> Jay seems to be just unhappy
> with IPP, suggesting perhaps that it is not good for anything (because
> to many compromises were made?)
Aw now, c'mon. I have pointed out several specific concerns I
have with IPP--in the role of a host-to-device protocol--on this
DL. Perhaps you've missed them?
I think a lot off pro-IPP folks are perhaps missing the point,
at least with regard to my concerns, and probably Paul Moore's
concerns.
Paul has repeatedly pointed out the problem of asymmetry of
IPP-over-HTTP in the context of host-to-device support. This
is a big concern for me, also (both of us being "driver"
developers that need a good "host-to-device" protocol).
The lack of symmetry in the protocol makes life unnecessarily
difficult to develop robust, highly functional drivers. For
example, during the job submission transaction with the printer,
the driver needs to be able to asynchronously receive events
from the printer that may cause a change in the transaction,
or at least cause the driver to "pass along" information to
the originator in a timely and reasonable simple manner (eg,
"Paper out", "PDL error", etc.)
IPP as currently defined (over HTTP) does not provide this
kind of protocol interaction...at least not very easily.
Or more specifically, not as easy as it should be when
compared to similar protocols in the Internet.
Hopefully Paul Moore will comment a bit about this,
but at least this describes one of my concerns about IPP.
Please don't say "Jay doesn't think IPP is good for anything".
...jay
----------------------------------------------------------------------
-- JK Martin | Email: jkm@underscore.com --
-- Underscore, Inc. | Voice: (603) 889-7000 --
-- 41C Sagamore Park Road | Fax: (603) 889-2699 --
-- Hudson, NH 03051-4915 | Web: http://www.underscore.com --
----------------------------------------------------------------------