[OT] DPP thinks the FSF is evil, most rational people disagree
David Pollak
I made some seemingly outrageous statements about the FSF on the list. Corey refuted and rebutted the statements. I'd like to spend some cycles walking through my logic. In order to do that, I'm going to ask Corey (and others on the list) a series of questions and as we work through the questions, I think the reasons for my position will become clear. I don't expect to change any minds about the FSF, but I do hope to clarify that I'm not a wacko.
So... here are a series of questions:
- Should producers of products that have a marginal cost of production that approaches zero be required to sell the products at marginal cost of production?
- Are there trade secrets that provide companies competitive advantage ( e.g., formulas that Fidelity uses to balance portfolios, manufacturing processes that save money, etc.)?
- Should companies be allowed to protect these trade secrets in order to derive the business advantages of these secrets and in order to recover the costs of developing these processes?
- Given that computers manage most aspects of business these days, are many of the trade secrets in #2 and #3 embodied in the software?
Thanks,
David
--
lift, the fast, powerful, easy web framework
http://liftweb.net
Corey
> Folks,
>
> I made some seemingly outrageous statements about the FSF on the list.
> Corey refuted and rebutted the statements. I'd like to spend some cycles
> walking through my logic. In order to do that, I'm going to ask Corey (and
> others on the list) a series of questions and as we work through the
> questions, I think the reasons for my position will become clear. I don't
> expect to change any minds about the FSF, but I do hope to clarify that I'm
> not a wacko.
>
> So... here are a series of questions:
>
Hello!
I'll go ahead and provide my opinions to these questions as directly as
possible.
> 1. Should producers of products that have a marginal cost of
> production that approaches zero be required to sell the products at
> marginal cost of production?
>
Producers certainly should not be 'required' - by _law_ (de jure) -to to sell
their products at marginal cost. However I can fathom a scenario wherein, due
to market factors, certain industries and/or some or many producers find that
they are operating/competing/entering within an environment where indeed they
must sell their products at marginal, or near-marginal, cost in order to
remain competitive within whatever niche they occupy (de facto).
> 2. Are there trade secrets that provide companies competitive
> advantage (e.g., formulas that Fidelity uses to balance portfolios,
> manufacturing processes that save money, etc.)?
>
I think that this warrants a fairly obvious 'yes' answer.
> 3. Should companies be allowed to protect these trade secrets in order
> to derive the business advantages of these secrets and in order to
> recover the costs of developing these processes?
>
I don't believe that question needs qualifying, simply:
'Should companies be allowed to protect these trade secrets [period]?'
Yes - assuming the mechanisms they use to do so are devoid of any unjust
or artificial means (e.g., special privilege granted via non-market
entities/forces such as the state).
> 4. Given that computers manage most aspects of business these days,
> are many of the trade secrets in #2 and #3 embodied in the software?
>
Yes, though I feel this must be qualified with the very important prerequisite
that said "trade secrets" are in fact founded on a coherently and rationally
understood basis and are not abused in such a manner so as to include obvious
and/or ubiquitous processes.
For an example of what I mean, just one current case in point:
http://www.duiblog.com/2007/09/04/secret-breathalyzer-software-finally-revealed/
"
Initially, Base One found that, contrary to Draeger’s protestations that the
code was proprietary, the code consisted mostly of general algorithms: “That
is, the code is not really unique or proprietary.“ In other words, the “trade
secrets” claim which manufacturers were hiding behind was completely without
merit.
"
> Please answer the above questions and we'll either discuss the answers or
> go on to the next set of questions.
>
Proceed, sir. (c8=
Beers,
Corey
David Pollak
I'll post more tomorrow, but wanted to interject one little thing...
> 4. Given that computers manage most aspects of business these days,
> are many of the trade secrets in #2 and #3 embodied in the software?
>
Yes, though I feel this must be qualified with the very important prerequisite
that said "trade secrets" are in fact founded on a coherently and rationally
understood basis and are not abused in such a manner so as to include obvious
and/or ubiquitous processes.
Agreed. I'm a lawyer by training (back in the day, I did IP transaction work and wrote a bunch of license agreements) and my wife is an IP litigator ( http://www.hellerehrman.com/en/attorneys/bios/Hurst_Annette.html)
Analysis in the law is broken down into "the law" and "the facts". Often the facts are difficult for lay people to parse, but experts are very helpful. Analyzing source code to determine if it actually contains trade secrets is a matter of "fact". Disputes that get as far as a jury are disputes of fact. Disputes of law (where all the parties agree on the facts) is taken care of reasonably efficiently by the judge.
So, for the sake of our discussion, there are always facts that can lead a rule not to be applied. But there also exist fact patterns that allow the rule to be applied.
An interesting example of this is http://cohesiveft.com They are putting together open source software stacks for the financial services industry. Their CEO was CIO of UBS/Swiss Bank. It used to be that trading shops developed custom applications from the OS level up (about 80% of the software stack was built internally.) Cohensive is taking the approach that 5% of trading floor software should be unique and the other 95% should be one of a few stacks. In this example, the 5% would really and truly be proprietary. Most fact patterns are not this clear.
Thanks,
David
Antoine Toulme
I am not (yet) a Lift user, I subscribed to the mailing list to follow a bit this community and eventually dive into Scala one of these days.
I am stepping up to say that I agree with David.
I think people should be able to develop, use and redistribute software without needing a lawyer.
I think the GPL is actually somewhat limiting the freedom of the users.
Regards,
Antoine
http://antoine.toulme.name
http://www.lunar-ocean.com
http://www.requirements-tracker.com
Corey
<snip>
> I think people should be able to develop, use and redistribute software
> without needing a lawyer.
>
So do I.
In fact, I can _imagine_ that an overwhelmingly large percentage of people
would agree; I mean, think of the obvious absurdity that would result
in reciprocating that statement:
'I think people should need a lawyer in order to develop, use and/or
redistribute software.'
> I think the GPL is actually somewhat limiting the freedom of the users.
>
Which users are you referring to?
The only subset of users who's "freedom" is explicitly limited via the GPL
are those users who redistribute software derived from GPL code -- and the
only subset of _those_ users that might reasonably 'need' a lawyer - are those
who desire or require that their new and/or pre-existing propriety code, which
is to be derived/dependent on open GPL code, stays closed.
So, there really is only a very particular subset of 'users' who should
in fact consider _really_ being certain that they understand the TERMS
(contracts _do_ have terms, you know) of the L/GPL; even if that means
consulting a lawyer.
It is in such instances, where the GPL is simply not an acceptable fit for a
propriety vendor, that such "liberal" licenses such as MIT show their
particular suitability, and I am glad such licenses exist.
To summarize, I think you failed to be persuasive here.
Stating that you don't think people should need a lawyer to develop, use
and redistribute software is hardly supporting evidence for why you agree
that the 'FSF is evil', and/or why you think the GPL limits the freedom
of users, because - #1 the GPL/FSF do not require people to use a lawyer to
do these things, nor do they even remotely suggest such an absurdity; and -
#2 you conflate 'all people' and 'users' with what is in fact just a very
certain specific subset of people: namely, propriety re-distributors.
Cheers,
Corey
Erik Engbrecht
The only subset of users who's "freedom" is explicitly limited via the GPL
are those users who redistribute software derived from GPL code -- and the
only subset of _those_ users that might reasonably 'need' a lawyer - are those
who desire or require that their new and/or pre-existing propriety code, which
is to be derived/dependent on open GPL code, stays closed.
(emphasis mine)
Corey
> > The only subset of users who's "freedom" is explicitly limited via the
> > GPL are those users who redistribute software derived from GPL code --
> > and the only subset of _those_ users that might reasonably 'need' a
> > lawyer - are those
> > code, which
> > is to be derived/dependent on open GPL code, stays closed.
>
> (emphasis mine)
>
> No, not proprietary. The code could not be released under an open source
> license other than the GPL, so it must stay closed. So the freedom of
> anyone who wishes to release their code that uses GPL code under a license
> other than the GPL is restricted.
>
> Ok, actually this isn't entirely true. You could have a codebase that
> is part BSD (for example) and part GPL, but the combined work would be GPL.
> So someone could use your work under the BSD license iff they first removed
> the GPL part.
>
Thank you - your correction is certainly more accurate; my use of the word
'proprietary' and 'closed' did not facilitate the nuance which you made more
clear.
That _does_ make the 'subset' of users explicitly restricted via the terms of
the L/GPL slightly larger. Rather than saying proprietary/closed, I should
have said non-L/GPL'd - and would have then also needed to remark on the
exception regarding those (somewhat rare) instances of originally
dual-licensed works as you mentioned.
Cheers,
Corey
Erik Engbrecht
That _does_ make the 'subset' of users explicitly restricted via the terms of
the L/GPL slightly larger. Rather than saying proprietary/closed, I should
have said non-L/GPL'd - and would have then also needed to remark on the
exception regarding those (somewhat rare) instances of originally
dual-licensed works as you mentioned.
David Pollak
Today's question is quite simple:
On a scale of 1 to 10 (1 being unimportant and 10 being critical) is it that the terms of a license conform to all of the following criteria:
- The terms are generally understood to mean the same thing by users of the licensed software
- The terms are generally understood to mean the same thing by the producers and users of the licensed software
- The producer of the software clarifies the meaning of the terms when there is ambiguity
- The producer of the software is willing to test the terms of the license in courts such that the courts will enforce the contract as it is understood by the producers and users
Corey
> Well... this is convenient.
>
> Today's question is quite simple:
> On a scale of 1 to 10 (1 being unimportant and 10 being critical) is it
> that the terms of a license conform to all of the following criteria:
>
> of the licensed software
> 2. The terms are generally understood to mean the same thing by the
> producers and users of the licensed software
> when there is ambiguity
> 4. The producer of the software is willing to test the terms of the
> license in courts such that the courts will enforce the contract as it
> is understood by the producers and users
>
I think the correct answer, ultimately - for any license - would be: 10.
