Some hopefully not so random thoughts.
Although I asked this question at the start of the MIME crisis, I have
seen no answer. What is the use, purpose, advantage or significance
of a MIME type identification to a printer? I do not ask this
argumentatively, but because if there is a proper advantage, it should
be the justification for making printers MIME cognizant. If there is
no advantage, then there is no reason to include MIME type
understanding in the printers or in the MIB.
If there is no purpose to using MIME to identify input types to
printers, why is it imperative that it is used in the IPP? Again, I
think that a reasonable statement of what purpose it serves is
desirable. The statement that it makes IPP compatible with faxes,
e-mail and whatever seems to be not germane because, as far as I
understand, IPP is intended to go to printers. The statement that our
area coordinators require it is, I believe doing them the disservice
of suggesting that they have no valid reason or are disinclined to
share it.
I would protest Steve's suggestion that, 'if a printer has implemented
the Printer MIB and also IPP, that printer must recognize both MIME
and lang enums' by observing that the statement is not applicable in
many cases. The printer MIB is concerned with printer parameters and
the printer must do at least the instrumentation for the MIB. On the
other hand, IPP may (and in many cases will) be done by the NIC or by
some enhanced HTTP server ahead of the printer. (I believe one
contention was that it would be done by an NT system.)
That, of course, suggests an answer the question of the MIME to Enum
conversion: if MIME types have no significance to printers other than
that they are necessary in IPP for other, unspecified non-printer
reasons, it should be the IPP server that does the conversion
(assuming that a simple conversion is possible... which is a different
issue that someone ought to consider).
At any rate, I agree that modification of the printer MIB at this
point is not warrented.
Bill Wagner, DPI/Osicom
______________________________ Reply Separator _________________________________
Subject: Re: PMP> Closure on MIME vs. Enums
Author: Harry Lewis <harryl at us.ibm.com> at Internet
Date: 9/4/97 10:35 AM
I agree with Lloyd and Jay, in principle. I was at the IPP call yesterday and
stated there should be no reason for IPP's choice of MIME representation of
printer languages to impact Printer MIB. Use of enums is entrenched (and
appropriate) in the MIB and double registry already exists (ex: PS - registered
both as MIME and prtInterpreterLangFamily enum with IANA).
The real question is where does the mapping between MIME and enum occur. Here,
you get into a debate over thin client vs. starved printer. This will not
resolve due to the wide variety of platforms, printing system views and printer
controller architectures our community represents.
The only reasonable rebuttal (to my initial premise, above) came from Steve
Zilles who pointed out that, if a printer has implemented the Printer MIB and
also IPP, that printer must recognize both MIME and lang enums. This fact (and
burden) seems unavoidable, given the choice of MIME in IPP. Still, one could
question the usage model that requires MIME in the MIB vs. in the IPP code. I
suppose it would alleviate a (very) simple mapping in the client, should they
uses SNMP for auto-configuration to determine languages supported. This would
seem like a fairly weak argument.
Just in case the prevailing forces (or our own consensus) leads us to include
MIME in the Printer MIB, it would be nice to know how the IETF process would
accommodate this. In other words, if we were to whip up a new, optional list of
prtLangMIMEType, as suggested, would the IETF allow this to slip in without
effecting the status and schedule of the Printer MIB?
In general, I think the greater harm may well be to IPP in that we have chosen
to take a very communications and "host side" view when making decisions,
failing to recognize the cost of storage in embedded devices... the very reason
enums exist.
Harry Lewis
------ Forwarded by Harry Lewis on 09/04/97 07:59 AM ------
pmp-owner at pwg.org on 09/04/97 12:12:43 AM
Please respond to pmp-owner at pwg.org @ internet
To: lpyoung at lexmark.com @ internet
cc: pmp at pwg.org @ internet
Subject: Re: PMP> Closure on MIME vs. Enums
Excellent statement, Lloyd. There is no real justification for the IPP effort
to be impacting the Printer MIB effort, particularly with regard to the
issue of MIME types vs. enum definitions.
...jay
lpyoung at lexmark.com wrote:
> Evidently there is some opinion in the IPP working group that
> the Printer MIB is stalled with the IETF because of the MIME vs
> enum discussion for interpreter types. I wanted to assure
> everyone within the PMP working group that as far as Chris and
> I are concerned this is not the case at all. Chris and I are
> working on taking the Printer MIB as currently defined forward
> to our Area Directors as a Draft Standard.
>> >From the minutes of the last IPP conference call, an issue is
> going to be raised against the Printer MIB in the Atlanta IPP
> working group session to "keep the same Printer MIB registry,
> do not deprecate it, add a new optional, printerLangMIMEType".
> My position is and will be that the Printer MIB is closed and
> moving forward on the standards track. I have not seen enough
> justification to halt this process to add this new object.
> Regards,
> Lloyd
>> ------------------------------------------------------------
> Lloyd Young Lexmark International, Inc.
> Senior Program Manager Dept. C14L/Bldg. 035-3
> Strategic Alliances 740 New Circle Road NW
> internet: lpyoung at lexmark.com Lexington, KY 40550
> Phone: (606) 232-5150 Fax: (606) 232-6740