Subject: Re: GNU License for Hardware
From: craig@jcb-sc.com
Date: 15 Oct 1999 15:50:15 -0000

>I fail to understand your logic, and you have yet to respond to the 
>hypocrisy allegations I have made against you. Do you or do you not believe 
>in personal freedom when it comes to GNU software? If you truly believe in 
>that freedom, then you must accept that people can and will take the GNU 
>code and incorporate it into their own projects. In doing so, there is no 
>requirement that they name the project after the GNU project (if you 
>believe they must, please point me to the relevant passage in the GPL).
>
>If you insist that people who incorporated the GNU code into their project 
>name their work after the GNU product, then we have a very monumental task 
>ahead of us, that being divesting ourselves of the GNU project for one that 
>truly believes in freedom, since that requirement is completely foreign to 
>any type of freedom that I am aware of.

Here's a clue: RMS *already* believes in and accepts that freedom.  His
"insistence", as you put it, that the operating system built out of GNU
compilers, libraries, linkers, assemblers, text utilities, shell utilities,
documentation utilities, and the Linux kernel, be called "GNU/Linux",
impinges *not one whit* on your freedom to call it whatever you want.

If you claim that his mere "insistence", or request, to call it by a
name that he believes more properly credits Project GNU, constitutes
such a trampling on your "freedom" that you must "divest [yourself] of the
GNU project for one that truly believes in freedom", then by all means
go ahead and do the 10+ years of hard work to realize that dream.

And be aware you're not the first person to threaten to do it in this
sad, silly little spat about naming (the history of which I won't go
into), and you won't be the last.  That's not just a sarcastic slam
on my part -- it's a suggestion you go and locate those other people
who've threatened the same thing, see how they've been doing vis-a-vis
making the threats real, and perhaps enlist them to help you (or you
help them).

Then, after the 10+ years of hard work you, along with all your friends and
supporters do, which will surely be under the umbrella of some name *you*
come up with to denote "true freedom"...

...I am sure you will not matter one whit when someone popularizes a
distribution of your "truly free" distribution under a name that gives
no credit whatsoever to all your work, because they name it after some
crucial component they developed using those tools, even though that
component constitutes about 1% of the total code base on which their
distribution depends.  And you won't matter when most of the enthusiastic
developers that distribution attracts via the one component completely
miss the importance of your contribution, and even make changes to your
code that you can't easily use because they're system-specific, even
when made to areas of your code you had carefully designed to be portable,
causing your own code to appear to have more features on their system
than it does on other systems.

The point I'm making here is not that "GNU/Linux" is a better name than
"Linux" or "Lignux" or "Debian Linux" or whatever.

The point is that RMS, for all his faults (some of which contributed
to this naming controversy, IMO), cares *vastly* more about freedom than
you know or, probably, do yourself.

Why do I say that?  Because, based on your post, all you seem to care
about is to call something what you want, and you have a fit because
somebody (who has *no* legal authority to impose his will on you, something
you are too clueless to understand or too desperate to engage in arguments
to admit) suggests you use a different name.

Whereas, based on RMS' history, what *he's* done is not only *insist*
on freedom, and *write* lots of code (including some famous code, like
GCC and Emacs) that actually *was* free, he's even undertaken the vastly
more difficult task of creating and nurturing a community that *values*
those freedoms, *talks* about them (constantly, if not overmuch ;-), and
defends them even against carrot-style persuasions to give them up in
specific instances for short-term benefits.

In the end, I know that RMS will have contributed far more actual *freedom*
to the computing community in an average *year* of his life than you will
in your entire life, because you seem to care more about whining about
minor things that are done (sometimes necessarily) to achieve major things.
(I'd love to be proved wrong, of course.)

Those major things including ensuring *continuing* support for the
GNU project and its evolution, which is one of *the* most crucial aspects
of making sure Redhat, Debian, SuSE, and all those other Linux-based
systems continue to evolve...and that such support largely depends
on people being aware of the need to provide it, something that the
use of a good name can do more than almost anything else.  Though I
expect most X users know they are using something called X, I very
much doubt most "Linux" users are aware they are using software highly
dependent on GNU components, components that *will* need care and
feeding over the next many years, just like the kernel.  Someday X
will be "just there" for lots of users the way kernels, libraries,
compilers, and assemblers are -- part of the wallpaper underlying
whatever they *think* they're interacting with -- at which point
they might think they're just using Enlightenment, or whatever, and
not realize they're also using X, glibc, Linux, GCC, etc.  (I realize
it's a problem that name games alone cannot hope to solve.)

In the meantime, if you want to really start on the de-GNU-ing Linux
project -- which I wholeheartedly endorse, for reasons going well beyond
promising an end to this naming fiasco (and ridding the GNU project of
much of the resentment directed at it by Linux enthusiasts) -- I suggest
you start by modifying Linux itself so it can be effectively compiled
by *any* C compiler, not just GCC, and getting Linus to accept your
patches.  (It won't be easy, but it's probably the most important thing
to do, in terms of getting widespread acceptance that you *have* created
a truly GNU-free Linux-based system.)

This will be the Nth time I suggest this course of action since pretty
soon after the GNU/Linux naming controversy started so long ago.  Check
the gnu.misc.discuss newsgroup archives to confirm that if you like.
My suggestion wasn't taken up then, it wasn't taken up when I repeated
it on the gcc2/EGCS/GCC mailing lists, and I predict it won't be taken
up now.  (Most, if not all, of my public posts suggesting this course
of action were in threads in which Linus himself participated, so he's
probably read my suggestion several times over already.  Oh, I suppose
my suggestion *has* been taken up, and I'm unaware of it -- but then
so are the people who show up regularly and claim they're going to
*start* such a project, so I'm not the only one.)