You could replace 'users' with 'citizens', and 'producers' (ironically)
with 'government', and 'terms' with 'laws' and 'software' with 'society',
and rudimentary nature of the underlying question would stay intact:
Fact is, for any protocol to work at its most optimal capacity it must in
theory suffer the least ambiguity possible. Although in practice this is
difficult to achieve because it's a messy world out there, and interests
often directly conflict between different groups of variously opposing
actors - even while they're all playing at the same table so to speak.
Getting back to your question specifically though, I'm certain you'd get
different levels of importance from different users/producers/distributors.
'Mere' downstream end-users/consumers might not give a wink about any of
those 4 points, so long as they get them their free stuff, and it works
good.
More interested/involved users, especially those with a collective mindset,
might place more importance of your 4 points, seeing a means to end via the
license, and therefore the efficiency of the license towards it's affiliated
end.
Developers are of course a motley bunch, and you see the gamut of opinions
from "hell if I care, I do this for fun", to "I want to be damned certain
people are respecting the terms of the code I write", to "whatever works".
It's generally within the court of commercial enterprises with profit motive,
and existing open source communities ( *bsd camp, Apache, GNU, etc. ), where
we really start seeing the terms of these licenses put to the test in terms of
actual legalese and courts and posturing and whatnot.
Cheers,
Corey
Corey
> > That _does_ make the 'subset' of users explicitly restricted via the
> > terms of
> > the L/GPL slightly larger. Rather than saying proprietary/closed, I
> > should have said non-L/GPL'd - and would have then also needed to remark
> > on the exception regarding those (somewhat rare) instances of originally
> > dual-licensed works as you mentioned.
>
> It hits LGPL software, too. For example, most (all?) of KDE is licensed
> under the LGPL,
>
Well, _much_ of KDE is actually licensed under the GPL, however most of the
kde _libraries_ ('kdelibs') package is licensed under LGPL.
> However, if you do not have a commercial Qt license, then the KDE libraries
> are for all practical purposes GPL,
>
Untrue.
You can use any gpl (v2) compatible license (which includes lgpl) - and
there's lots of them:
http://www.gnu.org/philosophy/license-list.html#GPLCompatibleLicenses
Emphasis mine:
"
If you develop using Qt Free Edition, your program must also be Free (it must
be released under a license that is *compatible* with either the GPL or the
QPLv2).
"
from http://kdemyths.urbanlizard.com/myth/59
Also:
"
With the release of Qt2, the GPL has been added as a licensing option. Free
Software can link against and use Qt using a broad range of licenses,
including the GPL, LGPL, X11, MIT and BSD licenses.
"
http://kdemyths.urbanlizard.com/myth/10
> To insert my opinion, I don't think that the GPL is bad or the FSF is evil.
> I do think the FSF is slightly deceptive in the way they describe how the
> GPL interacts with other open source licenses. I will not use GPL
> libraries or frameworks in software that I intended for commercial use
> (even internal use) or in an open source library or framework. Much of
> that concern is that I think the FSF leaves a lot up to their own
> interpretation, so I would be afraid of being sued even if felt I was in
> compliance. I think it's more a matter of being ammoral with an agenda
> than being evil. I have no problem using LGPL software libraries.
>
Fair enough - your decision to use or not use GPL'd software is certainly
your own.
I must say that don't see how it would be beneficial for FSF to be deceptive
about how GPL interacts with other open source licenses, nor have I personally
seen or detected/suspected any amoral behavior from them - in fact my
experience has been that they are extremely consistent and informative of
their goals and motives -- although I would definitely be interested in seeing
evidence of what you describe. Has the FSF actually unjustly sued anyone?
I do however, have a tough time with their tendency to occasionally define
morality; personally, I wish they'd completely depart from that particular
aspect, I don't react well to people telling me what's moral, and from a
purely tactical standpoint, I think that particular expression of theirs tends
to work against them; aside from that, I think gnu, the gpl and the fsf are
overall very positive in their affects and accomplishments.
Beers!
Corey
Erik Engbrecht
Well, _much_ of KDE is actually licensed under the GPL, however most of the
kde _libraries_ ('kdelibs') package is licensed under LGPL.
You can use any gpl (v2) compatible license (which includes lgpl) - and
there's lots of them:
http://www.gnu.org/philosophy/license-list.html#GPLCompatibleLicenses
What does it mean to say that two licenses are "compatible"?
In order to combine two programs (or substantial parts of them) into a larger work, you need to have permission to use both programs in this way. If the two programs' licenses permit this, they are compatible. If there is no way to satisfy both licenses at once, they are incompatible.
For some licenses, the way in which the combination is made may affect whether they are compatible--for instance, they may allow linking two modules together, but not allow merging their code into one module.
Just to install two separate programs in the same system, it is not necessary that their licenses be compatible, because this does not combine them into a larger work.
