Rekall not longer available from theKompany.com - a fabrication

Alex Martelli aleax at aleax.it
Fri Nov 7 15:11:37 CET 2003


Paul Boddie wrote:
   ...
>> licensing scheme. There would be a free for non-commercial use GPL
>> version and the normal commercial version.
> 
> Well, given the history of Rekall, combined with TheKompany's problems
> with selling GPL software because people apparently demanded the
> source code for free without buying anything, it's hard to know

*blink* surely the source has to be given only if the binaries are?
Can't stop others from redistributing that, but "demand the source
code w/o buying anything" doesn't seem something the GPL mandates.


>> The GPL would not include any kind of support whatsoever, therefore
>> you would need to take out a support subscription.
>> The GPL version would be free for non-commercial use.
> 
> You have to ask yourself the question: if your knowledge of the GPL
> was sketchy or non-existent, how would you interpret that last
> sentence?

Presumably in the "obvious way", yes:-).


> Isn't this known as being "economical with the truth", or is it just
> downright contradictory? It's like saying that "nude bathing is
> allowed but you aren't allowed to get wet", only to explain such a
> ridiculous statement away by then saying that "they just forgot to
> mention that you are allowed to get wet as well".

No, your analogy is misplaced: the "aren't allowed" would be
overtly false in your "it's like".

A correct analogy would be: "I never strangle somebody with the
initials PB on a Friday".  This statement _is_ perfectly true: I
never strangle anybody at all, and therefore, in particular, not
people with the initials PB, and even more specifically, not
on a Friday,

However, it's _misleading_ because, although Aristotelian logic
has nothing to say in the matter, people "naturally expect" that
a qualification restricting a statement is there for SOME reason:
typically because, without the restriction, the statement would
not hold.  But that's just a heuristical inference based on "if
the restriction wasn't necessary he wouldn't bother stating it" --
there's nothing either dishonest or contradictory in putting on
a statement _more_ qualifications than strictly necessary (it's
the difference between "precondition" and "WEAKEST precondition").

E.g., a similarly misleading statement would be "In Euclidean
geometry, there is one and only one perpendicular to a given
line through a given external point".  The qualification "In
Euclidean geometry" is way stronger than necessary, since
existence and uniqueness of the perpendicular follow from the
first _four_ axioms of Euclid only, _not_ needing the fifth
one which is not valid in non-Euclidean geometries such as
Riemann's and Lobachevsky's.

And yet, unless I have specifically undertaken to tutor
somebody in a field including non-Euclidean geometries, I
might well make such a statement and consider it defensible.
Indeed, I have seen _proofs_ of this very statement, based
on other results which _do_ hold in Euclidean geometry only...
Presumably, depending on context, the responsibility for
learning about NON-Euclidean geometries may be held to belong
to the reader of my statements, without any duty on my part
to instruct said reader in this subject.

Similarly, the asserters of the above statement (falsely
claimed to be contradictory, though truly claimable as being
misleading) might contend that the responsibility for learning
about the GPL belong to their customers, without any duty on
their part to instruct said customers.  I'm gonna stay neutral
on the specifics, but I've always been fascinated by the issues
of NON-weakest preconditions and qualification in many fields
(natural language, programming, and logic) so I couldn't resist...:-)


Alex





More information about the Python-list mailing list