Article View: gnu.misc.discuss
Article #73952Re: Patent on a standard-byte order for intermachine communication
From: Don Wells
Date: Mon, 15 Jul 1991 23:29
Date: Mon, 15 Jul 1991 23:29
76 lines
5106 bytes
5106 bytes
In article <gnusenet9107131607.AA14833@mole.gnu.ai.mit.edu>
rms@gnu.ai.mit.edu (Richard Stallman) writes:
> US patent number 4,956,809 covers using a single standard byte
> ordering for transfer of data between machines whose normal byte
> ordering is different. Unless invalidated (by prior art from
> mid-1982 or earlier), this patent prohibits the principal means for
> transfering data between disparate types of computer systems...
> If you find prior art dated 1982 or earlier, please inform
> league@prep.ai.mit.edu.
During the 70s the worldwide astronomy community had a problem with
interchanging digital data, especially digital images, between
observatories due to software and hardware incompatibilities. After a
series of prototyping experiments, an interchange format agreement was
negotiated between three major observatories, and subsequently this
agreement was formally endorsed by the International Astronomical
Union as the recommended practice for all of astronomy worldwide.
During the 80s the format design was extended to cover the interchange
of additional types of data objects (e.g. tables), and was adopted by
numerous projects and agencies. This interchange format is named
"FITS", the "Flexible Image Transport System".
The FITS Agreement was negotiated in March 1979, interchange was
demonstrated in April-May 1979, and a technical report on the
interchange trials was presented to an international image processing
workshop in June 1979. In March 1980 a paper was submitted to the
journal Astronomy and Astrophysics Supplement Series and, after
refereeing, it was accepted for publication in September 1980, and
appeared in the June 1981 issue of that journal:
D.C.Wells, E.W.Greisen, R.H.Harten, "FITS: A Flexible Image Transport
System", Astron. Astrophys. Suppl. Ser. vol.44, pp.363-370, June 1981.
This paper is still considered to be the formal, official definition
of the FITS standard. The abstract begins: "A format for the
interchange of astronomical images and other digital arrays on
magnetic tape is described..." Section 2 ("The physical
representation of the data", on p.364) begins by discussing the use of
a special record size to "[facilitate] packing and unpacking the data
on a wide variety of computers." (The logical record size of FITS
files is 2880 bytes, a number which is "evenly divisible by both the
byte and word lengths of all computers that have been sold in the
commercial market." If you think that this is an irrelevant
consideration, you should consider that the very first FITS file
passed in interchange in 1979 from an IBM 360 to a CDC 6400, and that
the bytestream of that file is still readable by FITS readers
installed in astronomy software everywhere.) The section then proceeds
to discuss the conventions to be used for file structures and for
pixel ordering within files. The second half of the third paragraph
discusses the representation of the pixel values:
"...We will support three types of pixel value representations: 8-bit
unsigned binary integers, 16-bit twos complement signed binary
integers, and 32-bit two-complement signed binary integers. We require
that the 2 or 4 bytes of 16 or 32-bit pixel values be written on the
tape in order of decreasing significance. The byte that includes the
sign bit will be first, and the byte that has the 1-bit of the 16 or
32-bit integer as its least significant bit will be last. Thus, users
of the DEC PDP-11 series will be obliged to correct for the
'byte-swap' action of that series' architecture."
This sounds like "prior art" to me. A reprint of the 1981 paper will
go out in tomorrow's mail addressed to RMS at the League address.
Acknowlegement: It is a pleasure to thank Chris Flatters
(cflatter@nrao.edu) for pointing out the possible prior art aspects of
the 1979 FITS Agreement.
--
Donald C. Wells Associate Scientist dwells@nrao.edu
National Radio Astronomy Observatory +1-804-296-0277
Edgemont Road Fax= +1-804-296-0278
Charlottesville, Virginia 22903-2475 USA 78:31.1W, 38:02.2N
Technical Details
Message-ID:
<DWELLS.91Jul15182940@fits.cx.nrao.edu>Path:
arc1-ca.pugleaf.net!gmdzi!unido!unidui!math.fu-berlin.de!ira.uka.de!sol.ctr.columbia.edu!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!agate!stanford.edu!mcnc!uvaarpa!polaris.cv.nrao.edu!polaris.cv.nrao.edu!dwellsReferences:
<gnusenet9107131607.AA14833@mole.gnu.ai.mit.edu>Show Full Headers
Complete Article Headers
X-Google-Language: ENGLISH,ASCII-7-bit X-Google-Thread: 1025b4,9a4c815047ed049b X-Google-Attributes: gid1025b4,public X-Google-ArrivalTime: 1991-07-15 22:11:41 PST Path: arc1-ca.pugleaf.net!gmdzi!unido!unidui!math.fu-berlin.de!ira.uka.de!sol.ctr.columbia.edu!zaphod.mps.ohio-state.edu!cis.ohio-state.edu!ucbvax!agate!stanford.edu!mcnc!uvaarpa!polaris.cv.nrao.edu!polaris.cv.nrao.edu!dwells From: dwells@fits.cx.nrao.edu (Don Wells) Newsgroups: gnu.misc.discuss Subject: Re: Patent on a standard-byte order for intermachine communication Message-ID: <DWELLS.91Jul15182940@fits.cx.nrao.edu> Date: 15 Jul 91 23:29:40 GMT References: <gnusenet9107131607.AA14833@mole.gnu.ai.mit.edu> Sender: news@polaris.cv.nrao.edu (News Manager) Followup-To: gnu.misc.discuss Organization: National Radio Astronomy Observatory, Charlottesville, VA Lines: 75 In-Reply-To: rms@gnu.ai.mit.edu's message of 13 Jul 91 16: 07:45 GMT