minor changes to the netfax rfc

Clifford Lynch <CALUR%UCCMVSA.bitnet@uccvma.ucop.edu> Wed, 26 August 1992 20:19 UTC

Received: from ietf.nri.reston.va.us by IETF.NRI.Reston.VA.US id aa07255; 26 Aug 92 16:19 EDT
Received: from NRI.NRI.Reston.Va.US by IETF.NRI.Reston.VA.US id aa07249; 26 Aug 92 16:19 EDT
Received: from stubbs.ucop.edu by NRI.Reston.VA.US id aa17996; 26 Aug 92 16:21 EDT
Received: by stubbs.ucop.edu (5.57/1.34) id AA02598; Wed, 26 Aug 92 12:38:42 -0700
Message-Id: <9208261938.AA02598@stubbs.ucop.edu>
Received: from UCCVMA.UCOP.EDU by uccvma.ucop.edu (IBM VM SMTP V2R2) with BSMTP id 0582; Wed, 26 Aug 92 12:33:58 PDT
Received: from UCCMVSA.BITNET (CAL$UR) by UCCVMA.UCOP.EDU (Mailer R2.08) with BSMTP id 7063; Wed, 26 Aug 92 12:33:58 PDT
Received: by UCCMVSA.BITNET Wed, 26 Aug 92 12:33:52 PST
Date: Wed, 26 Aug 1992 12:33:52 -0800
From: Clifford Lynch <CALUR%UCCMVSA.bitnet@uccvma.ucop.edu>
To: NETFAX@stubbs.ucop.edu
Subject: minor changes to the netfax rfc

I got a couple of suggestions from the folks at the
CMU mercury project about RFC 1314 (the image transmission
format) which I would suggest that we consider incorporating
into any subsequent revision.
They have some considerable experience now with network
image retrieval/display, and looked at the RFC in the context
of some actual library-type applications.


first, evidently TIFF lets you stick the image file
directory anywhere; it need not be at the front of the file.
By convention, it usually, is, but not always. We should
consider requiring it at the front of the file so that
programs processing these things know what they are
dealing with prior to having to buffer the whole image.

second, for random access applications (eg image databases)
things are frequently stored as one TIFF object file for each
page (either as files or database records). I think we might
want to be less strong about our suggestion of using single
large multipage TIFF files and suggest that this choice will
be driven by the context of the specific application.

I don't recall if this RFC is technically standards-track or
not, but if it is we should get some logical revision points
which would let us incorporate these changes if people think
they are a good idea.

Clifford Lynch