On 22-Mar-1999, D. Tweed <[EMAIL PROTECTED]> wrote:
> On Fri, 19 Mar 1999, Fergus Henderson wrote:
> 
> > Generally programming languages themselves are always free, i.e. very
> > few people have ever tried to copyright a language, and when they have,
> > the courts have for the most part rejected such attempts (e.g. see [1]).
> > It is of course possible to trademark the _name_ of a language, as for
> > example Sun have trademarked the name "Java", and as was the case with
> > "Miranda".  It is also possible to patent techniques that might be
> > required to implement a language.  And finally you can of course copyright
> > the language's reference manual.  However, the law doesn't really provide
> > any form of intellectual property that can cover the language itself.
> > Copyright only protects expression of an idea, not the idea itself.
> 
> Could you elucidate on what constitutes the programming language? I'm
> curious because I'd always understood that there was a patent on the
> syntactic form
> 
> f x y = x + y , if x == y
>       = x - y , if g x
>       = x * y , otherwise
> 
> in functional programming languages held by the person which holds the
> various licences for Miranda. (I've been told this by two people;  however
> I've never actually seen any published reference so I could be wrong.)

I too would like to see a published reference for that patent.
Personally I doubt that such a patent would be valid.

One of the major problems with the current patent system is that
the patent office frequently issues patents for things which are
not really patentable.  Once a patent has been issued, the
patent owner can often bully people into accepting the patent,
by threatening them with legal action; this works because few
will be willing to risk challenging a patent in court.

> However there's not a problem with the isomorphic Haskell
> construct
> 
> f x y | x==y = x + y
>       | g x  = x - y
>       | otherwise = x * y
> 
> because it's the syntactic construct that's covered, not the idea of
> `equations defined by clauses with side conditions'.

There's several different kinds of patents.  The usual kind are
called "utility patents", but there's also another kind called
"design patents".  Whereas utility patents protect the functional
aspect of an invention, design patents protect the ornamental aspects.
(The other kind of patents in the USA are the new "plant patents",
allowing patenting of genetically engineered plants, but that is
not relevant here.)

So if the above syntactic form were patentable, it would have
to be a design patent.  However, in US law, a design patent protects
a "new, original and ornamental design for an article of manufacture",
and I don't think a program or a programming language qualifies as
an "article of manufacture".  The courts have ruled that typefaces
don't qualify in that category, so I find it hard to imagine that
computer programs could qualify.  On the other hand, the exact
law varies from country to country, so the situation might be
different elsewhere.

If the patent mentioned above does exist, wouldn't the program
Miracula (a free implementation of a subset of Miranda) violate it?
I couldn't find Miracula on the WWW anymore, perhaps this patent
is the reason why?

-- 
Fergus Henderson <[EMAIL PROTECTED]>  |  "I have always known that the pursuit
WWW: <http://www.cs.mu.oz.au/~fjh>  |  of excellence is a lethal habit"
PGP: finger [EMAIL PROTECTED]        |     -- the last words of T. S. Garp.


Reply via email to