"proprietary", was Re: ada on linux

"proprietary", was Re: ada on linux

[tmoran]

> > > Well most often proprietary in software has been used to
> > > describe software which is only released under restrictive
> > > license terms.
>
> > Such as requiring that changed source be released with changed
> > binaries ?
>
> No, such as not permitting redistribution at all, and not
> providing the sources at all.
"Anyone may use in any way" is not restrictive.  "You must send money
to ___ if you give out copies to third parties" and "You must send source
code to ___ if you give out copies to third parties" are both restrictions.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <u5XX4.297$q86.81131@news.pacbell.net>,
  tmoran@bix.com wrote:

> "Anyone may use in any way" is not restrictive.  "You must
> send money
> to ___ if you give out copies to third parties" and "You must
send source
> code to ___ if you give out copies to third parties" are both
restrictions.

I will repeat what I said earlier, I think there is a
significant and important distinction between the position
that you cannot redistribute under any circumstances, and
the position that you can freely redistribute what you
received!

The above second restriction is *NOT* an accurate
characterization of the GPL restrictions. If you get
a copy of something under the GPL, you can freely
redistribute it. To fulfill the requirement for
distributing sources, you can simply point to the
original distribution. For example, if you download
the public binary version of GNAT, you can freely
redistribute it to others, you do NOT have to
distribute the sources yourself (the GPL never
requires distribution of sources along with the
objects, it just requires they be made available).
So when you pass on the binaries, people can get
the sources from the original location (e.g. cs.nyu.edu)
and that's just fine.

Tom, you seem to bat almost 100% when it comes to
misinterpreting the GPL. Time to reread it carefully :-)
Like any software license agreement, it needs careful
reading! To be fair, you did not mention the GPL in the
above, but I am willing to bet it is what you had in mind
when you wrote:

  "You must send source code to ___ if you give out copies to
   third parties"

The restriction that Tom alludes to above is a different
one. It affects only derived works. Here the restriction
is that if you create derived works (not permitted at all
with typical proprietary software), then you can still
distribute them, but you must distribute your mods in
source form (e.g. as a patch file).

Once again, with regard to the use of the word proprietary.
There is a very big difference between Free Software that
permits unrestricted redistribution and guarantees that the
sources are always available, and typical software (e.g. from
Microsoft or Ada vendors other than ACT) that significantly
restricts or forbids redistribution. Usually people use the term
proprietary software to refer to the latter. I don't really care
what term people use, just so long as they recognize the
distinction!

People are still often confused by this distinction. Some months
ago, I had to spend quite a bit of time on the phone explaining
to a procurement guy at Lockheed that he did NOT need us to
execute a source escrow agreement, because we had already
supplied all the sources, and Lockheed had all the access that
they would get from opening a source escrow (and more) already.
Finally he understood, and agreed that "OK, I guess we don't
need a source escrow in that case" :-)

Robert Dewar
Ada Core Technologiesh


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Ken Garlington]

"Robert Dewar" <dewar@gnat.com> wrote in message
news:8gppqa$og7$1@nnrp1.deja.com...
> Some months
> ago, I had to spend quite a bit of time on the phone explaining
> to a procurement guy at Lockheed

Lockheed Martin, please. The distinction is subtle but important.

> that he did NOT need us to
> execute a source escrow agreement, because we had already
> supplied all the sources, and Lockheed had all the access that
> they would get from opening a source escrow (and more) already.
> Finally he understood, and agreed that "OK, I guess we don't
> need a source escrow in that case" :-)

A procurement guy who finally understood? OK, maybe it's not Lockheed Martin
after all :)

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <gj_X4.2373$s64.80377@news.flash.net>,
  "Ken Garlington" <Ken.Garlington@computer.org> wrote:
> Lockheed Martin, please. The distinction is subtle but
> important.

Indeed! Sorry for the excessive abbreviation in my post



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

>I will repeat what I said earlier, I think there is a
>significant and important distinction between the position
>that you cannot redistribute under any circumstances, and
>the position that you can freely redistribute what you
>received!
  I fully agree.  How frequently does either position actually occur?  I
don't know of anybody who sells software with the requirement that you
cannot redistribute under any circumstances.  Even Microsoft allows
retailers and system vendors to redistribute their software.  The
cirumstances, of course, include paying Microsoft royalties.
  It's true that I understand the GPL to say that you cannot "freely"
redistribute what you received.  As I (mis?) understand it, if you
redistribute, there are certain requirements, ie, you are not free to do
whatever you want.  Perhaps the restriction that you must include a sentence
pointing to where the source can be found is not an onerous one in the usual
case, but it *is* a restriction on "freely".
  To construct a thought experiment, suppose you write a study guide for
the SAT.  To let your readers test themselves, you include on a CDROM the
binary of a GPL'ed SAT simulation program.  As I understand it, you
haven't fulfilled the requirements.  But if you include the source code,
or even a pointer to it, some people are going to look at the source code,
find out the questions and their correct answers, and then "ace" the
simulated test.  Now it may be that someone who does such a thing with a
practice SAT test is a fool.  Suppose we change the scenario slightly, and
have the test be one for an on-line college course.  Now the cheater will
get credit for something he does not in fact know, and perhaps be hired in
some position where an ignorant person can do significant damage.  In the
absence of the "source availability" requirement, you could have avoided this.
If you used a non-GPL program, its author almost surely would be happy to
let you redistribute, though he might want a small royalty per CDROM.
  If you offer a product that is, or uses, GPLed software, are you
effectively required to charge for support?  If you don't, and some
users have made "just a little improvement", you face a heck of a lot
of support time tracking down their "just a little mistake"s.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <ek2Y4.340$q86.98777@news.pacbell.net>,
  tmoran@bix.com wrote:
>   I fully agree.  How frequently does either position actually
> occur?  I don't know of anybody who sells software with the
> requirement that you cannot redistribute under any
> circumstances.  Even Microsoft allows retailers and system
> vendors to redistribute their software.  The
> cirumstances, of course, include paying Microsoft royalties.

Seems like sophistry to me. The license does NOT permit any
redistribution. Microsoft retailers do NOT redistribute copies
in the sense we are talking about. Yes, system vendors can
redistribute software under special licenses, but I was talking
specifically about the typical license. FOr example, if you
buy a copy of PowerPoint, you MAY NOT redistribute copies of
this copy under the license you received, so yes, it happens
all the time. In fact I assume that CLAW, being a proprietary
product in this sense, also comes with a license that forbids
unrestricted redistribution.

>  It's true that I understand the GPL to say that you cannot
> "freely" redistribute what you receive

Well you misunderstand, if you receive a software package under
the GPL, you are absolutely free to redistribute what you
receive. Restating your misunderstanding repeatedly does not
change that it is a misunderstanding. After a while I have to
wonder why you are insisting on repeating this false statement.
The whole idea of the GPL in connection with freely distributed
products like GNAT is not only to permit, but to encourage
such secondary distribution.

d.  As I (mis?) understand it, if you
> redistribute, there are certain requirements, ie, you are not
> free to do whatever you want.

In terms of redistributing the original that you received,
assuming the original distribution was legal (i.e. conformed
to the GPL), then yes, you can freely redistribute what you
received. How many more times does this need to be said? It
happens all the time with GNAT. People grab the file from
cs.nyu.edu, and then send copies of that file to their friends
or put it up on their own site. That is perfectly fine and
perfectly in accordance with the GPL.

As I said in my note, the requirement for distributing sources
comes into play when deriviative works are created, a whole
different ball game.

> Perhaps the restriction that you must include a sentence
> pointing to where the source can be found is not an onerous
> one in the usual case, but it *is* a restriction on "freely".

You really *insist* on this misunderstanding I guess.
There is no requirement for you to include a sentence of this
nature if you are simply redistributing what you received
assuming that this information was present in the original.
(which it most certainly should be).

>   To construct a thought experiment, suppose you write a study
> guide for the SAT.  To let your readers test themselves, you
> include on a CDROM the
> binary of a GPL'ed SAT simulation program.  As I understand
> it, you
> haven't fulfilled the requirements.

Let's assume you are NOT the original author of the simulation
program here (the original author is of course not restricted
in any way by the GPL, if you do not understand that, you are
really at sea -- a license *I* issue cannot restrict me!

Assuming that, the distribution of the GPL'ed SAT simulation
program is a copyright violation on its face.

<<irrelevant stuff snipped>>

(irrelevant since it simply seems to be an argument that the
GPL is not the right license to use in this situation. So
what? What's that have to do with it? Have you seen me argue
that all software should be distributed under the GPL? The
answer is no you have not, since I have never made such a
statement.)

> If you offer a product that is, or uses, GPLed software, are
> you effectively required to charge for support?

AARGH! This is so far off base that I don't know what to say.
I guess I will just say no and leave it at that

> If you don't, and some
> users have made "just a little improvement", you face a heck
> of a lot of support time tracking down their "just a little
> mistake"s.

Again, completely out of left field. Just because you offer
a product which is GPL'ed does not create any support
obligations at all. If you do offer support, then the
terms and conditions of the support will deal with the
issue of modifications. This is not something new with the
GPL, mainframe software has often been issued with full
sources, with the understanding that the customer may
(or even must) make changes, and then the support agreement
has to spell out the obligations on both sides.

In the case of GNAT, our support is premised on use of the
binary versions that we distribute. In the cases where
customers make changes, we continue to support on a best
efforts basis, or in some cases, special support agreements
are arranged.

This is all really quite simple, I can't understand why you
are making such heavy weather of it, unless your goal is
simply to encourage people to agree that it is complex.
Certainly I understand that you (or rather the company you
work for) have decided to use restrictive licenses, and that
is definitely your privilege, and may indeed be the best
choice for you, since I know nothing about your business I
could not comment.

But the GPL model is quite simple, and works very well for us,
and is really not that difficult to understand. These days we
even find that procurement agents at companies like
Lockheed Martin understand it quite fine :-) :-)

Robert Dewar
Ada Core Technologies


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

>The license does NOT permit any redistribution.  Microsoft
I think you mean "the kind of license typically given to an end
user allows him to run, but not redistribute, the program".  And
other licenses, typically given to retailers, allow
redistribution, and still others given to system vendors may
allow a variety of different things.  The point is, your statement
sounds like a general one about licenses (or at least "proprietary
software" licenses) but in fact it's a specific statement about
just one kind of license, and in that way misleading.

>Well you misunderstand, if you receive a software package under
>the GPL, you are absolutely free to redistribute what you
  If you receive a cake, are you free to redistribute it only in
its entirety, or in slices?  Does the GPL allow you the freedom to
redistribute portions of what you received, for instance, just the
binary?  Does it allow you to examine the software, decide it has
a very low probability of bugs, and then to redistribute it backed
with a warranty?  Your "absolutely free" is in fact a freedom with
very definite, and sometimes onerous, restrictions, and in that
way misleading.

> > If you offer a product that is, or uses, GPLed software, are
> > you effectively required to charge for support?
> Again, completely out of left field. Just because you offer
> a product which is GPL'ed does not create any support
> obligations at all. If you do offer support, then the
> terms and conditions of the support will deal with the
> issue of modifications.
  Is it really practical to offer GPLed software with source code
and support like "We offer no-cost support for this product, as
long as you have made no changes.  If you have changed anything,
it makes it harder for us to help you, so you must pay $xxx for
support.  If you say you've made no changes, and we spend time
helping you, and it turns out you have in fact made changes, then
you must $yyy for the time we've spent."  I suspect that in
practice, ie, "effectively", you will find that you *must* charge for
support, or you'll lose a lot of time and money giving free
support to some people who create problems and then lie about the
source of those problems.

>GPL is not the right license to use in this situation.
If Jones writes a program, can he offer it under different
licensing arrangement to different people, one of those
arrangements being the GPL?

>Certainly I understand that you (or rather the company you
>work for) have decided to use restrictive licenses, and that
  Calling other people's licenses "restrictive" and your own
"absolutely free", is not, IMHO, a fair statement of the case.

Re: "proprietary", was Re: ada on linux

[David Starner]

On Sun, 28 May 2000 05:41:30 GMT, tmoran@bix.com <tmoran@bix.com> wrote:
>  To construct a thought experiment, suppose you write a study guide for
>the SAT.  To let your readers test themselves, you include on a CDROM the
>binary of a GPL'ed SAT simulation program.  As I understand it, you
>haven't fulfilled the requirements.  But if you include the source code,
>or even a pointer to it, some people are going to look at the source code,
>find out the questions and their correct answers, and then "ace" the

Why would questions and answers be stored in the source code?

>simulated test.  Now it may be that someone who does such a thing with a
>practice SAT test is a fool.  Suppose we change the scenario slightly, and
>have the test be one for an on-line college course.  Now the cheater will
>get credit for something he does not in fact know, and perhaps be hired in
>some position where an ignorant person can do significant damage.  In the
>absence of the "source availability" requirement, you could have avoided this.

