Subject: Re: "On Virus" -- get real
From: "Stephen J. Turnbull" <>
Date: Mon, 4 Oct 1999 15:45:44 +0900 (JST)

CC's trimmed to FSB, as promised.

>>>>> "Andrew" == Andrew C Greenberg <> writes:

    Andrew> Right, from now on, let us use terms such as: "GPL version
    Andrew> 2, clause 6, line 3"

That was part of my point, and no, I don't think anyone, even you,
will adopt that suggestion, which is most of the rest of my point.

It takes superhuman restraint to resist catchy phrases and metaphors.
But those are, at least by my thinking, already in the domain of

    >> What "proffered standard of proof"?  Are you stereotyping the
    >> whole marketing community?

    Andrew> This one, your words, not mine:

    >> It is _you_ that suffers the burden of showing that the
    >> terminology you advocate that _we_ use (of course, you may use
    >> whatever terms you please yourself) will have the connotations
    >> we wish to project to an audience of those who are

You implied this is a bad standard; I still see nothing wrong with it,
as a _partial_ standard.  I did not intend that it be the sole

I would like an analogy which expresses "clause 6" as precisely as
possible without the positive and inaccurate connotations that
"partnership" brings with it, and without the negative and inaccurate
connotations that "viral" brings with it.  But I am willing to accept
a somewhat inaccurate metaphor (and the burden of explaining the
inaccuracy as I see it), in order to gain the benefits of using
terminology with wide currency in the community.  This is one place
where "marketing" enters: in the judgement whether the level of
inaccurate connotation is unacceptable for one's purposes.

And I see nothing wrong with using words that connote properties
beyond the technical (here, legal) ones.  For example, my objection to
Bruce Perens' use of "partnership" is not that legally there is no
partnership present, although that's the way I phrased it.

I think that the GNU GPL does help promote social partnership in some
limited situations (eg, where the downstream developer can use it to
convince his employer to donate the modifications back to the
community rather than reinvent the whole wheel for the sake of keeping
"knobby tires" proprietary).  But I oppose using that word because the
GPL a priori seriously restricts the options of all entrepreneurs who
don't already believe in the GPL.  Using "partnership" to describe the
nature of clause 6 has the unpleasant consequence that those who work
in BSD projects or X11-based projects are somehow junior partners:
they are not allowed to use my (GNU GPL-licensed) code, although I can
use theirs.  It is that connotation that decided me.

There is nothing ethically wrong with such restrictions, and I use
them and intend to continue using them myself; it simply isn't
"partnership with the public" to my mind, it's a way of restricting
derivation from my software to people who already think like me in
this respect, or are willing to adopt the GPL in one case despite
preferring something else in principle.  I will not use that word.

    >> Or do you think that it is unnecessary that we consider the
    >> interpretations our audience may assign?

    Andrew> I think we need to use words that communicate meaning.
    Andrew> Period.

Sore de wa, nihongo de mo yoroshii deshou ka.  [In that case, I
suppose we can just as well continue in Japanese?]  An extreme
example, but I do not see where you can draw a dividing line in the
continuum running from different languages, to different dialects,
different accents, and on to "merely" different life experiences.

    Andrew> Maybe we are not far off.
    Andrew> What I advocate is whether the audience will properly
    Andrew> understand the meaning of our words.

Me, too.  But...

    Andrew> My suggestion is this: marketing considerations are
    Andrew> irrelevant where words adequately convey meaning.

A given phrase may "adequately convey meaning" in your daily
professional life; it certainly does not in mine.  Even mathematics
does not communicate unambiguously, without context.

    Andrew> I have been able to articulate the whatever-it-is property
    Andrew> of GPL with the term "virus," without having people shy
    Andrew> away in fear of catching cold.

No, but some of them[1] do "catch" the idea that the GNU GPL is a
Movement, complete with "Massacree" and five-part harmony, out for
World Domination Now, by "infecting" unwitting BSD programs.  It's not
and it doesn't; it's a legal tool for authors to preserve public
access to their works even in modified or derivative form, and must be 
adopted intentionally.  But those connotations remain common.

    Andrew> Ideally, I would like one that is neutral in the lights of
    Andrew> those who oppose the use of "virus."  I asked for thoughts
    Andrew> and advices in that regard.  I'm not really interested in
    Andrew> finding good marketing terms.

What's wrong with "GNU GPL version 2, clause 6," then?  I don't see
how you can avoid use of that perfectly accurate phrase, or of the
actual license language itself, without injecting yourself into the
realm of marketing.  You may not be intentionally "doing marketing"
(although some in this thread certainly are), but you are subject to
the same considerations whenever you use a metaphor.

[1]  I assume not those listening to you.  I would expect that you
explain where the analogy falls down.  In particular, that you can't
"catch GPL" without explicitly including text from a GPLed source,
which is different from "catching RSA" by reinventing the algorithm.
Or maybe your audience (eg, IP lawyers) already understands that.

But others are not so careful.  In fact, some are intentionally "doing
marketing" of various positions.  As am I; although I believe that I
am looking for the most "value-free" terminology possible, I know that
Richard Stallman and Bruce Perens consider some words I have found
useful to be a distortion of the most important aspects of the GNU
GPL.  Better (for me, YMMV) that I acknowledge the nature of what I am
doing than to pretend that I am "simply searching for truth."

University of Tsukuba                Tennodai 1-1-1 Tsukuba 305-8573 JAPAN
Institute of Policy and Planning Sciences       Tel/fax: +81 (298) 53-5091
What are those two straight lines for?  "Free software rules."