Subject: Re: Advantages of freed vs. proprietary software
From: Jonathan Ryshpan <>
Date: Fri, 26 Aug 94 11:57:36 PDT

Some of the folks writing in the recent burst of activity in fsb seem to
be conflating a number of different issues:

    Software copyright
    Software licensing
    Sortware patent
    Interface licensing

This posting gives my ideas about these different issues.  Please excuse
its length.  Please keep these issues seperate in the future (8-)!

Software copyright and licensing present few problems.

    Very few people are opposed to software licensing or copyright.
    The shrink-wrap software industry has brought computer users many
    useful programs at reasonable prices.  Software licensing is a way
    to provide the authors of programs a fair reward for their labors,
    bearing some relation to the cost and value of what they sell.  It
    does not seem to impede the ability of others to write and to sell
    or give away similar programs.

    I am under the impression that fsb exists to discuss the possibility 
    of other business models for the distribution of software.  But this
    doesn't mean that most members of the group believe that the shrink-
    wrap model is "bad" in itself -- only that it isn't the *only* model
    worthy of consideration.

Software patent presents serious problems of several kinds:

    The patent office appears to have granted many specious patents,
    mainly to "inventions" which are obvious to anyone skilled in the

    There is a problem billing for software patents.  A reasonable fee
    can be established for a patent on an end user device or a part for
    which there is a market.  It seems reasonable for a patent holder
    to ask for (say) 5% of sales of (say) nuts with a plastic insert to
    take the place of a washer.  A manufacturer orders (say) 100,000 of
    them; the cost of the patent is bundled with cost of the bolts.  I
    haven't see a lot of orders for (say) 100,000 natural order recalc
    modules or 5,000 register coloring optimizers.

    Software is unusually complex.  In fact complexity is almost the
    only cost of software.  A software project of moderate complexity
    may involve a large number of inventions, many of which are
    reinventions of techniques already patented.

    Software patents affect programs available in source more than 
    those available only in executable, since patent holders can 
    inspect source for patent violations much more conveniently than
    they can inspect the executable.  This ought to be an issue of 
    some importance to members of this mailing list.

    [BTW: Maybe it should be a requirement for holders of software
    copyrights that they make the source available (to licensed users
    only and at an additional, but modest, fee) so as to allow their
    programs to be inspected for patent infringement in the same way
    that "free" software can be inspected.]

Interface copyright is a serious problem.

    Interface copyright was unheard of till the last few years.  By
    definition it affects only arbitrary elements of the interface,
    otherwise they would be immune to copyright as being in the nature
    of things or else would be patentable; therefor it protects only
    those parts of the interface that *don't* involve any creativity.

    Interface copyright protects monopoly and impedes progress.  (More
    info on demand.)

    The basic concept of interface copyright used to be considered so
    undesireable that Maxtor fought and won a case of unfair business
    practice against IBM not for copyrighting an interface (the
    connection between IBM mainframes and disk drives) but only for
    changing it so rapidly that Maxtor was not able to provide reliable
    operations of its IBM "compatible" drives across changes in IBM's
    mainframes.  IBM was forced to pay damages and to publish the
    specs for its interfaces and its plans for future interfaces (as I
    remember it).

If you want to see a good discussion of the arguments against software
patents and interface copyrights, read the League for Programming
Freedom's papers on these issues.  They are available from:
and similar locations on many other servers.

Many Thanks:

Jonathan Ryshpan                <>

	===============> I write for myself ONLY. <===============