No, you couldn't have. If you wrote the obvious code, the test will be
extractable by running 'strings' over the binary. Even without that,
you can still disassemble the code / watch the memory of the running
program and get the test. If you give them the answers in any form, 
someone with enough patience will be able to get them, and if it
actually matters someone will do it.

Don't distribute the test. Run it over the net, then you don't have to 
send them the answers before hand.

-- 
David Starner - dstarner98@aasaa.ofe.org
http/ftp: x8b4e53cd.dhcp.okstate.edu
Roleplaying Anonymous - a 1d12 step program

Re: "proprietary", was Re: ada on linux

[David Starner]

On Sun, 28 May 2000 22:48:38 GMT, tmoran@bix.com <tmoran@bix.com> wrote:
>support, or you'll lose a lot of time and money giving free
>support to some people who create problems and then lie about the
>source of those problems.

Is this not true for any type of support? I've read enough horror
stories about tech support on stuff that doesn't come with source
to find this limited to stuff that comes with source.

-- 
David Starner - dstarner98@aasaa.ofe.org
http/ftp: x8b4e53cd.dhcp.okstate.edu
Roleplaying Anonymous - a 1d12 step program

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <anhY4.1121$M72.360470@news.pacbell.net>,
  tmoran@bix.com wrote:
> >The license does NOT permit any redistribution.  Microsoft
> I think you mean "the kind of license typically given to an
end
> user allows him to run, but not redistribute, the program".

> And
> other licenses, typically given to retailers, allow
> redistribution

That's just completely wrong. You are inventing, and you
are inventing wrong. Don't guess and publish it as fact :-)
A retailer buys a copy of a program like Power Point. Now
under 117, the retailer owns a copy of the progrma (not the
program itself but a copy). The retailer is free under the
copyright law to sell this copy to someone else (if you
buy it, you inherit this right, and can then sell your
unopened copy to someone else.) This right comes from the
copyright law, not from any license. In fact the vendor
never executes any license agreement with regard to this
particular copy of Power Point (there may be a separate
sales contract, but that's different). If you sell your
unopened copy, you also never execute a license agreement
(the shrink wrap license comes into effect only when the
shring wrap is opened).

> and still others given to system vendors may
> allow a variety of different things.

Yes, that's right, see the Microsoft trial for many details
about such licenses.

>  The point is, your
> statement
> sounds like a general one about licenses (or at least
"proprietary
> software" licenses) but in fact it's a specific statement
about
> just one kind of license, and in that way misleading.

I am talking about end user licenses, and have always made
that clear. yes, system vendors do have other licenses.
Retailers do not, that's just a Moran invention :-)

>   If you receive a cake, are you free to redistribute it only
in
> its entirety,

For sure in its entirety, which is in practice the critical
issue, and is what distinguishes it from proprietary products
like CLAW. I realize that you are trying to minimize the
difference here Tom, but it is really a VERY important
difference, and trying to obfuscate this fact does not
change the fact that there is for example a very big difference
between CLAW and GtkAda here.

> Does it allow you to examine the software, decide it has
> a very low probability of bugs, and then to redistribute it
> backed with a warranty?

Yes, of course. Why are you asking questions like this whose
answer should be obvious if you read the license. These
questions seem designed to obfuscate.

> Your "absolutely free" is in fact a freedom
> with very definite

Noone said you are absolutely free to do anything you want.
That's an invention on your part.

> and sometimes onerous, restrictions, and in
> that way misleading.

Onerous in one case in particular, which is the case where
a proprietary software vendor wants to redistribute a
derived work in proprietary form. For example, I quite
see that the CLAW folks might like to use GtkAda to
make a proprietary version of CLAW for non-NT targets.
Well they definitely cannot do this.


>   Is it really practical to offer GPLed software with source
> code
> and support like "We offer no-cost support for this product,
> as
> long as you have made no changes.  If you have changed
anything,
> it makes it harder for us to help you, so you must pay $xxx
for
> support.

Yes, it is practical, and basically that is the situation at
ACT, we support GNAT as we distribute it, and will fix any
problems that are found. If you want us to guarantee to fix
problems in versions you have modified, we can do this but
we definitely write specialized (more expensive) support
contracts in this place.

What on earth would make you think it is impractical? It is
the only approach that makes sense.

> If you say you've made no changes, and we spend time
> helping you, and it turns out you have in fact made changes,
> then
> you must $yyy for the time we've spent."  I suspect that in
> practice, ie, "effectively", you will find that you *must*
charge for
> support, or you'll lose a lot of time and money giving free
> support to some people who create problems and then lie about
> the source of those problems.

You know nothing about distributing GPL'ed software and
supporting it, because you have never done it. I suggest
people take the actual experience of those who DO distribute
GPL'ed software over the suspicions of someone with no
experience at all.

I quite understand the kind of arguments people like Tom (CLAW)
use to justify choosing to go the proprietary route, but these
arguments are often misinformed, as in this case. This does not
mean that it is wrong to choose the proprietary route, but
certainly uninformed suspicions should not be the basis for
the decision.

> >Certainly I understand that you (or rather the company you
> >work for) have decided to use restrictive licenses, and that
>   Calling other people's licenses "restrictive" and your own
> "absolutely free", is not, IMHO, a fair statement of the case.

I understand how those who have chosen a proprietary route
would like to minimize the differences, since they have a
commercial interest in doing so, but in trying to make this
argument, you should not exaggerate. Only *you* have called
the GPL absolutely free. All I said was that one characteristic
of the GPL is that you are absolutely free to redistribute what
you received. By this I mean redistributing the entire product,
I am not talking about deriviative works.

It is definitely the case that the GPL is like the Microsoft
License in that, for example, both ACT and MS are selling
copies of copyright license with a license that places
restrictions on the user. For more details on this, see
the transcript of the GEAC vs GRACE trial in Newark Federal
District Court, where I testified in detail on this point
(the attorney for the other side being confused like Tom :-)
The difference is in the qualitative nature of the restrictions.
The reason that lawyers find the GPL odd as a license agreement
is that it reads as being more interested in ensuring what the
user CAN do than what they CANNOT do.

If you can't see that CLAW has a much more restrictive license
than GtkAda (I am choosing vaguely competing products here),
then you are deluding yourself, and you will not serve your
customers well by this delusion.