What does it mean to say a license is "compatible with the GPL".
It means that the other license and the GNU GPL are compatible; you can combine code released under the other license with code released under the GNU GPL in one larger program.
The GPL permits such a combination provided it is released under the GNU GPL. The other license is compatible with the GPL if it permits this too.
As for my "the FSF is amoral comment," don't confuse "amoral" with "immoral." I can't think of a single large corporation or organization that I wouldn't call amoral. That doesn't mean they don't operate within acceptable ethical and legal bounds. With corporations that doesn't really matter. The purpose of a corporation is usually to make money and generate as high of a return for its shareholders as possible. But the FSF is pursuing what I believe amounts to a religious cause. This is what scares me, because it is very easy to accidentally violate the GPL if you are not keenly aware of to what your software is linking. This is especially true when you involve scripting languages in an environment where all the libraries come pre-installed.
Erik Engbrecht
#2 - 8
#3 - 10 - this is tough for OSS because the producer usually isn't the same as the entity defining the license
#4 - 7
Those are extremely insightful criteria. I never looked at it from that angle but, wow, that's an epiphany. Thank you for openning my eyes.
http://erikengbrecht.blogspot.com/
Corey
> #1 - 6 - users are numerous and common understanding is hard
> #2 - 8
> #3 - 10 - this is tough for OSS because the producer usually isn't the same
> as the entity defining the license
> #4 - 7
>
The way I read it, he was asking for a single score, to be applied to all
those criteria as a whole, not each individually:
"
On a scale of 1 to 10 (1 being unimportant and 10 being critical) is it
that the terms of a license conform to all of the following criteria:
"
... I guess we can take the average of your scores - 8.
<grin>
Although, I'm still wondering what the significance of his question is - it
wasn't the epiphany for me that it was for you.
I see a very subjective number pulled out the mind of who ever is answering,
which is very likely to change according to current mood and/or inclination
of the participant, in addition to a number which would of course change
dependant on the _perspective_ of the participant -- with which interest group
does one identify oneself with?
A: the "downstream end-user user" actor
B: the "open-source/free-software developer user" actor
C: the "commercial/proprietary developer user" actor
D: the "distributor of GPL'd code user" actor
D.1: commercial distributor
D.2: open-source/free-software distributor
E: the "distributor of non-GPL'd code user" actor
E.1: commercial distributor
E.2: open-source/free-software distributor
It's been my observation that when people take an adamant stance against the
FSF and the GPL, they tend to be binding/aligning themselves most tightly as
a member of group C and/or E.1 -- even though they do infact often occupy one
or more of the other groups, and benefit greatly from the explicit terms and
aims of the GPL and FSF as a _direct_result_ of the restriction/constraint
applied to groups C and E.1.
Here's a question for you guys, I'm going to riff off David's:
On a scale of 1 to 5 (1 being unimportant and 5 being critical) is it
that the terms of a license effectively implement to all of the following
criteria:
0. The freedom to run the program, for any purpose
1. The freedom to study how the program works, and adapt it to your needs
2. The freedom to redistribute copies so you can help your neighbor
3. The freedom to improve the program, and release your improvements to the
public, so that the whole community benefits
Again, your answer to this would be subject to the same weaknesses I described
concerning David's early question.
You'll notice that the following requisite is, unsurprisingly, lacking from
the above set:
4. the freedom to change the program's license in order nullify one or more of
the above freedoms
Continuing - such language as "so you can help your neighbor", and "so that
the whole community benefits", are entirely irrelevant to those who, for
whatever reasons, prefer to sympathize/identify primarily with groups C and/or
E.1.
This is where I detect a gross hypocrisy in those who accuse the GPL of being
largely flawed (vs. merely being unsuitable for their interests) - they in
fact are upset that it does not allow them to do that which runs counter to
its entire purpose: which is to make more software that is in perfect
accordance with those 4 freedoms more available to the world at large.
The GPL and the FSF have always been 100% consistent, and informative of their
goals, in that regard.
CHeers,
Corey
David Pollak
On Sep 8, 2007, at 12:30 PM, Corey wrote:
>
> Here's a question for you guys, I'm going to riff off David's:
>
> On a scale of 1 to 5 (1 being unimportant and 5 being critical) is it
> that the terms of a license effectively implement to all of the
> following
> criteria:
>
> 0. The freedom to run the program, for any purpose
I'm assuming you mean any legal purpose and I'll put this at a 3, but
it depends on the circumstances. For example, I like student
discount pricing on software (or books or other items.) It allows a
manufacturer to have some level (at a very gross level) of
distinguishing ability to pay and pricing accordingly.
>
> 1. The freedom to study how the program works, and adapt it to your
> needs
This is a 2 parter. The 9th Circuit has already affirmed a person's
right to reverse engineer code which includes studying it. This is
something a license cannot take away.
The second part is a 1. I do not believe that I have an inherent
right to source code. I believe that this is market dictated (just
like pricing.) I like to tinker with stuff, but the fact that I
don't have source to TiVo or the Nav system in my car does not
decrease the utility of either.
