Path: utzoo!utgpu!jarvis.csri.toronto.edu!mailrus!csd4.milw.wisc.edu!uakari.primate.wisc.edu!ark1!dsill
From: dsill@ark1.nswc.navy.mil (Dave Sill)
Newsgroups: gnu.misc.discuss
Subject: Re: Questions
Message-ID: <55@ark1.nswc.navy.mil>
Date: 4 Aug 89 15:32:22 GMT
References: <115@isgtec.UUCP>
Reply-To: dsill@relay.nswc.navy.mil (Dave Sill)
Distribution: gnu
Organization: Naval Surface Warfare Center, Dahlgren, VA
Lines: 70

In article <115@isgtec.UUCP> robert@isgtec.UUCP (Robert Osborne) writes:
>Some comments/questions:
>    :
>2)   What if I think up some new, neat-o-keen, idea|algorithm (Fat Chance :-)
>     *patent* it and then use BISON/G++ to implement it,  I then distribute
>     my source.   Would not any future distribution of source be made
>     illegal by my patent.  Or would I lose this patent (by my subsequent
>     distribution of same source.).  Am I stupid/naive for asking
>     this question?

Your patent would be available to anyone wanting a copy.  If you want
to keep it secret you'd have to make it a trade secret.  (This is what
Coca Cola uses to protect its formula.)  You'd have to make everyone
that needed to know the algorithm sign a non-disclosure agreement.  Of
course any reasonably knowledgeable person could disassemble your
binary to discover the algorithm; just as chemists could theoretically
reverse-engineer Coke's formula.  I guess you'd also need to copyright
your algorithm, but I think there are circumstances under which
copyrighted material must be made public.

>3)   What ARE the *better* ways of making money from our software (assuming
>     we don't sell it, but instead distribute it).

I believe RMS' idea is that programmers should earn money doing
consulting and supporting.  A possible scenario would be that your
company would pay you to write the software (which would be
copylefted).  They'd distribute it; possibly bundled with a
maintenance agreement or warranty which would allow them to charge
whatever they want for it.  Their customers would have access to the
source, so they would have the option of purchasing maintenance or
support elsewhere.  The "danger" in this scenario is that your company
would be required to give its software to its direct competitors.
This is offset by the fact that the competitors would also have to
give your firm any enhancements or fixes they distribute to their
customers.  Another offsetting factor is that your firm, as the
developer of the software, would get a big jump on the competition in
the span between its first release and the time any competitor would
be able to distribute an enhanced version.  And of course, by that
time your firm would probably already be shipping their next version.

>4)   I'm a software hoarder (oh no!) and I work for a company (see .sig)
>     that provides hardware and software to create 3d images from CAT scans.
>     (simplified version of what we do,  *I disclaim*).
>     The development required from my department, has so far taken
>     approximately 100 man months of effort.
>     (This is only a small section of our company the whole effort is much
>     larger)   The users of out product are radiologists and surgeons.
>     Verrrry few of these people are computer literate,
>     (with some exceptions), nor do they desire to become so.
>     Where would hospitals (society) find 100+ man months of *volunteer*
>     FSF type of help to create, support, maintain such a software
>     project?   How does this kind of project fit into the FSF world
>     view?

See my response to #3.  Nobody's suggesting that programmers quit
earning money for writing software.

I'd suspect that in a vertical market such as 3D CAT imaging, it
wouldn't matter a whole lot if your company freely distributed its
sources.  Who would your customers buy support/maintenance from?  At
first, your firm would be the only choice.  Is there any proprietary
hardware bundled with your products?  If so, what good would your
software do your competitors, especially considering that they'd have
to "show you theirs" after you "showed them yours"?

>Robert Osborne                     ...uunet!mnetor!lsuc!isgtec!robert
>                                          ...utzoo!lsuc!isgtec!robert
>ISG Technologies Inc. 3030 Orlando Dr. Mississauga. Ont. Can. L4V 1S8
-- 
Dave Sill (dsill@relay.nswc.navy.mil)