MIT vs GPL vs LGPL for open source project

Rails
61

On Apr 11, 2006, at 12:58 AM, Craig W. wrote:

Simply using the term ‘viral’ with specific reference to GPL is
pejorative, which I suppose is your right as you are entitled to an
opinion but it is only that - an opinion and it is the language
that is
unfair.

While you may decide for yourself whether it is pejorative, it is
intended to be denotative and to describe the property of the
license. The analogy to virus is quite apt, explains the situation
well to a layman. Understandably, the lack of any better term has
led to its wide adoption among attorneys working in the arena.
Perhaps you would care to suggest an alternative? While you are at
it, you might ease up in calling perfectly reasonable licenses “non-
free.” (or did you mean that in a denotative and non-pejorative sense?)

In the sense that you are suggesting that the GPL license is
viral…that doesn’t begin to approach viral strategies such as Sony’s
embedding a rootkit on people’s computers merely by inserting one
of the
many CD’s that they have published, such as DCMCIA legislation being
pushed by lobbyists or the absurd patent practices in the United
States.

I tend to refer to the latter as criminal violations of the Computer
Fraud and Abuse Act and various state law claims. Cute analogies to
defend unconsented seruptitious code infused without consent or
knowledge is indefensible. We agree. So what? What bearing does
this have on the licensing issues?

I would suggest that the American consumer is footing the bill for the
digital rights protections, the pharmaceutical companies protections,
etc. and there are many who feel that software patents should
simply not
ever be allowed.

We agree. So what? What bearing does this have on the licensing
issues?

Anyway, I digress - but only to one more point…that samba is an
excellent example of GPL license. Microsoft can’t buy it…IBM
can’t buy
it, etc. As someone who has seen projects go cold when the company
released the software under BSD/MIT type license was sold to another
company that moved all development to proprietary/in house software, I
readily see the value of GPL.

So that’s your best case in support of the GPL? If so, quod erat
demonstrandum! Your anecdotal suggestion is interesting, and details
would be most welcome if provided. Frankly, your suggestion that the
BSD license resulted in software “going cold” doesn’t make any legal
or commercial sense to me – the open source code remains publicly
available, and if the company is the only one willing or capable of
keeping it up to date, I’m not sure I see the value in the software.
This is the scenario that forks were made for.

And, by the way, there is no reason to believe that, had the code
been released under the GPL the same result could occur. A company
that owns a copyright, and has not accepted donated code from others
subject to the GPL, is free to relicense the code on such terms as it
sees fit. Even if it had accepted some contributions, it could
simply excise those changes or clean room them out of the code. Dual
and multiple licensing is not only common, these days, but old news.

An even sadder proposition is the possibility that any mode of
license might ultimately be revocable.

GPL doesn’t stop this parade of horribles, either as a matter of law
or as a practical mattter.

For most purposes, a GPL license is mostly
void of opportunities for an IP lawyer to make a buck. That has its
own
appeal.

And what would a vapid argument be without a lawyer-bashing
conclusion? In fact, Craig is mistaken. I just last week gave a CLE
presentation at the mid-Winter meeting of the Business Law Section of
the ABA on this very subject. Based on the talks given by me and my
bretheren on the panel, it would appear, quite to the contrary, that
there is quite a buck for IP and M&A lawyers who deal with open
source licenses.

ps…if you respond, please don’t refer to me as ‘your brother’ as I
would interpret that as condenscending.

Of course you would. The prior fraternal reference was not intended
as condescending. The first sentence of this paragraph most
certainly was. Between you and me, Craig (who is not my brother),
trying to dictate my choice of words --the sole basis of your
response-- is a hopeless and inadequate mode of argument. Personal
and ad hominem remarks add little more.

62

On Apr 11, 2006, at 9:21 AM, Craig W. wrote:

pejorative, considering the connotations of the word viral, especially
in the sense of computer technology. It’s simply being argumentative.

