On Tue, Jan 06, 2004 at 12:02:45PM +1100, Craig Sanders wrote:
> One thing that all of the advocates for dumping non-free have in common is a
> complete disregard for the actual contents of non-free.
This statement is without foundation, and probably unfalsifiable (as you
are not telepathic).
Unfalsifiable statements have no utility as premises for a practical
argument (as opposed to a formal one), because their truth cannot be
determined. Practical arguments require not only that their reasoning
be cogent and valid, but that their premises are factual.
> they like to pretend that it's all proprietary software, that it
> doesn't even come close to free, that source-code isn't available.
This statement is without foundation. Cite evidence of an advocate of
removing non-free misrepresenting the availability of source code.
Moreover, since I advocate the resolution, and since I know that source
code is available for the packages in non-free in many (but not all)
cases, your argument ("One thing that all of the advocates for dumping
non-free have in common") commits the fallacy of composition[1].
> Nothing could be further from the truth.
The truth value of an unfalsifiable statement, or of a fallacious
statement that is without foundation, is indeterminate, and not of
utility in practical reasoning -- consequently this statement is null.
> while there are a handful of packages
> in non-free that don't have complete or usable source code,
While imprecise (I'll assume "a handful" is something less than 50%),
this statement is not particularly objectionable apart from its lack of
foundation (you have not enumerated which packages in non-free have
incomplete or unusable source code).
> and even fewer that don't have any source code,
Again, lacks foundation, but not otherwise objectionable.
> the vast majority of software in non-free is there because the license
> doesn't quite meet the requirements of the DFSG,
Actually, by definition, *all* of the software in non-free is there
because the applicable licenses don't meet the requirements of the
DFSG.[2]
> just as much GNU documentation does not quite meet the requirements of
> the DFSG.
Indeed; once a distinguishing criterion is defined, that some things
satisfy it and others don't is a truism.
Your second paragraph does not appear to raise any points under
contention.
> The majority of programs in non-free come with source code and allow the user
> to modify and use it as they like.
Again, lacks foundation, but not otherwise objectionable.
> However, some prohibit commercial exploitation or sale, some prohibit
> distribution of modified versions, some prohibit use by government
> agencies, some allow free use only for educational or private
> purposes.
You cite no examples (and thus provide no foundation), but it is true
that all of these are ways to fail the DFSG.
> some of it is affected by software patents, so it is "free" in certain
> countries but "non-free" in others.
You cite no examples (and thus provide no foundation), but it is true
that the presence of a patent on software that is incompatible with the
DFSG renders the software unable to be legally or non-tortiously used in
conjunction with all of the freedoms under the DFSG.
> In short, almost all of the software is almost-free or (using RMS'
> terminology) semi-free software.
I take it your definition of "almost-free" is as follows: prohibits
commercial exploitation or sale; prohibits distribution of modified
versions; prohibits use by government agencies; prohibits use for
non-educational or non-private purposes; or is restricted in any way by
patents.
> Debian doesn't distinguish between the types of non-free...
That is apparently true. I know of no nontechnical position statement
issued by the Project that attempts distinguish among varieties of
non-freeness and by whose definitions anyone is bound.
> whether it is non-free because it is proprietary
What is your definition of "proprietary"? Some would define it as
"prohibits commercial exploitation or sale; prohibits distribution of
modified versions; prohibits use by government agencies; prohibits use
for non-educational or non-private purposes; or is restricted in any way
by patents", among other restrictions.
Because you offer no definition for this term it is difficult to
understand how you use it as foundation for further argument.
> or non-free because use by spammers is prohibited, it is treated the
> same: if we can distribute it at all, it can go in non-free.
It does seem to be the case that any package which is not DFSG but still
distributable, at least in certain countries where prominent Debian
mirrors reside, can be distributed in the non-free section.
> if we can't distribute it under any circumstances, then we can ignore
> it.
More precisely, if we cannot legally or non-tortiously distribute it
under any circumstances, then we endeavor not to do so.
> Aside from the convenience for our users, this has also been useful in
> motivating some software authors to get their programs out of the non-free
> ghetto by changing the license to one that is truly free. there have been
> numerous examples of this happening over the years.
Of which you cite exactly none. Consequently these statements are
without foundation.
> As well as software, the non-free section also contains documentation and data
> that, for one reason or another, does not have a DFSG-compatible license.
You distinguish "software" from "documentation" and "data" here without
defining any of these terms, whereas before your argument did not rely
on making any such distinction -- "software" could be interpreted as
synonymous with "the contents of a package".
> the reasons for this are pretty much the same as for the software -
> prohibition of sale or commercial exploitation, discrimination
> for/against a class of users, no right to distribute derived works,
> etc.
