On 2003-09-28, Barak Pearlmutter <[EMAIL PROTECTED]> wrote:
>> > - No other free software organization eschews such snippets.
>>
>> I disagree with the premises of those two, as well. For instance: no
>> other free software organization edits out the non-free fonts from
>> XFree86 or the non-free firmware from the linux kernel; and this seems
>> like a relatively minor change, as changes in Debian go.
>
> I'm not sure I follow your reasoning there. Your email address
> implies that you are associated with a math department, so let me
> phrase this in mathematical jargon: a proof of this form
>
> A -> B
> A -> C
> Therefore: B and C are the same
>
> is not valid.
I am not claiming that non-free firmware or non-free fonts in XFree86
are the same as unmodifiable snippets, only that your point above ("No
other free software organization eschews such snippets") does not
serve to distinguish them. (Well, the FSF would eschew those bits in
the kernel or XFree86 as well, but they don't actually do the work of
separating them out.)
But in any case, I think I missed making my larger point: in the
section I quoted, you were arguing that we should avoid taking out
snippets because it is a lot of work. To me, that is not an argument:
either it's wrong or it's right (per the DFSG/courtesy to authors) to
include snippets, and if it's wrong we should remove them when we
notice them. If it's very wrong to include snippets, we should in
addition go through the major work of identifying them all, but that
doesn't follow from just the assertion that they violate the DFSG.
Compare this with the situation with non-free code. It is nearly
certain that there is some code that violates the DFSG in main; do you
want to do an audit of all packages to root it out?
I'm much more interested in the arguments why it's a good idea in the
first place to include the snippets than in these arguments about how
much work it would be to remove the unmodifiable snippets.
Peace,
Dylan