As I said earlier, companies choose licenses to best meet their
interests. No one ever said that the GPL is the right license
for everything. In fact as you should know well, Ada Core
Technologies considers the GPL to be unsuitable for a
considerable chunk of the technology (including the Ada runtime
and ASIS for example).

Customers choose products on many factors, including the
license. Some license provisions are important some are not.
Actually for most of our customers, the redistribution right
is not particularly relevant, big companies are not in the
business of doing extra work to distribute copies of software
to their potential competitors. For them, what is important
about the GPL is the accessibility of the sources, and the
open source nature of the technology in terms of development.

To repeat, I have absolutely no criticism of the decision to
use a proprietary license, by which I mean a license that is
considerably more restrictive than the GPL, and in fact more
akin to the Microsoft license, for CLAW. Many, in fact most
of our tool partners (including the CLAW folks) choose to
license their software in this manner.

There are definitely arguments that can be made that this
helps increase revenue to the company involved. Whether these
are correct arguments or not depends on the circumstances.
Certainly the automatic assumption that a proprietary license
increases revenues is suspect.

However, it is really not possible to construct an argument that
a more restrictive license is beneficial to your users (other
than perhaps arguing that it decreases costs). In terms of
user benefits, a more permissive license is a clear win in
all terms.

Robert Dewar
Ada Core Technologies


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

Thank you for clarifying things:

> > a very low probability of bugs, and then to redistribute it
> > backed with a warranty?
> Yes, of course. Why are you asking questions like this whose
> answer should be obvious if you read the license.
  Sorry, I had looked at the license included in the Gnat library sources,
and not at the line in the NO WARRANTY section of the FSF GPL that does
indeed say the parties may contract in writing for a warranty.

>Noone said you are absolutely free to do anything you want.

>I am talking about end user licenses, and have always made
>that clear.

> and support like "We offer no-cost support for this product,
>...
>ACT, we support GNAT as we distribute it, and will fix any

>By this I mean redistributing the entire product,

>both ACT and MS are selling copies of copyright license with a
>license that places restrictions on the user.

  I'm glad you've now made those things clear.

>I quite understand the kind of arguments people like Tom (CLAW)
  You misunderstand here.  Tom is not CLAW or vice versa.  Tom
has worked on CLAW on occasion for RR Software, but is not an
employee, not an executive, and most certainly not a person with
authority over RR Software licensing terms.  Tom has, however,
read an economics book and recognizes how a monopolist may
increase profits by setting different prices for different
customers.

>Certainly the automatic assumption that a proprietary license
>increases revenues is suspect.
  Agreed. "Give away the razor, charge for the blades" often
brings greater revenues.

>In terms of user benefits, a more permissive license is a clear
>win in all terms.
  Agreed.  But a *differently permissive* license may or may not be a
win, depending on the particular user.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <9qBY4.1017$vx2.469023@news.pacbell.net>,
  tmoran@bix.com wrote:
>   Sorry, I had looked at the license included in the Gnat
> library sources,
> and not at the line in the NO WARRANTY section of the FSF GPL
> that does
> indeed say the parties may contract in writing for a warranty.

Be careful that the "license" in the GNAT library sources is
NOT the full text of the license, it incorporates by reference
the entire GPL, and then modifies this text.


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Ted Dennison]

In article <8gq6r9$un$1@nnrp1.deja.com>,
  Robert Dewar <robert_dewar@my-deja.com> wrote:
> In article <gj_X4.2373$s64.80377@news.flash.net>,
>   "Ken Garlington" <Ken.Garlington@computer.org> wrote:
> > Lockheed Martin, please. The distinction is subtle but
> > important.
>
> Indeed! Sorry for the excessive abbreviation in my post

For some reason, those who came up from the Martin Marietta side of
thengs tend to get touchy about that. Since I came from GE Aerospace
portion of Martin, I never really cared. Silly as it sounds, inside
the company they still keep track of who came from where. So
to an insider "Lockheed" is short for "Lockheed heritage", and means "an
LMC organization that used to be part of Lockheed". But to the rest of
the world, its about as silly as GNU/Linux. :-)

As for getting a bean-counter from LMC to understand anything that isn't
along their well-trod path, I sympathise completely. Trying to introduce
thought into a process that runs on the orderly flow of forms from one
desk to another is an uphill battle.

Plese don't confuse anyone here with a LMC paper-pusher though. Perhaps
Tom hasn't fully grokked the GPL, but that doesn't put him in their
league!

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Ted Dennison]

In article <anhY4.1121$M72.360470@news.pacbell.net>,
  tmoran@bix.com wrote:

> If Jones writes a program, can he offer it under different
> licensing arrangement to different people, one of those
> arrangements being the GPL?

If "Jones" is the sole holder of the copyright on said program, yes. To
avoid this potential route of imprisoning software for the Linux kernel,
Linus makes sure that everyone who makes a significant contribution to a
kernel source gets a piece of the copyright. That way if someone wanted
the Linux kernel distributed under a different license, they'd have to
get the permission (iow: pay off) every copyright holder for every
kernel source file.

I took the same approach with OpenToken, for the same reason.


> >Certainly I understand that you (or rather the company you
> >work for) have decided to use restrictive licenses, and that
>   Calling other people's licenses "restrictive" and your own
> "absolutely free", is not, IMHO, a fair statement of the case.

In principle, perhaps not. In this particular case, yes it is (unless
RR's license has changed quite a bit from the last time I looked at it).

I highgly suggest you go up to www.gnu.org and look through the relevent
documents.

The definition of "Free Software" is at
http://www.gnu.org/philosophy/free-sw.html . It would be hard to argue
that RR's software fits this defintion better than ACT's. An excerpt:

-----
   ``Free software'' refers to the users' freedom to run, copy,
distribute, study, change and improve the software. More precisely,
it refers to four kinds of freedom, for the users of the software:

     The freedom to run the program, for any purpose (freedom 0).
     The freedom to study how the program works, and adapt it to your
needs (freedom 1). Access to the source code is a
     precondition for this.
     The freedom to redistribute copies so you can help your neighbor
(freedom 2).
     The freedom to improve the program, and release your improvements
to the public, so that the whole community
     benefits. (freedom 3). Access to the source code is a precondition
for this.
---

Also, you really should read the GPL. Its much shorter and eaiser to
follow than this thread. :-) Any good software engineer should be
capable of fully understanding it if they read it as they would a new
API; with an eye towards what it allows you to do and doesn't allow you
to do. The URL is:
http://www.gnu.org/copyleft/gpl.html

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <8h0jgj$amb$1@nnrp1.deja.com>,
  Ted Dennison <dennison@telepath.com> wrote:
> In article <anhY4.1121$M72.360470@news.pacbell.net>,
>   tmoran@bix.com wrote:
>
> > If Jones writes a program, can he offer it under different
> > licensing arrangement to different people, one of those
> > arrangements being the GPL?
>
> If "Jones" is the sole holder of the copyright on said
> program, yes.