And, while I may reasonably disagree with you, I still await your
proposed alternative. Frankly, I see it more as quibbling than
offensive, but I am pleased to adopt any other term that reasonably
accounts for the behavior, and conveys the denotative meaning I
intend to non-technical individuals. Please note that the horse is
long since out of the barn – the adjective is the generally accepted
usage.

And what would a vapid argument be without a lawyer-bashing
conclusion?

Vapid or not, it’s a tough temptation to pass :wink: Present evidence
withstanding, especially for those that are incapable of adapting
their
language to the audience they are speaking to.

You seem to feel that I must “adapt” my language to suit your taste
and avoid the commonly used term to describe the behavior, all just
because you say so. And then you make an empty lawyer-bashing remark
in lieu of argument, and defend it on that ground. Fair is fair,
paper doesn’t refuse ink, and e-mails posted will be sent. Just
don’t pretend you are actually making an argument.

You are conveniently missing my perception of your comments. If your
intent was to refer to Matthew as ‘brother Matthew’ was not to be
condescending, then I missed it. In fact, I mentioned that I saw it as
condescending. I sought not to dictate your choice of words but rather
provide reference that I expected to you understand the boundaries
of a
discussion.

All evidence to the contrary. You seem to be hell-bent on dictating
my choice of words rather than discussing issues on the merits. If
it is any comfort, I apologize. I disagree with you, and don’t think
you are my brother.

general demeanor more suitable for your law practice then a mail list
for ruby on rails.

Indeed, my law practice is far more collegial than this colloquy.
Fear not, I shall not respond to you again.

63

Austin Z. wrote:

unfair.
“viral marketing.” It’s not pejorative. The GNU GPL practices "viral

What if we say that the GNU GPL is retroviral? :wink:

I did not follow through the entire thread, if I make mistakes below I
apologize ahead. Searching through the word ‘viral’ and ‘virus’ through
a couple of dictionaries (including Webster and Oxford) turned out
definitions and descriptions that are ALL negative, none are neutral.

I suppose it boils down to what you personally mean by ‘viral’. Do you
mean ‘viral’ to mean that it is self-replicating? Or that GPL is
‘catching’ like a flu virus?

GPL is neither. It is not self-replicating because GPL cannot reproduce
by itself – you need to consciously copy or derive from GPL protected
code for GPL to be mandatory on your code. It can’t do anything. It
is not ‘catching’ because it doesn’t force you to make your code GPL,
unless you want to. If you don’t want to, just don’t copy or derive GPL
protected code :slight_smile: Perhaps you might accidentally use some GPL code in
your own or link to it unknowingly. If you don’t want to make your code
GPL, just unlink it, or remove that GPL code and you’re home free. If
you insist on using it or deriving from it, then you can’t complain,
there is no such thing as a free lunch, if you use it, you must pay up
and in this case you need to make your code GPL.

Say you are the developer and you released code as GPL. Say a while
later you want to make it commercial and sell it for tonnes of money.
GPL doesn’t prevent you from doing that, because you are the copyright
owner (unless of course you took contributions from the FOSS community)
and you can readily release your code with another license. Point is –
no one else can, except you. You can even release it under the MIT
license later.

Having said as much, I’m happy that Rails is released under MIT but I’ll
be equally happy if Rails is released under GPL. I’m just glad that it
was released. :slight_smile:

http://lists.rubyonrails.org/mailman/listinfo/rails

–
Sau S.

http://blog.saush.com
http://read.saush.com
http://jaccal.sourceforge.net

64

I am quite certain that much of the Linux kernel development would not
be as pronounced if it weren’t GPL license. In fact, I would say that
the GPL license is a likely reason that Linux kernel development pace is
much more rapid than say BSD kernel development.

I completely agree. Some people may have a dislike of the GPL but
without it I seriously doubt we’d have such a rich open source eco
system.

65

2006/4/11, Craig W. [email protected]:

The simple fact is that contributing code to a GPL licensed project
means that your code contributions cannot be used commercially without
the commercial product also disclosing their code base too.

