Re: PROPOSED: interpretive guidelines regarding DFSG 3, modifiability, and invariant text

[Richard Stallman <rms@gnu.org>]

      Such material may not exceed 16 binary kilobytes (16,384 bytes)
    when viewed in plain-text form (treating all adjacent white space
    characters as one byte).

GNU Emacs comes with more than 16k of such material.  Much more.  And
that is not even counting the material that is part of the Emacs
Manual which is included in GNU Emacs.

It is arbitrary and useless to have any rule about the size of the
invariant material.  The *kind* of material is the important thing.
Technical material that users are entitled to modify must not be
invariant.

    4) For works whose primary purpose is not as instruction code for a
    computer[3], a percentage of the work considered as a whole is permitted
    to be held non-modifiable by the copyright holder(s); this proportion
    must not exceed 5%.

This quantitative requirement is also arbitrary and useless.

	    A) They must be incidental to the content of the work itself.
	    Notices of authorship, acknowledgements, and editorial markup
	    are examples of such incidental material.  A good rule-of-thumb
	    for judging whether material is "incidental" might be: "Is the
	    value of the work in its own right diminished if this material
	    is excised?"  Consider how the work may be republished after it
	    falls into the public domain when making this determination[4].

This is the real issue, and I think you are on the road to a good
criterion.  "Incidental to the contents" is not quite clear enough
though.

For instance, people might not consider the GNU Manifesto to be
incidental.  But the GNU Manifesto is incidental *to the technical job
that the Emacs Manual does as documentation*, which is to say how to
use Emacs.  That, I think, is what has to be expressed as clearly as
possible.