Actually, the reasons are identical[2].
> This non-free data & documentation can still be used and even modifed by the
> end-user, however,
Just as (some) non-free software can still be used and even modified by the
end-user.
> and the fact that modified versions can not be redistributed really
> makes NO PRACTICAL DIFFERENCE to anyone at all.
This statement is without foundation, and probably unfalsifiable, as I
suspect your definition of "practical" is "that which serves my
argument". Feel free to prove me wrong by defining it.
> no one really needs to modify doc-linux-nonfree-text, or povray-doc...
This statement is without foundation. You cannot know the minds or
needs of all people who have ever used doc-linux-nonfree-text or
povray-doc.
> ...any possible need to modify can easily be worked around with an
> errata sheet, or by submitting a change to the authors.
How is the same not true of "software", as distinguished from "data" or
"documentation"? All three of these are frequently distributed with
errata sheets, and submission of changes to the authors are known
practices for all. For examples, I refer you to the Debian Bug Tracking
System[3].
> ditto for data sets like the 'yale' star catalog.
So, no one really needs to modify doc-linux-nonfree-text, povray-doc, or
yale.
> this is the sort of data that end-users don't need to change
This statement is without foundation. How do you know what all end
users' needs are? Alternatively, please define the scope of users'
needs, in your view.
> and, more importantly, where changes should be managed by qualified
> experts.
This is a value statement that should probably be looked at more
closely. You posit that there is some data, publicly distributed, which
should be controlled by a restricted class of persons?
What sorts of data are these? What criteria have we for identifying
this data when we encounter it, and what means have we for identifying
legitimate modifiers?
Is it your position that the only criteria and means we require is an
assertion of copyright?
> for some data sets it is a Good Thing that there is only one official,
> authoritative (preferably peer-reviewed) source.
If it is preferable that there is only one "source", then Debian should
not be distributing it at all -- unless perhaps, the data set is a work
product of the Debian Project. Have you any such data in mind?
I should further note that DFSG-free licensing is not incompatible with
authentication of data. Releases of the Linux kernel, for example, are
commonly digitally signed using public-key cryptography, and yet the
Linux kernel is licensed in a manner compatible with the DFSG[4].
Therefore you have posited a false dilemma[5].
> The proposal to remove non-free is nothing more than rabid ideology,
> completely divorced from reality. if it succeeds, it will be a
> triumph of insanity over reason.
Strangely, I am unable to identify a specific fallacy for this sort of
name-calling. Perhaps you have mistaken it for cogent reasoning; given
the nature of your contributions to the Debian mailing lists over the
years, this certain seems possible.
Actually, I see why. This isn't an argument at all, it is just a set of
stridently expressed assertions, with no rationally persuasive value --
though, of course, it may inflame the emotions of people who are already
inclined to feel as you do.
> a far better use of everyone's time would be to:
This statement presumes that you know better than everyone else how
their time should be spent. Touting the superiority of one's own
intelligence is a poor substitute for reasoned argument.
> - write DSFG replacements for non-free software (or encourage & assist others
> to do the same).
I agree that this is a worthy activity, but it does not preclude the
cessation of support for the non-free section of the Debian archive.
> - try (politely!) to convince the non-free authors to change to a DFSG-free
> licence.
I agree that this is a worthy activity, but it does not preclude the
cessation of support for the non-free section of the Debian archive.
> - get a life
Belittling your audience is a poor substitute for reasoned argument.
> and stop worrying about what other people run on their own
> computers.
This statement is irrelevant to the discussion. The destruction of
software such that it is henceforth unavailable from any source has not
been proposed (and would be infeasible even if it had been).
Finally, it is worth pointing out that the subject line of your message
appears to have little relation to the body; it promised "one reason",
and instead what was delivered was a large number of statements which
were either groundless, not in dispute, or simply abusive of those who
do not share your opinions.
On a personal note, you continue to underwhelm with your demagoguery and
carelessness of thought (which I believe I have pointed out above). I
hope that one day you will be a better exemplar of the Debian Project.
[1] http://www.swif.uniba.it/lei/foldop/foldoc.cgi?composition+fallacy+of
[2] http://www.debian.org/doc/debian-policy/ch-archive.html#s-non-free
[3] http://bugs.debian.org/
[4] ...apart from some bits that we strip out. I am assuming that the
utility of signing kernel releases is not restricted to those
non-DFSG-free bits. If you refuse to stipulate this, please let me
know.
[5] http://www.nizkor.org/features/fallacies/false-dilemma.html
--
G. Branden Robinson |
Debian GNU/Linux | If existence exists,
branden@debian.org | why create a creator?
http://people.debian.org/~branden/ |
Attachment:
signature.asc
Description: Digital signature