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

Re: Let's stop feeding the NVidia cuckoo



On Fri, 4 Mar 2005 12:13:06 -0500 Luke Schierer wrote:

> assuming there is a meaningful definitionof source for that package.

I claim there is at least one: the one found in GPLv2.

> But as this thread has amply shown, it is entirely possible to come up
> with things that have no meaningful source.

I don't agree.

> Pictures, graphics in general.

Source for graphics is the preferred form for modification, as for other
cathegories of works.

On a case-by-case basis, source may be coincident with the form
exploited when the work is /used/:

 * scripts
 * PNG or even JPEG images that are modified in that very format
 * plain text files that are modified as such
 * ...

or may be distinct:

 * programs modified as C/C++/Java/Fortran/... code
 * PNG or JPEG images that are modified in layered form
 * plain text generated from DocBook source
 * ...

> sufficiently complex programs that require knowledge that 
> the average person doesn't have to understand the function of.

No, IMHO, knowledge is not part of source code.

Of course, it would be highly desirable that upstream author points you
to (or even writes him/herself) DFSG-free documentation[1] that teaches
the difficult topics you have to be familiar with in order to hack with
the program, but that would anyway be a plus (nice but not required for
the software to be DFSG-free).


[1] There is fairly too few DFSG-free documentation around, so it's
    always desirable that someone writes or spreads some!  ;-)

> 
> we've even seen someone suggest that if a programmer for some unknown 
> reason *wants* to write in machine language, he or she can never make 
> his program free software!  to me, this means that we are taking the 
> necessity of source code too far for it to be useful, important, or 
> meaningful.

I'm not the one that suggested that.
I agree with you that for a program written in machine language, source
code and executable form are coincident, as long as the program is
actually maintained and modified in that form.
Note that this is thoroughly consistent with the "preferred form for
modification" definition of source code found in the GPLv2 text.

> 
> cvs or other repositories are nice and all, as is knowing the history 
> of a project, what patches have been proposed, rejected, accepted, so 
> on.  and on the other hand, yes a sufficiently skilled person can edit
> a binary. but the importance and value of free software/open source 
> software necessarily comes between both extremes.  you simply cannot 
> require I distribute intangibles like my knowledge of gaim and its 
> history, or Sean Egan's knowledge of yahoo authentication, you can 
> require that he and I distribute the code we actually edit, in 
> whatever form we edit it, if we want to call it free software.

I agree.
Preferred *form* for modification, not preferred *brain* for
modification!

> but if
> we choose to edit it as a jpeg, or in machine language, or in hex, 
> that should not be sufficient to make our work, under the same license
> it was the day before we made that commit, giving you the same rights 
> it did the day before, the same access it did the day before, 
> non-free.

Again I agree.
And again it follows from the "preferred form for modification"
definition.

> Gaim has actually had hex (though never, to my knowledge, 
> true machine language.) in it, we did an april fools joke that way 
> once.  a hex string is substantialyl harder to edit than the 
> corresponding ascii that does the same thing, does that make that copy
> of gaim non-free? No, our copies had it in hex as well, some gaim 
> developers just happen to have the ability to read a hex string, your 
> lack of that ability does not restrict your rights.

If that hex string was actually modified (or going to be modified in
case a modification was needed) in that form, then, yes, it was source.

Hard to read source, but still source.


-- 
          Today is the tomorrow you worried about yesterday.
......................................................................
  Francesco Poli                             GnuPG Key ID = DD6DFCF4
 Key fingerprint = C979 F34B 27CE 5CD8 DC12  31B5 78F4 279B DD6D FCF4

Attachment: pgpicMAFglq65.pgp
Description: PGP signature


Reply to: