> At 10:06 PM 7/17/97 -0400, JK Martin wrote:
> >Yes, it does appear that we have been in violent agreement all along,
>
> >that everyone now believes the target job URI must be specified at
> the
> >IPP protocol layer, regardless of the transport used.
> >
> >It also sounds like we have consensus in support of the proposal
> >by Paul/Bob (that launched this interesting thread).
> >
> >If we could only resolve all IPP issues this quickly!
> >
> > ...jay
> >
>
> Hating to spoil the party, but is everybody now clear on which address
>
> goes into the HTTP as server address? I am not!
Using URI strings as object identifiers within the application/ipp body
does not meanwe cannot take advantage of the use of URI targets and
parameters at the transport
level. As in the current specification, the same URI specified within
the application/ipp
body, may or may not be the same URI as specified in the HTTP request
and/or
response.
In the case of the original issue with the CANCEL operation, it could
very
well be the same URI at both the transport level and within the
application/ipp
message. IMHO, I think of the CANCEL operation as being applied to a JOB
object,
so it kinda makes sense to "send a CANCEL message to a JOB object". If
we issue
a CANCEL operation to the PRINTER object, it doesn't seem very easy to
understand,
given the object-oriented nature of the model document.
Just my $0.02
Randy
>
>
> Carl-Uno