That's entirely correct. It is one special case of the fact
that if you as copyright holder issue a license to someone,
it does not in anyway restrict what YOU can do.

> To avoid this potential route of imprisoning software for the
> Linux kernel, Linus makes sure that everyone who makes a
> significant contribution to a
> kernel source gets a piece of the copyright.

Note that another interesting way of achieving this is by
contract. In the case where a copyright assignment is made
to the FSF, the instrument of assignment includes a requirement
that the FSF make the software available under the GPL for
ever, so that they do NOT have the rights the original
copyright holder would have. This is relevant to CLA, since
the copyright for the basic GNAT technology was assigned
from New York University to FSF.

> That way if someone wanted the Linux kernel distributed under
> a different license, they'd have to get the permission (iow:
> pay off) every copyright holder for every
> kernel source file.

Or argue that a particular copyright holder does not in fact
have a right to protectable code. After all Altai contains
quite a long list of exclusions from what is protectable. Yes
I know that's only third circuit, but in practice it is widely
recognized.

Also there are drawbacks to the multiple copyright holders
situation. Has each of these copyright holders specifically
licensed his work under the GPL. Note that merely putting
a notice in a source file may or may not constitute granting
such a license.

Consider the following situation. Party A holds the copyright
to some code X and releases it under the GPL.

Party B makes a useful modification. He holds the copyright
on this modification. He posts on the net

"I have a really nice modification here to the code X"

That does NOT mean that anyone is free to use this modification
unless party B very explicitly releases the modification under
the GPL.

What is the status of the modified software in B's posession?
Simple, it is a deriviative work, certainly permitted under
the GPL, in which there are two interests A's and B's. The
rub is that unless B agrees to the GPL licensing, the
deriviative work simply cannot be distributed.

One can even imagine B hiring himself out to make the same
change to individual copies of program A, retaining full
rights to this change.

So if you go the multiple copyright holder route, be absolutely
sure that you have proper legal instruments signed by every
contributor. The requirement here is no different than if
a copyright assignment was required.

> I took the same approach with OpenToken, for the same reason.

Have you been careful to get proper legal documents signed
by all contributors. If not, it may well be the case that you
are distributing something that cannot in fact be legally
redistributed.

Legal stuff gets complicated fast :-)

Robert Dewar


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Ted Dennison]

In article <8h0mt3$d5j$1@nnrp1.deja.com>,
  Robert Dewar <robert_dewar@my-deja.com> wrote:
> Consider the following situation. Party A holds the copyright
> to some code X and releases it under the GPL.
>
> Party B makes a useful modification. He holds the copyright
> on this modification. He posts on the net
>
> "I have a really nice modification here to the code X"
>
> That does NOT mean that anyone is free to use this modification
> unless party B very explicitly releases the modification under
> the GPL.

Well, I don't go and grab random code off the net to put in the
baseline. Everything in OpenToken was either written by me or was
submitted to me via (acrhived) email for the express purpose of being
incorporated into the GMGPL baseline. But it certainly couldn't hurt to
formailze this process a bit more. I'd hate for there to be a
misunderstanding.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <8h0qd0$g1q$1@nnrp1.deja.com>,
> Well, I don't go and grab random code off the net to put in
the
> baseline. Everything in OpenToken was either written by me or
was
> submitted to me via (acrhived) email for the express purpose
of being
> incorporated into the GMGPL baseline. But it certainly
couldn't hurt to
> formailze this process a bit more. I'd hate for there to be a
> misunderstanding.

Incorporating something into a GPL'ed program does not make
the incorporation GPL'ed. That is a common misconception, but
it is just that a misconception.

The issue is one of
redistribution. A derived work that contains GPL'ed code
can (under the rights obtained from the GPL) only be
distributed if all the other code has no more restrictive
licensing requirements.

So unless you have a very definite statement, preferably in
writing signed, an email message is unlikely to be taken as
a formal contract, or at least authenticity questions can
easily be raised in this context, you may be in trouble.
The statement must CLEARLY give the author's permission
for incorporating the code under whatever licence you are
using.

There is a lot of sloppiness in these kind of procedures with
code wandering around, but that sloppiness can come back to
haunt you in court later on.

Suppose someone offers you ten million dollars for a copy
of OPENTOKEN with all these changes. You might well find
that some of the other contributors object to getting $0.

Yes, that is far fetched, and of course in the case of
OPENTOKEN, there is unlikely to be a scenario in which
the legal issues are ever explored, but that does not mean
they do not exist. And Linux is after all quite another
kettle of fish when it comes to $$$$, I can easily imagine
some of the copyright holders for Linux getting persnickety
if things have not been done exactly right :-)



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Ted Dennison]

In article <8h0rne$h92$1@nnrp1.deja.com>,
  Robert Dewar <robert_dewar@my-deja.com> wrote:
> In article <8h0qd0$g1q$1@nnrp1.deja.com>,
> So unless you have a very definite statement, preferably in
> writing signed, an email message is unlikely to be taken as
> a formal contract, or at least authenticity questions can
> easily be raised in this context, you may be in trouble.
> The statement must CLEARLY give the author's permission
> for incorporating the code under whatever licence you are
> using.

I can see where you are comming from here. But I'm afraid that requiring
a formal written and notarized contract for every little submission
would have a very chilling effect on such a small project. I know for a
fact that I don't really have the money to develop such a contract. Note
that my main co-contributor so far (Christ Grein) is based in Germany.
How do I know what the legal ramifications of that are? Perhaps there's
some law there that doesn't allow people to sign away copyright intrests
in the way my homespun contract requires. It would take a true
international IP laywer to sort that out, and I just can't afford to do
that.

> There is a lot of sloppiness in these kind of procedures with
> code wandering around, but that sloppiness can come back to
> haunt you in court later on.

True. But when you are talking total hobbyist/volunteer efforts, I'm
afraid a certian amount of such legal "slop" is going to be unavoidable.
If I do something email-based that makes it clear what is going to
happen to a submission, and that the submitting person agrees to that,
that ought to get me about 90% of the way there, without turning off
prospective submittors too much. It should at least prevent
misunderstandings, like the one in your scenario. It would be absolutely
no defense against a company with deep pockets and a commitment to
breaking the license. But at my present financial level, no amount of
formal written documents would save me in that sceario either.

