Subject: Re: [Approval Request] Kallisys Reflexive License
From: Abe Kornelis <abe@bixoft.nl>
Date: Sat, 10 Nov 2001 21:12:18 -0800

Paul Guyot wrote:
> 
> >- The text contains a large number of diacriticals
> >   (in the french sections). Though I liked the
> >   bilingual version, my browser died while trying
> >   to print it. Please consider using &egrave;
> >   &eacute; etc. for the characters with diacritics.
> Both versions are bilingual. The American English or French text
> can't be separated. But one is in one column (the one I'm working on)
> and the other one in a single column (French, English, French,
> English, etc.). They're encoded in ISO 8859-1 and they say so in meta
> tags. I wonder what kind of problems you have. I encoded  with an
> HTML entity (i.e. &OElig;) because an old version of Netscape I had
> didn't print it properly. I'm ready to put entities, but the W3C
> recommends to not to when you can send 8 bits texts.
-- I don't know on which character exactly the browser died.
   It *is* a pretty old one (Netscape 3). It works fine with the
   'old' standardized diacritical names that can be encoded in
   7-bit ascii. I don't know about the newer standard for 8-bit
   encoding. I assume it is more efficient with regard to bandwidth.
   If this standard has been approved, I'll not complain.
   Then it simply is my own problem.

> I had a quick look, especially at the interesting (for me) preamble.
> You're right about patents, but this isn't an European problem, and
> probably not a problem at all.
-- Sorry, you're wrong.

> Software cannot be patented in Europe, AFAIK.
-- Wrong again. Has been patentable for over thirty years, even in
Europe.
   The pertinent patent laws claim that software is not patentable,
   but attorneys have find a very simple way around it. Now the de facto
   point of view is: software is not patentable as such, but combined 
   with hardware it's a 'device', which is absolutely patantable.
   And without the hardware, what use is any software?
   It's an old and troublesome misunderstanding, fact remains that
   software can be patented. The european patent office has granted
   thousands of them over the last few years. If you want to make
   certain, check the EPO (European Patent Organization), or check
   the IBM website, which hosts a monstrously large patent database, 
   containing american, european, and japanese patents.

> And if they actually are, it's via weird protection methods.
-- What you call weird is in fact standard legal practice.

> I'm not sure that if you hold a patent (on a technique, since it's
> the object of patents) and give free use (under an open source
> license) of the form of the technique in a software that you can sue
> someone for using this software.
-- You could, if you did *not* include a patent license with the
   copyright license for the software.

> What you can do is suing someone who would have not copied the form
> (you gave a license to) but the technique (you cannot sue him for
> copying the idea in any case, but with a patent, the part of the idea
> which isn't the technique is quite small). The problem is that with
> programs, the form is very close to the technique.
-- That is indeed a problem. Now I'm not a lawyer, but AFAIK there is
   a lot of jurisprudence (Jurisprudentie in dutch, hope english word
   is correct) that defines the border, though it still may be a narrow
   one, there absolutely is a distinction - at least in legal terms and
   practice.
 
> To give an example, let's say that Unisys distributes under BSD some
> C code to compress data with the LZ algorithm. I doubt that they can
> sue you if you use their code (they gave you the right to).
-- They might license the code wothout the patent. That would allow
   you to do anything with it that the copyright license allows, 
   *except* to actually tun the program without first obtaining a
   patent license as well. So you could study it, create derivatives,
   translate, redistribute, etc. etc. But as soon as anyone plans on
   actually using that code, or any derivative of it, then the patent
   license would have to be acquired first.

> I also
> doubt that they can sue you if you translate their C code to pascal
> if the form is respected (except the language of course).
-- That would be regarded as a derivative.

> So in the
> end, I think they can't sue you as long as you're doing computer
> programs out of their code.
-- Sorry don't understand what you're meaning to say.

> I don't know if you had a look at their
> patent, but as a software patent, it includes pseudo code. So if they
> give you a license on a similar code, they still own the technique,
> but no longer this expression of it.
-- That might be, I'm not a lawyer. Still I'd be cautious to use 
   such ciode without first checking whether or not the patent 
   holder agrees that you may run that particular software without
   infringing on their software patent(s).

Regards, Abe.
-- 
Abe F. Kornelis, B.V. Bixoft
Het Jaagpad 63, 3461 HA Linschoten
The Netherlands
phone: +31-6-22755401

To visit our website:
either: http://www.bixoft.com
    or: http://www.bixoft.nl

--
license-discuss archive is at http://crynwr.com/cgi-bin/ezmlm-cgi?3