I can accept that 1:1 interactions should have 
the RPY/ERR distinction in the header type while the 1:N interactions 
should have no equivalent distinction, but it would be easier for me 
to do so if I understood that this asymmetry was intentional and 
served a purpose.


-- 
j h woodyatt <jhw@wetware.com>
http://www.wetware.com/jhw