All,
I agree with Michael that we ought to use SHOULD instead of MUST if we
want this to be an editorial change; using MUST seems to require a new
WG Last Call.
Carl-Uno
-----Original Message-----
From: Michael Sweet [mailto:mike@easysw.com]
Sent: Thursday, March 30, 2000 8:07 AM
To: McDonald, Ira
Cc: 'harryl@us.ibm.com'; Hastings, Tom N; anthony.porter@computer.org;
ipp@pwg.org; venky@teil.soft.net
Subject: Re: IPP> Document-format attribute.. [ipp-mod] clarification
"McDonald, Ira" wrote:
>
> Hi folks,
>
> I agree with Harry that we should further revise this paragraph
> to indicate that a client MUST specify a particular document
> format when known and MUST NOT use 'application/octet-stream'
> instead.
Um, that probably won't work too well, since many printer-specific
data streams do not have registered MIME types (e.g. Canon, ALPS,
EPSON, Lexmark, etc.), and a generic print server (e.g. JetDirect,
Axis print server, etc.) probably won't know enough to be able to
enumerate the supported MIME types for the actual device.
SHOULD and SHOULD NOT are probably more appropriate if we are
trying to "encourage" rather than "enforce".
Also, the application/octet-stream information should probably be
updated to reflect a special case for printer objects that list
only application/octet-stream for document-format-supported.
That is, if a client knows the MIME type but the printer object
only supports application/octet-stream, then the printer object
is just acting as a "dumb" printer buffer and the client must
only use the default document format or pass
application/octet-stream.
-- ______________________________________________________________________ Michael Sweet, Easy Software Products mike@easysw.com Printing Software for UNIX http://www.easysw.com
This archive was generated by hypermail 2b29 : Thu Mar 30 2000 - 17:12:34 EST