🚀 go-pugleaf

In article <58801@bbn.BBN.COM> pplacewa@bbn.com (Paul Placeway) writes:
>karsh@trifolium.esd.sgi.com (Bruce Karsh) writes:

>On the other hand, "Look and Feel" is a recent, and not fully
>established concept.  Further, it is not yet clear if there is any
>protection for pure algorithms.

It's new for software, but it's been in effect for other areas for a long
time.  For instance, music copyrights.  Also, there is an analogue to
look and feel protection in publishing.  You are not allowed to steal
someone else's book by making simple changes to names, places, etc.  If
you write substantially the same book, even if not one name is the same,
you will infringe.

>The problem is that usually when one of these patent-pending good
>ideas appears in a journal, there is no warning that this might be
>propriatary.  Other people will start to use it as if it was free
>(because they don't know any differently -- it was PUBLISHED after
>all), and then they find out later, AFTER THEIR PRODUCT DEPENDS ON THE
>GOOD IDEA, that they can no longer use it (or SOMETIMES, but not
>ALLWAYS) that they can but have to pay some amount.

OK, I've been saying lots of good things about patents.  This is a problem
area.  The problem is that the patent office is overworked and they don't
have enough money to hire all the examiners that they need.  It often takes a
couple of years for a patent to be issued.

>So now the original person gets BOTH rewards -- one for making the
>idea available, and one for keeping it propriatary.  This, I feel, is
>not fair to the rest of the community.

This seems OK to me.  It's good to make ideas be public.  It's OK for
the inventor to be honored for doing this.

>A (perhaps radical) solution to this would be a small change in the
>patent laws: if you publish before the patent is GRANTED, you loose
>any claims to the idea.

Wouldn't it be adequate to require that the article state that a patent is
pending?  This seems like an improvement to me.  I kind of wonder if you
want to hold up publication until after the patent is granted.  In the
first place, it could be discovered by somebody else in the meanwhile and
in the second place others could use the information in the disclosure to
advance the field.

>Journals could also help out by requiring the
>author to sign away any propriatary claims to the ideas described.
>Perhaps some organization similar to the FSF should start such a
>journal.

Why should an inventor have to sign away propriatary claims in order to
publish.  The inventor doesn't have to publish at all.

>A few things to keep in mind: if the protection you desire were
>allready in place 5 years ago, your company probably would not exist,
>because the command structure and interface of UNIX would be
>propriatary to AT&T.

Ack!  I hope that this posting is not representative of the views of the
opponents of software inventor's protection.  I'd hope that the discussion
could be made on the basis of understanding of the issues, not on the basis
of misconceptions.

UNIX is proprietary to AT&T.  We, like other UNIX system vendors, pay a
license fee to AT&T to be allowed to use their invention.

>If using recursion as a programming technique were patented, the
>renewal of that patent would expire about 4 years form now.

This is false.  There's prior art for recursion going back at least to
the fifties.

>Quick Sort, Heap Sort, Shell Sort, etc. would all be patented.  YOU
>COULN'T SORT DATA WITHOUT PAYING A ROYALTY.

This is false too.  If they were patented when they were discovered, the
patents would have long since expired.  Further, in order to have a situation
where "YOU COULDN'T SORT DATA WITHOUT PAYING A ROYALTY", all the sorting
techniques would have to be patented at the same time.  Since there are
still undiscovered sorting methods out there, it'd be difficult to imagine
a hypothetical situation in the past where sorting was totally locked up
with patents.

I think that the discussion of the merits of patents can proceed without
any upper-case fear-mongering.  If software patents should be done away
with, it's not for some simple and obvious reason.  Other fields have
been patenting things for a long time and there have been no disasterous
effects.

>Associative arrays (which appear in many languages and _are_
>databases) would be patented.  Maybe you could use them, maybe not.

There's prior art for this back to the 50's.

>I could go on.

Spare me.

>  The point is if the kind of protection that Bruce
>seems to want were actually in place, we wouldn't be working on
>computers at all, because they MIGHT have progressed past the early
>60s by now, but you certianly wouldn't be able to afford to buy one
>yourself.

Computers have always been protected by patents.  There's no evidence that
software patents will dramatically increase the cost of computers.  More
likely, they'll decrease costs, as they have in other fields of invention.

>The software industry was working along JUST FINE in 1982.  If it
>ain't broke, don't fix it.

You must be kidding.

			Bruce Karsh
			karsh@sgi.com