>
> 2. The freedom to redistribute copies so you can help your neighbor
1 (or 0). If my neighbor wants software, then my neighbor should
contract with the software vendor. This goes back to my very first
question about pricing. I do not believe that software is any
different than any other item. There is no inherent requirement that
software be priced at the marginal cost of production. There are
good reasons to price software in such a way, but there is nothing
about software that distinguishes it from prescription drugs, chips,
cars, etc. all of which are priced to allow the producer to recover
fixed costs as well as make a profit. The fact that it's easier to
copy software doesn't change the equation.
>
> 3. The freedom to improve the program, and release your
> improvements to the
> public, so that the whole community benefits
1. People have the right to reverse engineer software. That's
good. Making derivative works falls into the arguments above.
With all that being said, there are market reasons for having open
source software. There are market (and moral) reasons for having
licenses that do not allow folks to make derivative works and profit
off the contribution of the "open" authors. Having a continuum of
licensing options for the whole software ecosystem is good.
>
>
> Again, your answer to this would be subject to the same weaknesses
> I described
> concerning David's early question.
>
> You'll notice that the following requisite is, unsurprisingly,
> lacking from
> the above set:
>
> 4. the freedom to change the program's license in order nullify one
> or more of
> the above freedoms
5. This is very important. The authors of a work should have the
right to change and control the distribution of their work. If they
have pre-existing contractual obligations (e.g., everyone who is a
lift committer has to assign their committed IP to an entity which
owns the lift IP and in exchange, that entity has agreed to keep the
lift code available under an Apache 2.0 license) then they have to
honor those obligations (or re-negotiate or risk litigation). But,
just as with any other form of property, intellectual or otherwise,
the owner of the property has a broad set of rights with regard to
that property. I see nothing that distinguishes software such that
it must always be "free" especially when, as you agreed in the last
exchange, software freedoms can interfere with the protection of
other IP.
Thanks,
David
>
>
> Continuing - such language as "so you can help your neighbor", and
> "so that
> the whole community benefits", are entirely irrelevant to those
> who, for
> whatever reasons, prefer to sympathize/identify primarily with
> groups C and/or
> E.1.
>
> This is where I detect a gross hypocrisy in those who accuse the
> GPL of being
> largely flawed (vs. merely being unsuitable for their interests) -
> they in
> fact are upset that it does not allow them to do that which runs
> counter to
> its entire purpose: which is to make more software that is in perfect
> accordance with those 4 freedoms more available to the world at large.
>
> The GPL and the FSF have always been 100% consistent, and
> informative of their
> goals, in that regard.
>
>
> CHeers,
>
> Corey
>
>
> >
--
David Pollak
http://blog.lostlake.org
Corey
I appreciate your answers, however I think you clearly illustrated my gist -
one's answers will not only be highly subjective, but they will also be
extremely biased towards the preferences of a particular interest group.
Case in point: you gave the _highest_score_ to the one criteria that is
logically not available via the GPL, and you gave exceedingly low scores
to each of the criteria the GPL places in high regard.
Additionally, you provided your answers in terms of absolutes - you answered
based on a universal sense of '_all_ software'. That is a very collective
mindset. What about the less general case of '_some_ software', on down to the
specific instance of '_this_ software'?
How would your answers have changed if I were to have replaced "the program",
with "scala"? I'm guessing the scores you placed would have each been much
higher!
> > 1. The freedom to study how the program works, and adapt it to your
> > needs
>
> This is a 2 parter. The 9th Circuit has already affirmed a person's
> right to reverse engineer code which includes studying it. This is
> something a license cannot take away.
>
Not to be overly pessimistic - but I doubt you'd be very satisfied with
that answer were scala to be distributed in purely binary form, and I
daresay that it is unlikely the developer base you are hoping to facilitate
would be satisfied with such a platitude were you to do the same with /lift/.
I think your first part of your 2 parter was 9 parts smoke! (c8=
> The second part is a 1. I do not believe that I have an inherent
> right to source code.
>
One does not need to believe one has an _inherent_right_ to source code
in order to see the strengths and benefits inherent in having source made
available.
The GPL does not grant inherent rights, it grants privileges to those who
agree to its terms - namely, the privilege of using the GPL'd software in
question.
> I believe that this is market dictated (just like pricing.)
>
The market is, ultimately, dictated by (preferably rational) people and
processes. If the market changes due to the course of revolutionary or
evolutionary processes (as it chronically does), then, well - "the market
has spoken."
Free Software and Open Source software are fast becoming significant factors
in the market. If they become more and more prevalent, perhaps even to the
point of total ubiquity - then, by all means: "the market has spoken."
Although it's not surprising that at any given point in time, some people will
be satisfied with whatever direction the market is moving, whereas others will
be threatened whatever direction the market is moving.
When one is satisfied, one says: "the market dictates".
When one is threatened, one says: "<insert corrupting influence> is
going to destroy the market!"