--
T.E.D.

http://www.telepath.com/~dennison/Ted/TED.html


Sent via Deja.com http://www.deja.com/
Before you buy.

About AdaOS

[Didier Utheza]

Is the AdaOS site stillonline. I try http://www.AdaOS.org/ and I get a
forbidden access message. Is the access restricted or the site offline?
		Didier Utheza

Re: "proprietary", was Re: ada on linux

[bill]

Why deal with all this mess of GPL, when one can download 
freeBSD which is all free?

Bill

 

Re: "proprietary", was Re: ada on linux

[Ken Garlington]

"Ted Dennison" <dennison@telepath.com> wrote in message
news:8h0hmg$95u$1@nnrp1.deja.com...
> In article <8gq6r9$un$1@nnrp1.deja.com>,
>   Robert Dewar <robert_dewar@my-deja.com> wrote:
> > In article <gj_X4.2373$s64.80377@news.flash.net>,
> >   "Ken Garlington" <Ken.Garlington@computer.org> wrote:
> > > Lockheed Martin, please. The distinction is subtle but
> > > important.
> >
> > Indeed! Sorry for the excessive abbreviation in my post
>
> For some reason, those who came up from the Martin Marietta side of
> thengs tend to get touchy about that.

Actually, I'm from the Lockheed (nee General Dynamics) side. I just know
that the corporate position is to protect the trademark by not allowing the
name to be shortened. Aviation Week ran a (somewhat snide) article some time
back about the postcards that LM legal sends to publications that shorten
the name...

Re: "proprietary", was Re: ada on linux

[Florian Weimer]

bill@new writes:

> Why deal with all this mess of GPL, when one can download 
> freeBSD which is all free?

Does FreeBSD have an Ada compiler which is "all free"?  What about a C
compiler?

Re: About AdaOS

[Didier Utheza]

Thanks, I did. Very happy that the project is still on.
			Didier Utheza.

Re: "proprietary", was Re: ada on linux

[Geoff Bull]

This has got be the most pointless thread ever on CLA!

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <8h1747$i4a@drn.newsguy.com>,
  bill@new wrote:
>
> Why deal with all this mess of GPL, when one can download
> freeBSD which is all free?

Assuming this is not meant as a troll, which might well be
the case, then the answer is straightforward. Authors choose
to use the GPL to ensure that derived works remain all free.

The trouble with freeBSD, is there is nothing to stop R. Dewar
or anyone else making nonfreeveryexpensiveproprietaryBSD with
some small changes.

We want to ensure that all future versions of GNAT remain free!



There is no "mess" with GPL, though I agree some of the
obfuscation in this thread could be misinterpreted in this
manner. The GPL is really quite clear, and quite effective.

Usually I only see complaints or worries about the GPL from
those associated with manufacture of proprietary software.
Those people would of course prefer a BSD like license to
the GPL license, because then they can use the software in
their proprietary products.

Let's take a specific example. At one point Tartan suggested
the possibility of joining the GNAT front end with their
proprietary backend to make a proprietary result. We told
them they couldn't and that was that! On the other hand, we
know of important cases where major manufacturers are open
sourcing important code generators so that they can properly
use GPL'ed front ends.

Anyway, the important thing to remember is that the author
of software decides on the licensing, and the only reasonable
response anyone else can have to a license is to either
decide that you like the software and license, and acquire
it, or that you don't and decline!

Robert Dewar
Ada Core Technologies



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <39362445.3C630862@research.canon.com.au>,
  Geoff Bull <geoff@research.canon.com.au> wrote:
>
> This has got be the most pointless thread ever on CLA!


I don't agree, since GNAT is an important artifact on the
Ada scene, understanding what its status and licensing are
is of interest to quite a lot of people.

I do agree that some of the posts have been less than helpful,
but the thread contains useful information for many (I even
got a nice private thankyou note from someone for one of my
explanatory posts :-)



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

>Usually I only see complaints or worries about the GPL from
>those associated with manufacture of proprietary software.
  One would hardly expect people using the GPL to be frequent complainants.

>Let's take a specific example. At one point Tartan suggested
>the possibility of joining the GNAT front end with their
>proprietary backend to make a proprietary result. We told
>them they couldn't and that was that!
  So there could have been an Ada compiler with the advantages of both
the Gnat front-end, and Tartan's back-end, but you said no.  Was that
because you didn't have the rights to do it, or because you didn't want
Tartan to offer such a compiler?

>... those associated with manufacture of proprietary software.
>Those people would of course prefer a BSD like license to
>the GPL license, because then they can use the software in
>their proprietary products.
  Do you know as fact that all people who object to the GPL would
prefer the BSD like license?  Do you know as fact that's because then
they would be able to use the software in their proprietary products?
Or are you just guessing that anyone who disagrees with the OSF hype
does so solely because it prevents them from thievery?

Re: "proprietary", was Re: ada on linux

[Dale Stanbrough]

tmoran@bix.com wrote:

>   So there could have been an Ada compiler with the advantages of both
> the Gnat front-end, and Tartan's back-end, but you said no.

The "saying no" would have been more along the lines of 

   "that would violate the copyrights for this software which is
    owned by the FSF"

than

   "we are not going to let you do it"

Dale

Re: "proprietary", was Re: ada on linux

[Jeff Creem]

> >... those associated with manufacture of proprietary software.
> >Those people would of course prefer a BSD like license to
> >the GPL license, because then they can use the software in
> >their proprietary products.
>   Do you know as fact that all people who object to the GPL would
> prefer the BSD like license?  Do you know as fact that's because then
> they would be able to use the software in their proprietary products?
> Or are you just guessing that anyone who disagrees with the OSF hype
> does so solely because it prevents them from thievery?

I know in cases I have been concerned about, I have had to stay away
from libraries that are pure GPL (not talking about GNAT here) for my
projects
since there are projects where portions of the source code are classified
(by
the government).  We certainly give all of the source the the government and
they
in turn are free to give it to whomever they determine has the clearence and
need to
know. But, the clause in the GPL that says all third parties must have all
the licence rights
has always meant to me (and I am not a lawyer and I could be very wong here)
that if
some third party approached the company and said that they wanted a copy
of program X from us that we would have to give it to them (but we could
charge a fee)..

