RE: Re: MIT vs GPL vs LGPL for open source project

Austin, if it makes you feel better to call anyone who disagrees with
you a zealot, enjoy. But from where I sit, it’s a little more
complicated than that.

– You, and the people supporting licenses like MIT, use a definition of
‘freedom’ that maximizes choice for the largest number of people,
optimizing across the first round of choices after you release the
license. That trades-off more choices for more people about what to do
with the licensed code today for a risk that subsequent changes may
permit fewer choices for some people in future. Nothing wrong with that,
and there are plenty of people out there who use a similar definition.

– Stallman and the people advocating GPL use a definition of ‘freedom’
that maximizes choice for the largest number of people, optimizing
across all future rounds. That trades-off a reduced immediate
‘freedom’ for some (the GPL restrictions you mention) for a guarantee of
the same amount of ‘freedom’ for all people in all future rounds.
Despite your personal FUD, there’s nothing inherently wrong with that,
either. And clearly, there are a lot of people who use a similar
definition.

Despite what you may think, this doesn’t mean that MIT is better than
GPL – or that GPL is better than MIT – or that BSD is better than
Linux – or even that Betamax is better than VHS ;-). It is literally
impossible to predict which strategy will result in greater total
‘freedom’ without making some pretty strong assumptions about (a) the
relative size of the first v. subsequent-round audiences, (b) the
proportion of each audience that would actually want to do something
contrary to the license in question if given the opportunity, and ©
the current context and future direction of the particular OS
application and its community.

In other words, both options maximize choice for an arbitrarily large
group of people: they just each define the group differently. And as
both economists and psychologists can tell you, when used in a
predictive capacity, those two strategies are more a matter of
individual taste than right v. wrong. Some people ride motorcycles
off-road without helmets; others won’t get out of bed unless the
insurance policy is paid-up. Taste for risk varies, as does
future-orientation.

So feel free to “go with your gut” in your own decision process – but
why not cut a little more slack for those whose guts tell them
differently? --Chris

Austin Z. wrote:

Instead, imagine the alternative history if BSD hadn’t been in the
middle of a legal fight.

On Sunday, April 09, 2006, at 12:55 PM, Christopher J. Mackie wrote:

and there are plenty of people out there who use a similar definition.
Despite what you may think, this doesn’t mean that MIT is better than
In other words, both options maximize choice for an arbitrarily large
differently? --Chris

I was about to post along the same lines, but you totally nailed it.
Nicely said.

On 4/9/06, Christopher J. Mackie [email protected] wrote:

Austin, if it makes you feel better to call anyone who disagrees with
you a zealot, enjoy. But from where I sit, it’s a little more
complicated than that.

[… nonsensical and incorrect text elided …]

Please do me a favour and actually read what I write. The claim of
zealotry isn’t related to licence choice. In fact, in this very thread,
I have indicated that the aims of the GNU GPL (e.g., Share Alike) are
very good. There’s no FUD there. The FUD comes from GNU GPL zealots who
claim that the GNU GPL is “more free” than the MIT or BSD licences.
Measurably, it is not so. The choice of the restrictions provided by the
GNU GPL or a similar licence may be desirable or even preferable. But
where I have a problem – and where the zealot claim sticks – is that
is isn’t “my” side trying to change definitions and use terms in ways
that render them ultimatelyi meaningless.

Despite what you may think, this doesn’t mean that MIT is better than
GPL – or that GPL is better than MIT – or that BSD is better than
Linux – or even that Betamax is better than VHS ;-). It is literally
impossible to predict which strategy will result in greater total
‘freedom’ without making some pretty strong assumptions about (a) the
relative size of the first v. subsequent-round audiences, (b) the
proportion of each audience that would actually want to do something
contrary to the license in question if given the opportunity, and (c)
the current context and future direction of the particular OS
application and its community.

I have not made any statements about the relative better choice of a GNU
GPL-like licence over an MIT licence. Others in this thread have; I have
not. I have indicated that there is “something rotten in the state of
Denmark” with the way that the GNU GPL is evangelized; I have expressed
my preference for the MIT licence; I have not held back on pointing out
that there are people, primarily those who senselessly advocate the GNU
GPL to the exclusion of all ese, who play Humpty Dumpty[1].

[…]

So feel free to “go with your gut” in your own decision process – but
why not cut a little more slack for those whose guts tell them
differently? --Chris

I haven’t said that people shouldn’t ever choose the GNU GPL. Ever. I
challenge you to find where I have said so. I have said that people
shouldn’t choose the GNU GPL under false pretenses or for the wrong
reasons. I would argue that they shouldn’t choose any licence under
false pretenses or for the wrong reasons.

I just happen to think that the evangelism and groupthink around the GNU
GPL leads to both false pretenses and the wrong reasons. There is
software that probably should be under the GNU GPL. But that is
software where the authors have considered their licence choice
carefully.

-austin
[1] “When I use a word,” Humpty Dumpty said in a rather a scornful tone,
“it means just what I choose it to mean–neither more nor less.”
http://sunsite.tus.ac.jp/coll/alice/alice_25.html

Mon, 10 Apr 2006, Austin Z. skrev:

There is
software that probably should be under the GNU GPL. But that is
software where the authors have considered their licence choice
carefully.

I’m not totally in opposition to this. Generally I would say:

If you, or the group who are writing a software might think of making
it closed source in the future for commercial reasons: Then of course
you can’t use GNU GPL. If this is the case, then I think you should
inform eventually other contributors about this possibility (so
amateur’s like me can take that into consideration if we should think
of spending time on the software).

If making the software closed source not will be an option, then I
think you generally should use the GNU GPL license (of reasons I have
mentioned in other emails). There are no morally wrong with either of
this choices.

  • Henrik