> I like to tinker with stuff, but the fact that I don't have source to TiVo
> or the Nav system in my car does not decrease the utility of either.
>
I'm very surprised that you said this.
I think reasons why should be immediately obvious.
> > 2. The freedom to redistribute copies so you can help your neighbor
>
> 1 (or 0). If my neighbor wants software, then my neighbor should
> contract with the software vendor.
>
I wish that the FSF didn't refer to "neighbor" for that freedom. A
more generic entity is require to prevent people from (understandably)
pigeon holing the spirit of that freedom.
Replace "neighbor" with "client", and put yourself as a "vendor" who
is distributing a bundled enterprise application framework. I'm sure
you'd prefer to provide scala in that bundle, rather than telling your
clients - "This software you are about to purchase requires scala (along
with X number of other packages) - you'll need to go contract with those
other software vendors first before you can use this program."
> With all that being said, there are market reasons for having open
> source software. There are market (and moral) reasons for having
> licenses that do not allow folks to make derivative works and profit
> off the contribution of the "open" authors. Having a continuum of
> licensing options for the whole software ecosystem is good.
>
I agree.
> > You'll notice that the following requisite is, unsurprisingly,
> > lacking from the above set:
> >
> > 4. the freedom to change the program's license in order nullify one
> > or more of the above freedoms
>
> 5. This is very important. The authors of a work should have the
> right to change and control the distribution of their work.
>
Those who would prefer that their work be distributed under the terms of
a copyleft license would disagree strongly with you that point; which is
why they themselves use, develop and distribute GPL'd code. It's not a matter
of 'right' or 'wrong', it's just a matter of which ecosystem/market/community
one prefers to take part in.
I can understand why the dude(s) who implemented the scheduler for linux
might very well place terms on that code, via the GPL, so that any particular
copy of their work is ever enclosed. They very well may have never done the
work were it not protected thus; it's in their right to place terms/price on
how their code is used. This does not make the FSF or the GPL evil.
Chris Palmer
insert a question, two factoids, and a regular expression below. As a
former EFF Staff Technologist, it's sort of my (former) job to
profoundly disagree with David :D but nobody wants to hear it, so I'll
spare you. :)
David sez:
> > This is a 2 parter. The 9th Circuit has already affirmed a person's
> > right to reverse engineer code which includes studying it. This is
> > something a license cannot take away.
Which case? (Note that the DMCA does exempt reverse engineers from
bizarre punishment in the specific case of RE for compatibility, but
otherwise RE is pretty frowned upon. I assume you are talking about
some case that is not DMCA-flavored.)
Corey sez:
> Not to be overly pessimistic - but I doubt you'd be very satisfied with
> that answer were scala to be distributed in purely binary form, and I
> daresay that it is unlikely the developer base you are hoping to facilitate
> would be satisfied with such a platitude were you to do the same with /lift/.
You all realize that JVM code is essentially source code, right?
E.g.: http://www.kpdus.com/jad.html
> The GPL does not grant inherent rights, it grants privileges to those who
> agree to its terms - namely, the privilege of using the GPL'd software in
> question.
I'm sorry, I just have to pick a nit. It's my nature. :) Users --
people "merely" using GPL software -- do not have to agree to the GPL
at all. (Despite all those goofball programs whose installers make
you click through the "GPL EULA".) Only redistributors must agree to
its terms, and if they break the terms, they only lose their right to
redistribute. The GPL conditionally waives some copy rights.
> > I like to tinker with stuff, but the fact that I don't have source to TiVo
> > or the Nav system in my car does not decrease the utility of either.
>
> I'm very surprised that you said this.
>
> I think reasons why should be immediately obvious.
s/surprised/entertained/
s/obvious/hilarious/
;)
David Pollak
I'm not sure this discussion is valuable anymore.
You ask questions about "freedoms" and I answer, so you call me "biased".
It is good that some software is Free. It is good that there are a broad range of options for software licensing. I never suggested anything different. A software author should have the right to require authors of derivative works license those works in a particular way. But that should not mean that it's a requirement for all software as the FSF advocates.
Would I have used Scala if it were not open source? Yes. I used Java (and continue to use Java and the JVM) despite that they are not yet Open Source or Free Software (yes, the JVM is migrating towards being Free.) Would I have started using Scala if it cost more than a trivial amount to license? No.
The FSF is evil because:
- It intentionally keeps the meaning of its license terms vague (what is "linking"?) by avoiding litigation that would resolve the meaning of the license.
- It allows "friends" to interpret the meaning of the GPL differently (see how the folks at MySQL has tortured the GPL to require payment of license fees for using MySQL in commercial settings even though there's no "linking".)
- It is evil because it tried to torpedo Linux and rebrand Linux as a FSF/GNU project. This has nothing to do with helping users by enlarging their software choices and everything to do with ego and trying to maintain market clout.
- It is evil because it is mercurial with its would-be allies. The spat with the Apache foundation is the flip side of the same coin as the MySQL issue. The FSF could have commented on the Apache 2.0 license during the comment period, but decided to lie in wait and spring the who license thing on Apache at some politically opportune time.
- The FSF is evil because it strives to impose "Freedom" on all
software and all producers of software.
So, go ahead, claim (without any logical reasoning as you did early in the thread) that my logic is flawed. Go ahead, call me biased when I state my opinions. It basically boils down to name calling.
Thanks,
David
Corey
> Corey,
>
> I'm not sure this discussion is valuable anymore.
>
> You ask questions about "freedoms" and I answer, so you call me "biased".
>
My intention was not to use the word "biased" as though it were an ad hominem,
and I apologize for my lack of clarity if it was construed in that way.
My actual purpose was to express how I didn't think that your original series
of four questions were of any particular use within the context of this
debate without also identifying from which interest group one might be
identifying most closely with while answering those particular questions.
In order to try and help illuminate the nature of that view, I produced my
own series of questions (GNU/FSF's "four freedoms").
You answered those questions in a way which - in my mind - demonstrated the
very point I was getting at.
If I had answered those questions I would have done so just as honestly
and thoughtfully as you did -- and would have given a very different set of
scores!
Thus: what would have been proven aside from the fact that we do not all
universally share the same set of interests/biases on any given matter at
any given moment?
However, if I would have contextualized my answers with both scope (specific
software? all software? some software? certain software?), and an interest
group to which I was identifying (as an end-user? proprietary developer? open
source distributor?) - then we'd have something a bit more tangible that could
help us begin to find patterns of commonality and contention.
> It is good that some software is Free. It is good that there are a
> broad range of options for software licensing. I never suggested
> anything different.
>
I never suggested that you suggested anything different! (c8=
> A software author should have the right to require authors of derivative
> works license those works in a particular way.
>
Of course.
> But that should not mean that it's a requirement for all software as the
> FSF advocates.
>
It would be an understatement for me to say that I have an extremely tough
time imagining a future where all software is licensed accordingly to what
the FSF advocates.
Therefore, color me unconcerned.
In the mean time I accept and admire GPL'd software simply for what for it is
and for what it has facilitated.
There's certainly a small but not insignificant bit of rhetoric from the FSF
that often manages to rub me the wrong way, however I find no reason to
polarize my views against them simply because they do not tend to mirror my
own preferences/biases/interests/ideas to an extremely fine degree.
> Would I have used Scala if it were not open source? Yes. I used Java
> (and continue to use Java and the JVM) despite that they are not yet
> Open Source or Free Software (yes, the JVM is migrating towards being
> Free.) Would I have started using Scala if it cost more than a trivial
> amount to license? No.
>
I did not ask whether you would have used Scala if it were not open source, I
asked how your answers/scores would have changed if I were to have
replaced all instances of the phrase "the program", with "scala" in the
questionare I had provided (this is kind of getting back to the stuff I
attempted to clarify in the beginning of this email):
0. The freedom to run scala, for any purpose
1. The freedom to study how scala works, and adapt it to your needs
2. The freedom to redistribute copies of scala so you can help your
[client]/neighbor
3. The freedom to improve scala, and release your improvements to the
public, so that the whole community benefits
What's interesting here is how it is fairly easily demonstrated that my
assertion regarding how such questions are answered is dependent largely upon
one's existing bias or needs in relation to the software in question.
For instance, if you don't identify yourself as a developer or distributor of
the software in question - then you will likely place low scores for #1, #2 &
#3... being a 'mere' consumer of the software, you may only really concern
yourself with #0.
Now, were we to ask any of the scala core-developers the same 4 questions, and
we'd no doubt see high scores in #1 and #3 (in addition to #0 of course).
Go ask redhat the same question - you're sure to see a high score in #2 as
well.
There's no inherent crime in being biased differently accordingly to one's
dynamically re-organizing set of interests... isn't that basically a facet
of being a conscious being?
> The FSF is evil because:
>
> * It intentionally keeps the meaning of its license terms vague
> (what is "linking"?) by avoiding litigation that would resolve the
> meaning of the license.
> * It allows "friends" to interpret the meaning of the GPL
> differently (see how the folks at MySQL has tortured the GPL to
> require payment of license fees for using MySQL in commercial
> settings even though there's no "linking".)
> * It is evil because it tried to torpedo Linux and rebrand Linux as
> a FSF/GNU project. This has nothing to do with helping users by
> enlarging their software choices and everything to do with ego and
> trying to maintain market clout.
> * It is evil because it is mercurial with its would-be allies. The
> spat with the Apache foundation is the flip side of the same coin
> as the MySQL issue. The FSF could have commented on the Apache
> 2.0 license during the comment period, but decided to lie in wait
> and spring the who license thing on Apache at some politically
> opportune time.
> * The FSF is evil because it strives to impose "Freedom" on all
> software and all producers of software.
>
I keep trying to figure out how best to respond to the above; but I'm at a
loss. I'm not a spokesperson of the FSF, so I'm not going to take the time
to take the podium in their defense on each of the items you've listed.
