[Date Prev][Date Next] [Thread Prev][Thread Next] [Date Index] [Thread Index]

Re: Inconsistencies in our approach

On Mon, Aug 04, 2003 at 12:17:09PM -0500, Branden Robinson wrote:
> On Sun, Aug 03, 2003 at 02:10:37AM +0200, Sergey V. Spiridonov wrote:
> > If one does not see the difference between program and documentation, it 
> > is very hard to explain why they do not need the same kind of freedoms.
> If one cannot coherently and usefully *describe* the difference between
> programs and documentation, it is difficult for other people to see it.

Documentation consists of instructions primarily intended to be
human-readable regarding the operation of something such as a program.

Programs consist of instructions primarily intended to be machine-readable
that either contain machine language binary data or instructions designed to
be interpreted or converted into that at runtime.  Programs will always
contain source code or machine language code, and often both.

I will grant that these definitions are imperfect and improbable arguments
could be lodged against them; at the same time, I believe that reasonable
people not engaging in a Jesuit exercise to find logical needles in a
haystack of common sense are able to tell the difference between a manpage
and a C source file.

> I continue to suspect that people are indulging an Aristotelian
> categorization fetish solely as a means to an end, that end being to
> compel the Debian Project to ship their favorite w4r3z in main, heedless
> of the negative consequences to the freedoms that our users currently
> enjoy.

Actually, my goals are the opposite.  I see it as intellectually and
logically dishonest to claim certain requirements for some types of
non-program data in Debian, other requirements for other data, and do it all
under the guise that "everything binary is software."

I have a far smaller problem with people designing the Debian Free
Documentation Guidelines or the Debian Free Data Guidelines, and
implementing them fairly even to the exclusion of RFCs and GFDL than the
problem I have with people attempting to contort the Debian Free Software
Guidelines into something that covers non-software, because this attempt is
fraught with confusion and incoherency.

> After all, what utility would this distinction serve beyond providing
> one a means of routing around the DFSG's inconvenient restrictions?

The DFSG's restrictions prove inconvenient to those on your side of the
fence, too.  After all, if you claim that all documentation is software,
than you are ignoring the restriction in DFSG #2, which states:

   2. Source Code 
   The program must include source code, and must allow distribution in
   source code as well as compiled form.

There is neither source code nor compiled code for my King James Bible in
free-form ASCII text.  For we have a definition of source code as:

>From The Free On-line Dictionary of Computing (09 FEB 02) [foldoc]:

  source code

     <language, programming> (Or "source", or rarely "source
     language") The form in which a computer program is written by
     the programmer.  Source code is written in some formal
     programming language which can be compiled automatically into
     {object code} or {machine code} or executed by an


There is no formal programming language that KJV is written in, and nothing
that can compile it or execute it.  Moreover, there *never will* be anything
that can do that, because the KJV is not a set of instructions[1].  Other
books, such as works of philosophy or detective novels, similarly are not
instructions, and neither is the Mona Lisa, even if it is saved as a BMP.

The point is that this is not an exercise on my part to let less-than-free
bits into Debian.  It is rather an attempt to call a horse a horse, rather
than engage in this business of calling everything software.  I think we
have a clear weakness in the DFSG and a clear need for some guidelines for
non-Software components, and I advocate that.

I also advocate equal treatment, both over our entire archive and over time. 
(That is, an unmodified copy of something that was free in 2000 should still
be free today unless we have changed our definition to exclude it.)

We distribute the GPL all over on our system, and at the very top, it says:

 Everyone is permitted to copy and distribute verbatim copies
 of this license document, but changing it is not allowed.

Under the DFSG, this would fail.  If you call the GPL "software", then you
have declared the entirety of the most important pieces of a Debian system

If you do not call the GPL software, then why is documentation software?

Why is it OK to include the GPL in our system but not other bits of

Reply to: