[Pkg-doc-linux-devel] Re: LDP "non-free" documents

[Rick Moen]

Quoting Colin Watson (cjwatson@debian.org):

> > I have stated that, myself, on quite a number of occasions on LDP's
> > "discuss" mailing list.  I trust and hope that nobody was actually
> > mislead into thinking that nobody ever NMUs or otherwise alters a
> > developer's uploaded package -- having stated the reverse many times.
> 
> I wasn't referring to NMUs at all (although those happen too, of
> course). 

Just a point, since there has been a great deal of reckless and
mischievious inaccurate representation in prior posts (not by you):
I did not say just "NMUs" but rather "NMUs and otherwise".

(I regret having to reiterate what I have _not_ said, as it is
ultimately a poor use of time, but a certain party has debased the
discussion to that level, regrettable.)


> The simple fact is that authority to make licensing decisions
> in Debian does not devolve to package maintainers [...]

This statement in no way accords with my plain reading of Debian
governing documents -- unless one assigns a very peculiar meaning to the
word "decisions".

I have observed how Debian maintainers operate for many years, and it is
very evident to me that they _are_ charged with applying DFSG.  If they
err, their actions are subject to correction at various levels, which,
if they feel are in error, they can counter and appeal in various ways.

Denying that such actions aren't "making licensing decisions" would seem
very odd, but I would rather talk about mechanism and structure, to make
sure Emma Jane and others correctly understand same.


> Package maintainers do the implementation, but that's all.

Doing the implementation happens to be practically everything.
And portraying the maintainers as mere mechanics is obviously very
distortive; they are voting members of a polity with (moreover) _the_
primary responsibility to implement DFSG in their packages.


> Yup. Wasn't disagreeing with you on that front, just placing slightly
> different emphasis. (I've disagreed with debian-legal "consensus" myself
> plenty of times in the past ...)

And what _about_ that "consensus"?  The notion that the Web pages in
question accurately convey any such thing is both (1) entirely
unsupported, and (2) doubtful on the plain evidence of what's in there.

Please just _look_ at both the contents and how those get there.  Our
passive-aggressive friend Mr. Jensen has killed many innocent electrons
protesting the extremely high reliability and authority of those things,
but c'mon, let's be serious:  The uncredited page maintainer simply lobs
together, for each licence pretty much _all_ posted objections that profess,
however improbably, to be DFSG-based, concatenates and summarises them
_uncritically_, and posts them claiming that they are the "consensus of
debian-legal".

There is basically _no reason at all_ to believe that the ftp-master
admins, the DPL, or anyone else in a position of responsibility would
agree with many of those claiims.  It's just _stuff_ that uncredited
persons posted to a public mailing list.

I've already talked, elsewhere, of my view that one uncredited d-l
commentator claimed that the Creative Commons BY 1.0 licence's trademark
clause violates DFSG #3 _solely_ because he evidently failed to understand
trademark law and more evidently failed to understand the clause's
intent and probable judicial application.  Let's talk about one of the
other objections:

   When any Licensor asks, all references to their name(s) must be
   purged from the work. This restricts modification (DFSG §3).

Hypothetical:  I write a book on system administration, and publish an
instance of the work under CC BY 1.0.  Among the things that happen to
it is an utter dolt creates a derivative work that, in my view,
tarnishes my name by association, so I invoke the referenced clause to 
require my attribution be removed from the derivative.

_How_ has the DFSG #3 right of modification been materially infringed?  
The dolt is permitted to include or not include anything and everything
in his derivative, as long as he does not retain attribution to me.

That analysis -- and all the rest of the analysis of the cited licence
on that page -- is simply a bad joke:  It's an uncritical collection of
random, context-deprived wacky objections devoid of understanding of the
licensing situation and intent, apparently ignorant of trademark and
copyright law's actual real-world application, and is in general
obviously defective.

There's really no reason to suppose that those pages represent a
"debian-legal consensus".  Were they voted?  No.  Did anyone even take
any kind of poll?  No.  They were just individuals' posts to a public
mailing list that were picked up in a virtual net and summarised.

Thus my comment that the pages' implication of authority that they, in
fact, completely lack is deceptive in effect (though probably not in
intent).

And if a maintainer _were_ to decide to package a document on that
licence, and someone else were to block or move the package on grounds
that "When any Licensor asks, all references to their name(s) must be
purged from the work. This restricts modification (DFSG §3)", then I'll
bet that, on appeal, responsible parties would say "That's absurd.  We
doubt that that represents the consensus of debian-legal; and, if it
does, then debian-legal is simply mistaken."

Which then gets us back to the matter of Debian's governing structure. 
Debian-legal is not a rung on that ladder.  Its role is strictly
advisory.

At the _bare_ minimum, whoever edits that set of Web pages should make
its purely advisory role crystal-clear.  Which it most certainly is not.