I think that's the best we can do if we limit ourselves to the existing=
Web
infrastructure. Meanwhile, implementers will find ways to do asynchron=
ous
notifications outside the IPP box, as proprietary extensions. Later, t=
he IPP
v2 working group can look at what works and select or synthesize some a=
pproach
as standard for IPP asynch notification. And by then maybe the Web won=
't be so
asymmetrical.
Anyway, polling might not be elegant, but I think it can do the job. W=
ith
HTTP/1.1, the client will be polling over a persistant connection,
<RKD> So you assume that we would keep a connection open until
<RKD> a print job is completed and a notification provided? I
<RKD> don't know if I'd agree that this is a good idea.
so the
overhead should be low. Polling rates could be fairly slow, since ther=
e's no
need for instantaneous notification with an application like printing.
-Carl
ipp-owner@pwg.org on 02/04/98 01:53:15 PM
Please respond to ipp-owner@pwg.org @ internet
To: ipp@pwg.org @ internet, Carl Kugler/Boulder/IBM@ibmus
cc:
Subject: RE: IPP> Notifications
No argument at all. This is othogonal to the XML debate.
I am looking at expanding IPP to cover many needs beyond the relatively=
simple feature set currently defined, the extensibility issue led me to=
the
XML proposal, the unsolicited message issue led me to this thread.
The point I am making is that using HTTP asymmetrically (i.e the client=
always POSTs, the printer always listens for POST - which is the 'natur=
al'
use of HTTP) precludes the core IPP protocol from generating asynchrono=
us or
unsolicited reverse messages. This is a major limitiation - I want to b=
e
sure that everybody knows that we are doing it and that we all accept t=
he
trade-off. I'm sure we could invent lots of hacks later on the will wor=
k
round this but that's not an ideal solution. What will actually happen =
is
that we will all poll :-(
> -----Original Message-----
> From: Carl Kugler [SMTP:kugler@us.ibm.com]
> Sent: Wednesday, February 04, 1998 9:40 AM
> To: ipp@pwg.org
> Subject: Re: IPP> Notifications
>
> Paul-
>
> I would like to point out that the XML/new method proposal is no bett=
er in
> this
> respect. The problem is not that IPP is asymmetric: the underlying =
HTTP
> transport layer is asymmetric, and that is common to both approaches.=
>
> - Carl
>
>
>
> ipp-owner@pwg.org on 02/03/98 12:24:44 PM
> Please respond to ipp-owner@pwg.org @ internet
> To: ipp@pwg.org @ internet
> cc:
> Subject: IPP> Notifications
>
>
> Has anybody noticed that IPP will be useless for notifications due to=
the
> asymmetry of the protocol? As currently constituted a printer cannot =
send
> an
> unsolicted message to anybody.
>
> Was this discussed later on on the Thursday brainstorm?
>
>
>
=