Perhaps this is just another misconception..A while back a sent e-mail to
FSF asking specific
questions about  the GPL in this and other areas...I got some answers but
the rest of the
reply said these are hard questions and I got no answers...

We do of course have our own legal department that could look at these (and
has) but
I am often concerned that they give me answers for what the believe they can
get
away with based on the licenses and not what the original intent of the
license is. (again
another personal belief that is probably wong).


Jeff

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <sv1_4.104$%a3.95362@news.pacbell.net>,
  tmoran@bix.com wrote:
> Do you know as fact that all people who object to the GPL
> would prefer the BSD like license?  Do you know as fact that's
> because then they would be able to use the software in their
> proprietary products?

a) Yes, this corresponds to my experience in these discussions.

b) Yes, it makes sense, since that's really the only major
issue in practice.

> Or are you just guessing that anyone who disagrees with the
> OSF hype

OSF? [what does OSF have to do this with this discussion??)
hype? (what hype are you talking about here??)
what on earth are you talking about here.
This discussion is just about what various licenses permit.

> does so solely because it prevents them from thievery?

Who said anything about theivery? That's very very odd rhetoric.
We are just talking about licenses here and what they permit.
I really have no idea what you are talking about here either!


Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <8haraq$k3j$1@pyrite.mv.net>,
  "Jeff Creem" <jcreem@atetola.mv.com> wrote:

> I know in cases I have been concerned about, I have had to
stay away
> from libraries that are pure GPL (not talking about GNAT here)
for my
> projects
> since there are projects where portions of the source code are
classified
> (by
> the government).

Absolutely! An author who decides to use the unmodified GPL
is most definitely making the decision that you CANNOT use
the software in this manner without negotiating a separate
license. And that's the author's right under current
copyright law.

> that if
> some third party approached the company and said that they
wanted a copy
> of program X from us that we would have to give it to them
(but we could charge a fee)..

This is complete nonsense. But it is a VERY common
misconception. The GPL never forces you to distribute
a program. If ACT suddenly decided to make no more public
versions available, that would be perfectly in accord
with the GPL requirements. As I have said many times before
the fact that ACT makes public versions available is not
required by the GPL, it is simply something that ACT decides
to do for the general good of the Ada community.

> Perhaps this is just another misconception..A while back a
> sent e-mail to FSF asking specific
> questions about  the GPL in this and other areas...I got some
> answers but
> the rest of the
> reply said these are hard questions and I got no answers...

There is absolutely NO reason to expect the FSF to provide
you with free legal services. The fact that they did answer
some of your questions is entirely up to them.

> We do of course have our own legal department that could look
at these (and
> has) but
> I am often concerned that they give me answers for what the
believe they can
> get
> away with based on the licenses and not what the original
intent of the
> license is. (again
> another personal belief that is probably wong).

You always have to ask your own lawyers what you can and cannot
do under a license if it is not clear to you!



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

>Those people would of course prefer a BSD like license to
>the GPL license, because then they can use the software in
>their proprietary products.

  If the software is GPLed, it can be used in "Those people"'s proprietary
product simply by agreement of the copyright holder, since he can release
his software under different licenses to different people.  So the only
difference is that with BSD "Those people" do not need permission of the
copyright holder.  While legal, adding significant value to one's
proprietary product by copying/including something BSDed, without its
author's permission, is usually considered unethical.  So saying that all
people who object to the GPL would prefer the BSD like license because
then they would be able to use the software, without it's author's
permission, in their proprietary products, is saying that all people who
object to the GPL do so because they wish to engage in that kind of
unethical behavior.  That is false.  There are many people who are not in
fact interested in stealing other people's software, but still believe the
GPL and its associated rhetoric very often have very negative consequences
for both developers and users.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <CQH_4.1477$C02.345830@news.pacbell.net>,
  tmoran@bix.com wrote:
>  While legal, adding significant value to one's
> proprietary product by copying/including something BSDed,
> without its
> author's permission, is usually considered unethical.

This is complete nonsense. If the author releases something
under the BSD, they are giving you *EXPLICIT* permission to
"add significant value ...", and there is absolutely NOTHING
unethical, inappropriate, or even unexpected about such
action. The whole idea of the BSD license is to give such
permission, and there are definitely authors who feel this
is appropriate and is what they want, and that is why they
use the license.

If an author wants to be asked for permission for such actions,
then they should not use the BSD license, it is as simple as
that.

> So saying that all
> people who object to the GPL would prefer the BSD like license
because
> then they would be able to use the software, without it's
author's
> permission, in their proprietary products, is saying that all
people who
> object to the GPL do so because they wish to engage in that
kind of
> unethical behavior.

Except that it is not unethical except in your confused
thinking.

> That is false.  There are many people who are not in
> fact interested in stealing other people's software,

To think that using BSD software in this way is stealing is
completely absurd, and actually really rather inappropriate.
You are accusing people of a crime, where there is no crime,
just misunderstanding on your part.

Really Tom, if you are going to make such pronouncements,
you must make more effort to understand how copyright
and licensing works!

> but still believe the
> GPL and its associated rhetoric very often have very negative
> consequences for both developers and users.

What "very negative consequences"? Developers use this license
if that's how they want to license their software, users acquire
the software if they find the license conditions acceptable.
This is the same situation as for any other licensed software.

The only negative consequence is that a particular user or
developer finds that they cannot make use of the software under
some situation because the license is unsuitable. Fine, go find
another product, that's what competition is about!





Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[Robert Dewar]

In article <393B554F.43C369C4@research.canon.com.au>,
  Geoff Bull <geoff@research.canon.com.au> wrote:
> tmoran@bix.com wrote:
> > While legal, adding significant value to one's
> > proprietary product by copying/including something BSDed,
without its
> > author's permission, is usually considered unethical.
>
> On what do you base this assertion ???
> There is absolutely nothing to suggest this in the BSD
> license.

Exactly, and indeed one of the main reasons that some authors
prefer to use the BSD is precisely that they want everyone to
be free to use the software without any restrictions. That's
perfectly legitimate, and may indeed make a lot of sense in
a particular situation. For example, when you buy the Palm
Developer's Kit, you get a completely unfettered license to
the source of the Palm utilities. Palm *wants* you to make
improved software to sell more Palms!

The choice of license really depends on the situation.

For the GNAT runtime, we think it is a good thing if developers
can freely use this code, and that is why we follow the pattern
established by the FSF in similar situations to ensure that
runtime code can be freely used.

For the GNAT compiler, we think it is a good thing for the Ada
community if we all have a guarantee that all future versions
of GNAT will remain in open source/free software status, and
that's why we choose to use the GPL for this part of the
technology.

