Newsgroups: comp.sys.amiga
Path: utzoo!utgpu!news-server.csri.toronto.edu!mailrus!ames!decwrl!bacchus.pa.dec.com!granite.pa.dec.com!mwm
From: mwm@raven.pa.dec.com (Mike (Real Amigas have keyboard garages) Meyer)
Subject: Re: thought Dillon and friends would like to know
Message-ID: <MWM.90Aug3161400@raven.pa.dec.com>
Sender: news@wrl.dec.com (News)
Organization: Missionaria Phonibalonica
References: <1990Jul27.204559.26305@athena.mit.edu>
	<1990Jul29.024841.1000@agate.berkeley.edu>
	<90210.100908GRGREF@BYUVM.BITNET> <10743@wehi.dn.mu.oz>
Date: 3 Aug 90 16:14:00
N-Reply-To: BAXTER_A@wehi.dn.mu.oz's message of 2 Aug 90 12:49:50 GMT
Lines: 84

In article <10743@wehi.dn.mu.oz> BAXTER_A@wehi.dn.mu.oz writes:
   > Actually, a patent *is* protection for ideas...a copyright is protection
   > for an implementation. 

   CRAP. A patent is protection for the owner (not necessarily the inventor)
   of an INVENTION. An IDEA cannot be patented in the US unless it has also
   been IMPLEMENTED. Then, it is the IMPLEMENTATION which can be patented
   IF it also meets other requirements (like not being public knowlege, being
   inventive (!), etc).

While the original comment wasn't quite correct, it was a lot closer
than what you're spouting. Copyright covers implementations. Patent
covers "inventions", which is a pretty broad area. The invention
doesn't even have to be implemented, just implementable - as well as
meeting those other requirements you mention. Once it's patented,
independent re-implementation - even if they aren't identical - are
infringements of the patent.

Copyright protection is relatively weak, covering only one
implementation (at least, that used to be the case). Patent protection
is very strong, covering any implementation, and in many cases even
similar products (consider the Bell patent on the Telephone, or the
Selden (non-)patent on the automobile).

   An algorythm per se can not be patented. Was probably is patented is the
   implementation of the algorythm in modems. (I don't know, I'm just guessing).
   In which case, there is no problem.

This is out of date. Up until about 1981, software could not be
patented at all. At that time, the Supreme Court decided (in Diamond
vs. Deihr) that inclusion of a computer program did not mean automatic
rejection of a patent application. The patent office took this to mean
that you could patent software. There are over 2000 patents covering
software, and nothing but software. Examples include:

#4,736,308 covers nesting scrollable objects inside other scrollable
objects. Apple is being sued for violating this one with hypercard.

#4,197,590 is the infamous "X-or cursor" patent, covering X-oring a
cursor onto the screen to make it visible against all backgrounds, and
for easy restoration of the screen. DEC (among others shipping X) is
being held up for license fees on this one.

#4,633,416 & 4,602,286 cover computerized graphics technics including
airbrushing, stenciling, and combining two images under control of a
third.

#4,398,249 covers "natural order recalc" in spreadsheets. If a company
sells a spreadsheet, they've probably got a lawsuit on their hands
over this one already.

#4,555,755 covers storing the obscured parts of windows in overlapping
window systems for quick restoration. AT&T is threatening to sue
everybody distributing X for this one (I wonder if I could purchase
the rights to use that in the Amiga market, and make everybody who
uses SMARTREFRESH windows pay me 1% of gross sales - including CBM!
Would beat the heck out of actually _working_ for a living).

#4,558,302 covers compress. My understanding is that there are
actually three patents on that algorithm, and Unisys is demanding 3%
of gross sales from anyone who wants to license it. Hey CBM - you
gonna ship BRU with compress turned on? And is MRBackup going to have
that feature vanish?

IBM has patents (#4,656,583 being the one I've got listed) on compiler
optimizing technics, including register coloring and computing
available expressions.

#4,742,450 covers "shared copy-on-write segments".

I understand that IBM also has a patent covering "stored-program
computers", but that was usually attached to a story of them only
dragging it out to wave at people who wanted them to pay a license
fee. Wander what's going to happen when they meet someone like REFAC
(holders of #4,398,249) who doesn't create anything but licenses &
lawsuits).

	Still depressed by the whole thing,
	<mike
--
Tried to amend my carnivorous habits			Mike Meyer
Made it nearly seventy days				mwm@relay.pa.dec.com
Loosing weight without speed				decwrl!mwm
Eatin' sunflower seeds
