🚀 go-pugleaf

Mark Harrison (harrison@necssd.NEC.COM) wrote:

> If someone who does speak for the FSF can clarify their position, I 
> would appreciate it.

And Paul Eggert (eggert@burns.twinsun.com) wrote: 

> Could you please clarify this in a later posting, or if you prefer,
> private email?  I'm interested in both your and the FSF's positions...

Okay, I asked and here's the official word (from Richard Stallman and Bob
Chassell of the FSF) on where the FSF stands on quoting non-free materials.  
First rms, then Bob Chassell, then my thoughts:

***********************************************
> From rms@ai.mit.edu  Sun Jun 24 12:57:54 1990

Pass this on if you want.

I am the person who commented out the references to the non-free
reference manuals that had originally been put in the GAWK manual.

We thanked Kernighan in the GAWK manual because we are grateful for
his help with GAWK.  We did not recommend that people buy the books he
worked on because they cannot be part of the documentation of the GNU
system.  No matter how well they were written, they are not available,
because the authors chose to make it illegal for GNU users to copy
them.

I don't see any contradiction between being grateful to a person for
one helpful action and not being grateful for another action which is
not helpful.  It is possible for a person to do some things right and
other things wrong.  I think it is a good thing to be able to judge
these actions separately instead of lumping them together.

	Why is it OK for GNU DIFF's README to cite a ``non-free'' journal, but
	not OK for the gawk manual to cite a ``non-free'' book?

One reason is that the former citation is not for purposes of
documenting the system for users.  Our documentation must stand on its
own.  If it isn't sufficient, then we must improve it, not refer
people to something that we cannot distribute along with copies of
GNU.

	Doesn't it hamper the free flow of information to refuse to cite work
	that is freely available in any good technical library?

The word "hamper" is not appropriate if we merely decide not to
publish certain information.  It is normal and inevitable for authors
to decide what they wish to say and what they wish to leave out.
Unless the author has a near monopoly on what people read, this is not
something that affects the free flow of information in general.

It is true that our decision will not give certain people some
information they would like.  However, the goal of the FSF is to
change society so that more information will be free in the long run.
Being maximally obliging to everyone's immediate wishes is not a
concommitant of this goal.  For the long run, we must encourage free
documentation, not non-free documentation.

	Should giving proper credit to other people's technical contributions
	depend on agreeing with their politics?

I think this question is a red herring.  As far as I know, the the
GAWK manual gives credit to everyone who contributed significantly to
GAWK, regardless of their political beliefs (which I generally did not
ask) and of their other actions.  In particular, it gives credit to
Brian Kernighan.  However, if you know of anyone specific who deserves
credit for contributing to GAWK but is not acknowledged in the GAWK
manual, please tell me.

****************************************************

>From labrea!grackle!bob@ai.mit.edu Wed Jun 27 09:52:26 1990

   > @comment We don't refer people to non-free information
   > @c We don't refer to hoarded information.
 
Here is a case where I agree with RMS:

    We did not recommend that people buy the books he worked on
    because they cannot be part of the documentation of the GNU
    system.  No matter how well they were written, they are not
    available, because the authors chose to make it illegal for GNU
    users to copy them.

Perhaps this fuller explanation should be included in the source.

   J. Greely writes:
   > harrison@necssd.NEC.COM (Mark Harrison) writes:
   >>I personally feel very bad on the Kernighan's behalf.  After putting
   >>so much work into his books (every one, IMO, a classic) and setting
   >>a new standard for excellence in programming books, and then graciously
   >>helping the gawk authors to copy his work, and then being called an
   >>INFORMATION HOARDER!!!!

But his books are hoarded.  I have looked at their copyright pages
in the bookstore.  Truth may be unpleasant, but it is truth.

Also, someone said

    Doesn't it hamper the free flow of information to refuse to cite
    work that is freely available in any good technical library?

I myself have never had good access to a good technical library except
for a short period in my youth.  (For practical reasons, I do not have
ready access to MIT libraries even now.)  Most people who will use GNU
will not have access to a good technical library.  To presume
otherwise is to discriminate against the majority.

RMS says:

    the goal of the FSF is to
    change society so that more information will be free in the long run.
    Being maximally obliging to everyone's immediate wishes is not a
    concommitant of this goal.  For the long run, we must encourage free
    documentation, not non-free documentation.

I agree.

*********************************************

And now me.  Here's my thoughts on the matter, a bit different from the official
``company'' view:

Paul Placeway (pplaceway@bbn.com) wrote:

> The entire essance of scholarly publication is an attempt to distribute
> information that can be used freely.  The fact that the physical
> manifestation (book, journal, etc.) may not be free is easily solved
> by going to a library.  The physical AWK book isn't free, but then
> neither is the screen that you are reading my words off of, nor the
> paper you put into your printer.  Neither, for that matter, is a GNU
> source tape (yes, you can FTP GNU stuff for free, but your network
> connection isn't free either).

That's an argument to run by rms.  Why not write him and ask him?  (His address
is rms@ai.mit.edu).
 
> On the other hand, if information in a work was used in another piece
> of work (including the implimentation of a program), the author has an
> obligation to cite all sources used.  To do otherwise is somewhere
> between sloppy and plagurism.

If the cited work was NOT used to help write the GAWK manual, and the author in
question simply felt that knowledge of such work would be useful to the GAWK
reader, then the citation should be placed in a ``Further Reading'' section and
not in the body of the main manual.  The FSF has no obligation to provide a 
section on further reading, nor any obligation to include non-free sources in
this section.  A section like that is up to the FSF folks to include.

However, if the cited piece was used to help write the GAWK manual, then it
*should* be cited (whether it is free or not) in a bibliography-type section.  Of
course, rms will argue that non-free sources should NOT be used to assist in
writing any FSF manuals (an argument I've already had with him).  After all, the
FSF is re-engineering sources from *scratch* and shouldn't be looking at the
original sources at all.  Therefore, there is no need to look at non-free
sources...(and 'round in circles we go :-)

As Ozan Yigit (oz@yunexus.yorku.ca) so elequently put it:

> This issue, however, is only peripheral to the issue of giving *proper
> credit* where one is due, whether that be books, personal communication,
> source code, whatever. So the question really is *whether or not the awk
> book is used as a reference* during the development of GAWK, rather than
> *the status of information* contained therein. If it *was* used, than the
> least I expect is a proper reference: it is my opinion that the quality
> of a tool, or its public usefullness does not absolve the authors from the
> responsibility towards other works that made the tool possible.

Now I didn't write the cited passage, so I can't tell you whether or not the AWK
book was used to help write the GAWK manual or not.  Arnold Robbins has already
said how he feels about this, and I think I'll let him have the last word:  :-)

> As I said above, I wasn't out to spread useless and irrelevant personal
> opinion.  Had the gawk manual been something being published as a regular
> book, it would have been appropriate to cite the other references.
> 
> That the commentary got left in is something of an oversight, but there
> are other things in the manual that are commented out because more work
> needs to be done on them or they are vestigial.  A manual is like a program
> in that it evolves over time.  How many of you have programs with NO code
> commented out?  Right, I thought so. :-)
--
Diane Barlow Close
	close@lunch.wpd.sgi.com
	also 	close@cygnus.com
	I'm at lunch today.  :-)