> -----Original Message-----
> From: Carl Kugler [mailto:kugler@us.ibm.com]
> Sent: Friday, June 05, 1998 5:04 PM
> To: ipp@pwg.org
> Subject: Re: RE: IPP> Implications of a new scheme, etc
>
>
> > Hi Roger,
> >
> > I've got some comments on this. (I dont imagine your surprised :)
> >
> > For the IPP: (new scheme proposals)
> > I think putting 'no impact' for proxies is 100% inaccurate.
> > a new IPP scheme will break *every* existing Proxy. I challenge
> > the wg to find a proxy which will pass this exception, if
> one exists.
>
> Josh, I think you've misread the chart. For the proposal
> that puts the "ipp:" scheme ON THE WIRE, we listed "Breaks
> most installed proxies. At best proxies have to be
> reconfigured".
Ok, I would make this stronger. It breaks all proxies.
It wont be a matter of reconfiguration. No proxy I know
of today can process this without a software change.
At best, the ipp: scheme will need to mapped to
a scheme that the proxy knows about, which these days
is one of :
http://
ftp://
gopher://
and maybe wais or mailto:
> The column that says "No impact" for Proxies
> is for the Larry's proposal
> (http://www.findmail.com/list/ipp/3754.html) which is
> actually HTTP on the wire. (See
> http://www.findmail.com/list/ipp/3785.html for > more info.)
>
I dont think larry's proposal meets the requirement that
keith is asking for. Since the client will just map ipp -> http
on the wire and the proxy or firewall will see it just as HTTP.
This doesnt allow the proxy to differentiate IPP from HTTP
in this case, which was what keith seems to be looking for.
> -Carl
>
> >
> > In the case of the new method, most current proxies will be
> > able to handle it with minimal effort, as you indicated, some will
> > handle it as shipped (MSproxy) and some will need a patch. (squid)
> > If this is a holdup for a new method, I volunteer to submit
> > a patch to the squid group to fix this.
> >
> > To use a new scheme means that proxies must understand the
> > IPP protocol inner workings (which means that it has to know
> > that its really just HTTP on the wire). To use a new
> > method means that IPP is a service on HTTP that is identified
> > by its different method (PRINT).
> >
> >
> >
> > > -----Original Message-----
> > > From: Roger K Debry [mailto:rdebry@us.ibm.com]
> > > Sent: Friday, June 05, 1998 1:00 PM
> > > To: ipp@pwg.org
> > > Subject: IPP> Implications of a new scheme, etc
> > >
> > >
> > > As suggested on Wednesday's teleconference, Harry Lewis,
> > > Carl Kugler, and I produced the attached table (in .pdf format)
> > > which hopefully summarizes the many views which have been
> > > expressed on this subject over the last couple of weeks. Our
> > > intent is that this would help support our position with
> Keith Moore.
> > >
> > >
> > >
> > > Roger K deBry
> > > Senior Technical Staff Member
> > > Architecture and Technology
> > > IBM Printing Systems
> > > email: rdebry@us.ibm.com
> > > phone: 1-303-924-4080
> > >
> >
> >
>