Robert Dewar
Ada Core Technologies



Sent via Deja.com http://www.deja.com/
Before you buy.

Re: "proprietary", was Re: ada on linux

[tmoran]

>The trouble with freeBSD, is there is nothing to stop R. Dewar
> or anyone else making nonfreeveryexpensiveproprietaryBSD with
> some small changes.
  That you find this perfectly acceptable is not the issue - your
belief that this is the only reason people object to the GPL,
is incorrect.

Re: "proprietary", was Re: ada on linux

[Geoff Bull]

Robert Dewar wrote:
> 
> OSF?

FSF?

Re: "proprietary", was Re: ada on linux

[Geoff Bull]

tmoran@bix.com wrote:
> While legal, adding significant value to one's
> proprietary product by copying/including something BSDed, without its
> author's permission, is usually considered unethical.

On what do you base this assertion ???
There is absolutely nothing to suggest this in the BSD license.

Re: "proprietary", was Re: ada on linux

[Jeff Creem]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

First let me start off by saying that I only jumped into this because
I find the whole area of software licensing interesting (apparently I
am
getting old and bizzare or something).


So in my prior postings (and Roberts response) we have :

> In article <8haraq$k3j$1@pyrite.mv.net>,
>   "Jeff Creem" <jcreem@atetola.mv.com> wrote:
>
> > I know in cases I have been concerned about, I have had to
> stay away
> > from libraries that are pure GPL (not talking about GNAT here)
> for my
> > projects
> > since there are projects where portions of the source code are
> classified
> > (by
> > the government).
>
> Absolutely! An author who decides to use the unmodified GPL
> is most definitely making the decision that you CANNOT use
> the software in this manner without negotiating a separate
> license. And that's the author's right under current
> copyright law.


Ok..Seems fair..And it is exactly what I always assumed one of
the ramifications of the unmodified GPL was (and again I understand
that an author is well within their rights to set these limits)..

But then curiously to one of my misconception ramblings we have this:

>
> > that if
> > some third party approached the company and said that they
> wanted a copy
> > of program X from us that we would have to give it to them
> (but we could charge a fee)..
>
> This is complete nonsense. But it is a VERY common
> misconception. The GPL never forces you to distribute
> a program. If ACT suddenly decided to make no more public
> versions available, that would be perfectly in accord
> with the GPL requirements. As I have said many times before
> the fact that ACT makes public versions available is not
> required by the GPL, it is simply something that ACT decides
> to do for the general good of the Ada community.


Hmm..But if that were the case then my first problem would be
no problem at all since I would have given source and binaries
to a customer who owns the secrets and I dont have to
give it to anyone else and neither do they..

My common misconception problably comes from the following
text from the GPL
"
3. You may copy and distribute the Program (or a work based on it,
under Section 2) in object code or executable form under the terms of
Sections 1 and 2 above provided that you also do one of the
following:

a) Accompany it with the complete corresponding machine-readable
source code, which must be distributed under the terms of Sections 1
and 2 above on a medium customarily used for software interchange;
or,

b) Accompany it with a written offer, valid for at least three years,
to give any third party, for a charge no more than your cost of
physically performing source distribution, a complete
machine-readable copy of the corresponding source code, to be
distributed under the terms of Sections 1 and 2 above on a medium
customarily used for software interchange; or,

c) Accompany it with the information you received as to the offer to
distribute corresponding source code. (This alternative is allowed
only for noncommercial distribution and only if you received the
program in object code or executable form with such an offer, in
accord with Subsection b above.)
"

So..If I ever once send someone a binary without source code sure
seems to me that for the next
three years I must give source to any third party for a nominal
fee...




>
> > Perhaps this is just another misconception..A while back a
> > sent e-mail to FSF asking specific
> > questions about  the GPL in this and other areas...I got some
> > answers but
> > the rest of the
> > reply said these are hard questions and I got no answers...
>
> There is absolutely NO reason to expect the FSF to provide
> you with free legal services. The fact that they did answer
> some of your questions is entirely up to them.

I agree..And my comment was not meant to say disparage the FSF
in any way however the FSF wants to promote "free" software. There
are theses licenses GPL and LGPL and based on discussions I have
seen I would bet that more than half of the people who put source out
there under the GPL or LGPL have no idea what terms they are
actually releasing the source under...Granted this is their problem..

Note that when I sent e-mail to them asking for help I did offer to
send them
a contribution (of course it would not have been enough to pay for a
lawyer since
I was looking into this for personal reasons).


>
> > We do of course have our own legal department that could look
> at these (and
> > has) but
> > I am often concerned that they give me answers for what the
> believe they can
> > get
> > away with based on the licenses and not what the original
> intent of the
> > license is. (again
> > another personal belief that is probably wong).
>
> You always have to ask your own lawyers what you can and cannot
> do under a license if it is not clear to you!
>


Again my point here is that lawyers will tell me what the license
says I can
do which is important from a legal point of view (and for example, I
neither could nor
would use something if my legal department said I could not)...But
there is another issue here which
is intent...It may be that the specific wording of a license allows
something but it could
be that the author never intended for it to allow it...For lawsuits I
am interested in the former..
For sleeping at night and looking in the mirror I am interested in
the latter.


Jeff

-----BEGIN PGP SIGNATURE-----
Version: PGPfreeware 6.5.3 for non-commercial use <http://www.pgp.com>

iQA/AwUBOTxJiS7VRsN7h5LhEQKJPACgzjDM+RVlkdeFe3h2HPHSwuDhiQoAoJTn
4anPf5Nkg+eVkdhngS8ji/1V
=uSnz
-----END PGP SIGNATURE-----

GPL distribution rules (was: "proprietary")

[Larry Kilgallen]

In article <8hhhej$mp2$1@pyrite.mv.net>, "Jeff Creem" <jcreem@atetola.mv.com> writes:

> 3. You may copy and distribute the Program (or a work based on it,
> under Section 2) in object code or executable form under the terms of
> Sections 1 and 2 above provided that you also do one of the
> following:
> 
> a) Accompany it with the complete corresponding machine-readable
> source code, which must be distributed under the terms of Sections 1
> and 2 above on a medium customarily used for software interchange;

I don't see any requirement around compilers.  Doesn't that leave
the "loophole" (from the FSF viewpoint) that someone could modify
some (non-compiler) software obtained under this license to use
special language features not generally supported in the source
language and release the modified source but not a suitable
compiler ?