This was affirmed by Cisco/Linksys in their usage of Linux in their
routers - http://en.wikipedia.org/wiki/Linksys

The take away of that really must be that Cisco/Linksys were not
“contributing code to a GPL licensed project” intentionally, to them
it must have felt like they were infected by the GPL, and had to shed
the whole thing to protect themselves.

GPL code doesn’t get into applications by itself, and it doesn’t need
a host to run, so by two counts it isn’t viral. Perhaps “trojan horse”
is a better description - if you let GPL code into your distribution,
Troy will burn.

Douglas

66

On Tue, 2006-04-11 at 23:06 +0800, Chang Sau S. wrote:

that is
The problem is, that it’s not an incorrect term. Consider the term
similarly neutral.

use some GPL code in your own or link to it unknowingly. If you don’t
community) and you can readily release your code with another license.
Point is – no one else can, except you. You can even release it
under the MIT license later.

Having said as much, I’m happy that Rails is released under MIT but
I’ll be equally happy if Rails is released under GPL. I’m just glad
that it was released. :slight_smile:

It is the pejorative usage of the term viral, the negative connotation
that has all of the detractors affixing the term to the GPL license in
an attempt to frame the argument itself which is unfair. It is not
correct to refer to GPL license as viral, only convenient for those
who wish to characterize the license negatively by adopting language
furthered by it’s detractors, such as Microsoft.

The simple fact is that contributing code to a GPL licensed project
means that your code contributions cannot be used commercially without
the commercial product also disclosing their code base too.

This was affirmed by Cisco/Linksys in their usage of Linux in their
routers - http://en.wikipedia.org/wiki/Linksys

and I note that they have since switched the base code from Linux so
they would not have to disclose their code in the future, as is their
right.

I am quite certain that much of the Linux kernel development would not
be as pronounced if it weren’t GPL license. In fact, I would say that
the GPL license is a likely reason that Linux kernel development pace is
much more rapid than say BSD kernel development.

Craig

67

On Wed, 2006-04-12 at 02:11 +0200, Douglas L. wrote:

The take away of that really must be that Cisco/Linksys were not
“contributing code to a GPL licensed project” intentionally, to them
it must have felt like they were infected by the GPL, and had to shed
the whole thing to protect themselves.

this is all conjecture as we have no explanation for why they chose
Linux in the first place or that they felt burdened (or infected) to
release their code. I merely recounted that they complied with the GPL
license.

GPL code doesn’t get into applications by itself, and it doesn’t need
a host to run, so by two counts it isn’t viral. Perhaps “trojan horse”
is a better description - if you let GPL code into your distribution,
Troy will burn.

hardly a ‘let’ situation - if the code maintainers are so slack as to
‘let’ GPL code into a project that they didn’t want to license under
GPL, they are at best ignorant.

In fact, the Linux based code base on the Linksys routers was part of
the acceleration of the many open source implementations - see

Linksys ultimately decided to use VxWorks to get by with less RAM/Flash
and ultimately make a cheaper product and that seems to follow the
industry.

There is no reason to conclude that Linksys is not interested in a Linux
router and in fact, they retooled their WRT54G into the WRT54GL
(L=Linux) and they are selling quite well.

Craig

68

On 4/11/06, Douglas L. [email protected] wrote:

the whole thing to protect themselves.

GPL code doesn’t get into applications by itself, and it doesn’t need
a host to run, so by two counts it isn’t viral. Perhaps “trojan horse”
is a better description - if you let GPL code into your distribution,
Troy will burn.

There are virii which require positive action by the host to acquire.

Between the two choices given so far (“trojan horse” and “viral”),
“viral” is the far better.

I do not think that the GNU GPL is necessarily a bad choice, and I think
that the “viral” nature is a feature to ensure that certain software
can’t be abused by others. But it is something to be cautious of.

-austin