OTOH, when the *GNU* project people decided they didn't like the
possibility of having to depend on non-free software to use a
product -- KDE -- they created *two* projects to solve the problem,
Gnome and Harmony.  Even today -- after the Qt licensing problem was
supposedly "solved" -- Gnome is (apparently) chugging along.  (I use
neither it nor KDE at this point, so I don't keep close track.)

In other words: Project GNU is, like it or not, populated by lots of
people who actually *care* enough about freedom to *do* something about
it.  (That's, IMO, partly due to RMS' obsessions about *true* freedoms:
people who don't care about them enough to do something about them are
*likely* to be put off by RMS' obsessiveness.  Though some who do care
enough might be put off by other things and go off and work on other
projects anyway, so I'm not saying *all* freedom-loving people work
for RMS...how scary would *that* be...shudder.  ;-)

And people who are aware that they might well *do* something about
a lack of freedom generally have more than a clue about exactly what
freedom *is*, and what it *isn't* -- because when it comes to *doing*
something, most people become a lot more rational about what is worth
doing, how difficult it might be to do, and what the risks are, than
when they care merely about having something done *for* them by others,
as GNU, Linux, and X was done for you.

(This is, by the way, why I expect 2nd-amendment supporters to win out over
gun-control advocates in the USA -- I believe a much larger portion of
the former are willing to take a bullet for their cause than of the
latter, who generally expect *others*, i.e. hired men with guns, to
do their dirty work.  This is an aspect of something schools don't
teach about democracy: 9 people wanting pepperoni pizza versus 1 person
who wants anchovies does *not* mean pepperoni gets ordered if that
one person is willing to die, or at least kill, for those anchovies.
Those extreme examples illustrate the reality behind Project GNU vis-a-vis
the overall Linux community -- we're in the minority, but we're a
minority that has already *sacrificed* for the freedoms we value,
and fully expect we'll have to continue doing it.  It's therefore
in our best interests to have a clear idea of what freedoms are most
important, what are achievable, etc.  That also gives more focus
to understanding just who are your friends, who are your enemies, and
who are bystanders, which can help make more rational decisions about
who to protect and who to leave hanging out to dry when there are
assaults on various groups.)

I am pretty sure there are those on the Linux project who *also* care
enough about freedom to do something about threats to it.

But I doubt *those* people whine that RMS, when he pushes the "GNU/Linux"
name, is somehow "taking away their freedom".  They might just ignore
his requests, or loudly reject them, but they probably have a better
notion of what "freedom" is than you do -- after all, they've *worked*
for it.

Yes, I'd prefer RMS (and GNU and everyone else generally) respect names
more than people tend to.  The abuse of names (whether to insult others,
as in "MS-DOG" or to claim mindshare that isn't deserved, as in "Windows"
or "linuxexpo.com"), to me, violates the Golden Rule.  It seemed pretty
clear to me that "GNU/Linux" came along too late to change the *actual*
name (as in "the name everyone uses") for "Linux", and, if it hadn't,
the "Lignux" debacle that preceded "GNU/Linux" tipped it over the edge.

That being said, I've felt the "GNU/Linux" naming proposal is, putting
aside the specifics of the campaign, among the *least* offensive attempts
I've seen to change the name of a thing.  (Especially given that,
at least for awhile, apparently, many people who said they were "hacking
Linux" were actually hacking GNU code in Linux-specific ways and not
caring one whit about interoperability with other systems.  I hope that's
no longer the case, if it ever was, but IIRC it was cited as evidence
for the need to make it clear that Linux is basically just a *kernel*
plus kernel-specific utilities, while GNU is a huge collection of codes
that are required to work properly on more than just Linux-based systems.)

But the "GNU/Linux" campaign would probably have worked a lot better if
it had constituted creating an FSF distribution of a complete GNU system
with a choice of Linux (perhaps as well as a BSD or Hurd) kernel and
promoting *that* so strongly that the "GNU/Linux" name would take over
only to the extent that the distribution was well-done.

(It is probably the case that if a few more people had felt "GNU/Linux"
was a name worth carving out mindshare for, there would have been
sufficient volunteer help available to make such a successful distribution
happen.  Maybe there will be someday.  So I'm not saying it's RMS's
job, personally, to make it happen, or even the FSF's, especially since
part of making it happen can include campaigning effectively for the
need in the first place -- which the "GNU/Linux" campaign did, in its
own way.)

Note that the "GNU/Linux campaign", as I am describing RMS's efforts
here, may well have been *designed* to have a positive effect on long-term
efforts to work on GNU, but might have been *executed* such that it
did not.  In particular, my impression is that some potential good workers
might have been so put off by the way it was conducted that, although
they might have gotten a clearer picture regarding which software was
GNU and which was Linux (and which was something else), they decided
against working on the GNU components any further.  That's always a risk
when undertaking a campaign like this anyway, but I think RMS learned
some things about how not to increase the likelihood, and breadth of
impact, of the risk when trying this sort of thing in the future.  Even
in failure, we all have the opportunity to learn lessons.  (And, yes,
some have been put off working on Linux by some things Linus and
others have done and said as well.  I am, myself, vastly less interested
in working on Linux now versus two or so years ago for that reason,
which is kinda too bad since OS/kernel work is one of my top areas of
expertise and I'm no longer working on g77, so I have more free time
to work on other things.)

        tq vm, (burley)