IPP Mail Archive: Re: IPP> Identifying jobs in requests

Re: IPP> Identifying jobs in requests

Randy Turner (rturner@sharplabs.com)
Fri, 18 Jul 1997 01:22:06 -0700

I agree that the 'getjobs' operation should, at a minimum, return the
URI
for any and all jobs. I think this statement pretty much stands on its
own.
The fact that a server may return a secured 'alias' for the actual job
URI
really wouldn't matter to the protocol or the end-user. I do agree that
returning a secured alias would be an interesting way for servers to
secure
access to jobs from certain classes of end-users.

Randy

Robert Herriot wrote:

> > From cmanros@cp10.es.xerox.com Thu Jul 17 18:17:07 1997
> >
> > 1) Either the Printer is open (non-secure), in which case the
> server
> > should return all the information.
> >
> > 2) Or the Printer is limiting access through some security
> restrictions,
> > usually requiring authentication and authorization. In this case,
> >
> > a) the client is either accepted and gets all the information,
> > b) or it does not pass the security check and does not get anything.
>
> >
> > In consequence, I suggest that there is no half-way house cases
> where the
> > server would return some, but not all of the information and we
> should
> > amend the model document to reflect this.
> >
> I disagree, a secure printer could reveal some but not all information
> about
> a job. We have left the degree of information purposely vague so that
>
> a Printer can return as little or as much information as it wants
> based
> on who is asking.
>
> What Paul proposed was that a Printer could not refuse to return a
> job-uri
> with Get-Jobs, though it could return a special job-uri that was
> different
> from what the job owner might get.
>
> Bob Herriot