🚀 go-pugleaf

Maciej Stachowiak <mstachow@mit.edu> writes:

> I think the proposed system is terrible, DATs should not be taxed at
> all. Of course, I am a libertarian and believe that what the
> government does is supposed to protect the rights to life, liberty and
> property exclusively, even if the citizens might want otherwise.

I disagree. If the government does not tax DATs (or software, books,
etc.), then publishing companies will. It's kind of like saying
"having the government protect people's life an property is just too
intrusive, so everyone has to hire a bodyguard." The only people who
win are the bodyguards or the publishing companies because they get
money.

I do agree with *most* of your libertarian stance, I just think that
freedom of information should be protected too. People are going to
want rewards for their work, so the government should facilitate their
receiving them *without* restricting others' freedom to use the work
in question. Of course this ideal government wouldn't provide anyone
with copyright anyway... :-)

I think the core libertarianism is that the most amount of freedom is
good, so the practice of it is basically figuring out what freedoms
you need to remove to maximize the total freedom availible. The GPL
takes away the freedom to restrict things, but it gives back much more
in the freedom to use, study, distribute, modify, etc. So in this way
it's more libertarian than X-style liscensing.

> * Look at any of the attempts to count Linux users by direct survey -
> do you think these are remotely accurate? Users are too lazy to
> respond to a survey unless it is mandatory, in which case it is evil.

I think that might change if voting meant a brighter future for free
software and real money for its developers. Most GNU users care about
freedom, but don't really care what the "usage" pie charts in IT rags
say. Of course, saying "free software" instead of "open source" can
only help more people realize that the issue is freedom...

> * How widely a program is distributed is not the sole measure of its
> usefulness. For example, whoever wrote the Linux `mount' would be on
> the gravy train with this system, since everyone has it; but it is a
> comparatively trivial program and requires little maintenance.

I have been worrying about this... Now keep in mind that the
voting/allocation process itself is independent of everything else and
can be revised whenever. So here's my first idea. People can first
report a list of everything they have installed; package managers make
this easy. That counts for a little, but not much. Then they can give
more points on top of that to whatever they feel is most important; a
scale of 1 to 10 or something would be nice. So `mount' would end up
with a bunch of tiny points, since no one voted for it exclusively,
and niche apps would get a small number of "very useful" big points.

> * A lot more users would probably use end-user applications than
> tools, but without tools development the end-user apps would
> languish.

This is the second issue. If, for example, GIMP uses GTK+, then a
portion of GIMP's points should be transferred to GTK+. But not that
much, as many other apps will use GTK+ too. The problem is deciding
just how much. But if we figure that out, people with not that much
time on their hands could just vote for "Gnome" and the points would
trickle down to the apps themselves, and from there to the libraries.
Concievably, really lazy people could just vote "Red Hat" or "Debian"
and have all their tax distributed however the maintainers of the
distribution felt best represented everybody. Or, hackers could set up
a vote that distributed points to underrated but really necessary
basic tools.

> Also, it may be hard sorting out the relative contributions of the
> 100s of contributors to the typical GPL work. I don't think lines of
> code is a good measure - some lines count for a lot more than others.

That's definetly true. I think that the appreciation of real people
for the work, just like with the DAT outline, is the way to go. 

-- 
please read: http://www.gnu.org/philosophy/
"cry me a bayou/and i'm leaving..." -tricky