The briefest, most direct comment I can make here is that you have not
actually provided anything to convince or persuade that the FSF is _evil_ -
merely that they are possibly (likely) fallable/imperfect... that they have
made one or more tactical/social mistakes of somewhat varying, relatively
confined breadth and significance. I see what might amount to one or more
minor transgressions - I don't see evil; especially when contrasted against
the demonstrably _good_ things the FSF have directly or indirectly been
involved in.
( I wonder what your list of Microsoft evils would look like... )
Heck, I could make a case that _I'm_ 'evil', were I to fixate on a shortish
list of a few bad/questionable/debatable things I've said or done. I could
also make myself look like a living saint were I to do the opposite.
> The FSF is more evil than Microsoft because Microsoft's motives are
> transparent: market share and profits. The FSF's motives are more
> mercurial and change at RMS's whim.
>
I disagree that the FSF's motives are less transparent or consistent than
Microsoft's.
I also disagree with your (outlandish?) assertion that the FSF is "more evil"
than Microsoft. I plainly disagree that the FSF is at all 'evil'[1] in any
real sense of the word. You'd get laughed off the stage if you made similar
claims regarding McDonald's vs. Food Not Bombs[2], or WallMart vs. The
Goodwill[3].
I fail to understand the sort of logic that downplays the oft-negative means
and ends of the pure profit-motive ("market share and profits") which
drives a mega corporation such as Microsoft, and exaggerates an arguable -
relatively benign at worst - flaw ("mercurial") of a grassroots, non-profit
organization such as the FSF which has helped put a truly amazing volume of
wealth and empowerment, via libre software, into the hands of untold thousands
of individuals, organizations and corporations the world over.
[1] http://www.hyperdictionary.com/search.aspx?define=evil
[2] http://en.wikipedia.org/wiki/Food_not_bombs
[3] http://www.goodwill.org/page/guest/about/howweoperate/shop
> So, go ahead, claim (without any logical reasoning as you did early in
> the thread) that my logic is flawed. Go ahead, call me biased when I
> state my opinions. It basically boils down to name calling.
>
Again I apologize for that misunderstanding; I don't think that it is
wrong/bad to be biased, or that it is an insult to claim that people and
markets are driven by interests.
I don't see quite where I went wrong in my attempt at providing logical
reasoning, in fact I was making every effort to do so! I'd be interested in
giving it another shot if you point out specifically where my logic was less
than satisfactory.
Regards,
Corey
Ian Holsman
if I can insert my 2c's.
every license is restrictive in some form or another. I belive that the
people who draft the license have the best possible interests, but are
not perfect. As such people find holes and drive trucks through them.
at the end of the day it is up to the author to decide which best meets
their criteria/objectives.
going GPL (2/3) or BSD is now a business decision, and in most cases
more about the competitive and control aspects of what the license
allows the company to do than the greater social good, in some cases
going GPL is about protecting revenue streams in other areas.
As far as the 'spat' between the ASF and the FSF, I didn't see one (I'm
a member of the ASF). They had different interpretations on warranty and
patents, and what was implied and what had to be explictly stated. There
were regular meetings between both parties lawyers during the formation
of GPL3.
as for the JVM being free, try to get it running on a JVM and you'll see
how it free it is to get a certified JVM there.
regards
Ian
David Pollak wrote:
> Corey,
>
> I'm not sure this discussion is valuable anymore.
>
> You ask questions about "freedoms" and I answer, so you call me "biased".
>
> It is good that some software is Free. It is good that there are a
> broad range of options for software licensing. I never suggested
> anything different. A software author should have the right to
> require authors of derivative works license those works in a
> particular way. But that should not mean that it's a requirement for
> all software as the FSF advocates.
>
> Would I have used Scala if it were not open source? Yes. I used Java
> (and continue to use Java and the JVM) despite that they are not yet
> Open Source or Free Software (yes, the JVM is migrating towards being
> Free.) Would I have started using Scala if it cost more than a
> trivial amount to license? No.
>
> The FSF is evil because:
>
> * It intentionally keeps the meaning of its license terms vague
> (what is "linking"?) by avoiding litigation that would resolve
> the meaning of the license.
> * It allows "friends" to interpret the meaning of the GPL
> differently (see how the folks at MySQL has tortured the GPL to
> require payment of license fees for using MySQL in commercial
> settings even though there's no "linking".)
> * It is evil because it tried to torpedo Linux and rebrand Linux
> as a FSF/GNU project. This has nothing to do with helping users
> by enlarging their software choices and everything to do with
> ego and trying to maintain market clout.
> * It is evil because it is mercurial with its would-be allies.
> The spat with the Apache foundation is the flip side of the same
> coin as the MySQL issue. The FSF could have commented on the
> Apache 2.0 license during the comment period, but decided to lie
> in wait and spring the who license thing on Apache at some
> politically opportune time.
> * The FSF is evil because it strives to impose "Freedom" on all