Re: [Gnu-arch-users] top posting and flame

[Thomas Lord]

Daniel> Did Canonical somehow interrupt your work on GNU Arch?  Do you
Daniel> feel they are responsible for your not having been an active
Daniel> Arch developer for well over a year now?  If so, can you
Daniel> please list how?  Be specific.

For background, Stephen's account gets some details wrong but
is close enough for now.  (It is especially wrong about the nature
of negotiations between Canonical and me.)

I have two views of the events, one about ethics, the other about
economics.  I think that Canonical behaved with poor ethics.  I also
think that they shot themselves in the foot, economically speaking.
I don't think that Canonical is alone among free software businesses
who make similar mistakes -- they just happened to effect me,
personally, in this case.

Ethically, the situation began with a public project.  The project was
endorsed as part of the GNU project by the FSF.  The project was
supported by busking revenues and, during part of the time, earmarked
contributions to the FSF.  By "public project" I mean one that is
administered in the interest in the public and paid for by public
support.  For example, feature design was aimed to benefit "everyone
in general" rather than, say, a specific company.  While I was not
myself an NPO when getting busking revenues, none of those revenues
were donated as part of a contractual exchange -- they were analogous
to money put in the instrument case of a street musician.

(Later, the funding situation changed a bit -- as did how I spent my
time.  Most of my funding at this later stage came from an "angel
investor" friend.  Some of that funding worked as busking revenues --
leaving me (reduced) time to spend on the public project.  Some,
however, was spent on a private-interest project of trying to rescue a
start-up opportunity for the mess.  I mention this for completeness.
One early sale resulted from this -- code I wrote for Arch for a
private interest rather than as the most efficient available means to
advance the public project.  Canonical's harm predates this shift.  We
were of course very careful to make sure that the code written for
private a interest did, above all, no harm to the public project and,
additionally, provided benefit.)

As Stephen notes, Canonical arrived on the scene and perceived that
they had a private interest in using the code base of Arch in ways
that I expressed grave concerns about.  Of course, they are free to
use the code in any GPL-permitted way.  Nothing per se unethical about
forking and doing whatever the heck you like with the code (consistent
with the GPL and the values of software freedoms).

However, Canonical pursued its plans using a tactic of participation
in the public project.  They solicited volunteers from the project to
work on their fork.  They solicited users of the public project to
switch allegiance.  The rhetoric employed by representatives of
Canonical in public forums to do these things was not, as we might
say, specifically good.  I think that people were misled about
Canonical's dedication to the public value of the project -- an
opinion confirmed by the outcome of their fork.  They were misled
about Canonical's technical competence.  Somewhat obscenely, as far as
I can tell, Canonical was "outspending" my funding to work on the
project month to month by amounts well beyond 10x, prbly under 100x.
They were able to create the illusion of a flurry of useful activity and
promote a false theory of why that activity was useful.  They were able
to "out market" me.

Between two competing companies, those kinds of tactics may be just
part of the game.   As a way to compete against the public project
that created the very opportunity your company is exploiting, I find
it to be exploitative.   It exploited my labor.  It exploited
community participation in a public project.   At the end of the day,
Canonical left behind a public project in ruins.

Now some might counter that it is a corporation's *duty* to spend its
own resources as efficiently as possible to generate a return for
shareholders (even a not-yet-fully-formed or net-yet-public
corporation).   That, therefore, Canonical's spoiling of the public
project was entirely justified -- it's a dog-eat-dog world, no?

I agree with the the part about "*duty*".  I disagree that, in this
case, the spoiling of the project was justified -- simply because
there were plenty of same-cost alternatives that would almost
certainly have led to a better outcome *for both sides*.  They split,
they did not win the developer and user communities.  They left me in
a state of discredit based on highly misleading rumours about why I
had made certain decisions.  They left me without even the fraction of
their resources I had been working on.  They left their own fork in
such a state that even they had to abandon it.  And why?  Because the
Canonical hackers couldn't be bothered to format code correctly?
Choose algorithms a little more carefully?  And make interface
enhancements in any other way than the first little idea that popped
into their collective heads?  Because it was just too much bother to
help set up a more efficient patch flow to upstream (the project that
was underway when Canonical so rudely interrupted)?

In my family, when you borrow something, you do your best to return it
in as good or better shape as when you got it.   When you enter a room
full of strangers, you don't bust in like a bull in a china shop.
When you see a man trapped under a burning car, you stop to try and
help effect a rescue.  When you come across an expert in some topic that
interests you, you act with respect towards that expert.

Canonical seems to know nothing of those ethical values and the
outcome of their participation in the public project is a direct
reflection of that.   Given that there is no reasonable theory
showing a necessity for such bad behavior, yes, I lay a lot of the
destruction that resulted at their doorstep.

So that's ethics.  What about economics?

The punch-line of Canonical's fork seems to be that, from the
perspective of just a few years, they wasted a lot of their own
money.   They reduced their options (by holding GNU Arch back).
They did, indeed, drive their own fork up a tree to be abandoned
there.

You know, if they respected my technical judgement well enough to want
to use my work in the first place, they might have had the foresight
to at least hedge their bets by assuming my judgement might continue
to be useful going forward.  They might have tried to help, rather
than brush aside.

-t