[conspire] external storage recommendation

[Rick Moen]

Quoting Nick Moffitt (nick at zork.net):

> On 25Sep2021 07:03pm (-0700), Rick Moen wrote:
> > Which is cheeky, transparent, insultingly obvious corporate bullshit.
> > What they really mean is that they're taking an educated guess that the
> > Linux kernel copyright stakeholders will not give a damn sufficient to
> > bring copyright infringment against them and extract damages.
> 
> And yet this is kind of the way of civil litigation, no?

Of course.  If the victim of a tort doesn't bring effective litigation, 
it is always and everywhere the case that the tortfeasor will get off
scot-fres.  Way of the world.

> And in the case of Ubuntu, it so far seems to have borne out.  We
> shall see, though, I suppose!

As long as Canonical, Ltd. continue to be the parasite of the Linux
world (and we could cite endless examples, such as all the
special-pleading about mandatory copyright ownership-assignment
agreements before patches would be accepted to Canonical, Ltd. codebases 
-- with the obvious intent that Canonical and nobody else remain free to
sell proprietary forks.  When caught doing that, Canonical would
reliably trot out Shuttleworth to claim these so-called "contributor
licence agreements" ought to be seen as unobjectionable because FSF
uses them -- Shuttleworth making the bet that readers are too stupid
to spot the difference between his for-profit corporation and a
501(c)(3) foundation required by law to advance its charitable purpose
of supporting free software.

And when that con (inevitably) fails to fool readers, Shuttleworth rolls
out the Plan B excuse that Canonical are a Special Company That Need
Protection of Their Revenue Stream.

I saw the above cycle play out a couple of times in a row on LWN.net
comment threads.  

Anyhow, parasites, and they deserve a heaping of disgust, occasionally.



> I did a licence review at the actual FSF in Temple Place, and found
> some code in something we were hoping to release with the FSF's name
> on it.  And as is the case with the software we all tended to rely on,
> we'd find stuff from the 70s and 80s that was kind of unclear and had
> copyright statements on it that to which we couldn't find
> corresponding grants of licence.
> 
> And the answer was "what is the risk here?"

Oh, indeed.  Yes, somewhere, a copyright stakeholder, either the
original author or a successor in interest, is probably breathing and
could bring copyright infringement litigation, but the downside is
limited by laches (legal doctrine of limited recourse if you delay and
people come to rely on the time granted), among other things.  

OTOH, _since_ the 1980s, we've all been aware of the problems from 
reusing code of unclear or disputed provenance or having murky
licensing.  By "we", I mean we in the Linux ecosystem.  Other OSes,
certainly not so much.

That leads to a story.  You may recall that I was a significant devotee
of PalmOS PDAs.  While we were both working at VA Whatever /
Sourcewhatever, Inc., I went over to the managers in charge of
Sourceforge.net and told them I'd uncovered a troubling problem:  I'd
studied all of the allegedly open source codebases there that were fore
PalmOS, essentially licence-auditing them.  A shockingly high percentage
of them, and I had a printout, were actually either under variously
weird sorts of proprietary (but undisclosed) licensing, or had clashing
licence terms created by incompetently borrowed code from poorly chosen
sources.  Those had been created in Sourceforge with project metadata
claiming licence category of either "Other Licence" or "Public Domain".  

The two guys in question looked at me with undisguised apathy and
impatience.  They advised me that VA Whatever / Sourcewhatever company
management assumes no responsibility for the accuracy of project
maintainer claims of fact, and they didn't actually want to know about
any that were currently, or in the future, actively misleading the
public.

Well!  That certainly told me.  There being really nothing else to say,
I wished these smug asshat middle-managers a pleasant day and left.  
I did _not_ (at that time) tell the public "For Ghu's sake, don't _ever_
believe licence claims on Sourceforge, because company management knows
about wrong licence claims and outright-unlawful hosted software, but
deliberately does and says exactly nothing."  Employees need to 
actively avoid conflicts with company interests, so I carefully did so.

I did, however, build my own separate collection on my Web/ftp software 
of all publicly distributed software for and about PalmOS, and warned
about the packages that were incompently licence-labelled or unlawfully 
constructed -- without anywhere pointing a specific finger at VA
Whatever / Sourcewhatever by name.  ("Er, Sourcewhat?  Sorry, never
heard of them.")


> Another fun thing from that licence audit was that I found code
> claiming to be in the Public Domain as part of the old pre-gzip
> Compress library.

That was one of the lessons of my PalmOS-codebases licence audit:  
Anything that is described as "public domain" needs careful scrutiny,
if you want to be careful, because, sure, it could be actual public
domain, but more often means someone just wanted to borrow something
publicly available and _wanted_ it to be public domain.

> It had a date on it from before the Berne Convention was penned (let
> alone the USA's membership in same), and to this day remains the only
> bit of code I've ever used that I am reasonably confident is literally
> without copyright.

Pre-Berne _could_ have been public domain, but it's a heisenstatus.
That is to say, it was possible in pre-Berne days for a codebase to 
be adjudicated to be public domain by court action on grounds of the 
stakeholders permitting too much redistribution without copyright
notices.  You may recall that the judge in the AT&T v. UC Regents (BSD)
case shot that across plaintiffs' bow in the preliminary ruling that 
caused the hasty settlement:  The judge observed that plaintiff's own 
UNIX source code had been bandied about rather a lot without copyright
notices, hence it was questionable whether it had copyright title to
litigate over.  And plaintiff then caught the judge's broad hint and
made tracks settling with UC.  1993, was it?  Around there.

But absent a judge's ruling that the stakeholder of a pre-Berne work had
forfeited copyright title, all that could be fairly said about such a
codebase's copyright title was that it was vulnerable to legal
challenge.

One of Berne's immediate effects was, for better or worse, to end the 
practice of copyrights getting voided through stakeowner halfassedness:
Legal title persists whether instances bear copyright notices or not.