new extension generator for C++

[Carl Banks]

While I think most of the disagreement in this long thread results
from different beliefs in what "freedom" means, I wanted to add, that
most of the responses that argue that the MIT license permits the user
more freedom than the GPL, suffer from the broken window fallacy.
This fallacy results from the short-sided-ness of the user base, as it
is only considering the first generation of derivative works.

I agree, that under an MIT license, the first generation of derivative
works have more freedom.  But any extra freedom gained here comes at
the direct expense of all future generations of derivative software.

Under a GPL license, it is true that the first generation will have
less freedom to distribute their software as they would like.  But it
also ensures that all subsequent generations of derivative works have
the freedom to access all previous derivative works.

I believe the you have the fallacy backwards.

The thing you GPL fanbois refuse to understand or accept is that, in
the real world, a person or company who doesn't want to open source
their "derivative work" will only rarely be forced to by the GPL.
They'll work around it instead, vast majority of the time. They
could:

1. Derive their work from a project with a license that grants the
user more freedom
2. Reimplment the functionality seperately (*cough* PySide)
3. Ignore the license

And no, a small number of anecdotal counterexamples is not any strong
evidence to the contrary.

On the other hand, those who intended to release their work as open
source are going to do it even if the license is permissive. The way
some of you GPL fanbois talk you'd think the MIT license prohibitied
open source derivatives.

So, you see, the rights given to users of first generation works (as
you say) are far more important than requiring people who create
"derivatives" to also offer those rights. Those who intend to do that
will anyways, those who don't intend to will find a way not to.


Carl Banks

PS The word "derivative" is quoted throught because the GPL's
definition of "derivative" is ludicrous and not in accordance any
common defintion of the word.

 

[Aahz]

All my position has ever been is this:

A copyrighted work denies recipients virtually all rights to do stuff
with that work, such as modify and redistribute it. Copyleft licences
grant some privileges and uphold some obligations in order to ensure
that these privileges are universally maintained in all forms and
extensions of the work. Permissive licences grant more privileges to
immediate recipients but do not uphold as many obligations.

You rejected the suggestion that people using permissive licences
afford users fewer privileges than those using copyleft licences, yet
on balance when considering all forms and extensions of the work, they
do. And the only such privilege that the copyleft licences withholds
from recipients is that of withholding any other privilege from others
who receive the work.

IMO this only makes sense if one agrees that people should not be allowed
to sell software for money. Absent that agreement, your argument about
freedom seems rather limited.

 

[Patrick Maupin]

While I think most of the disagreement in this long thread results
from different beliefs in what "freedom" means, I wanted to add, that
most of the responses that argue that the MIT license permits the user
more freedom than the GPL, suffer from the broken window fallacy.
This fallacy results from the short-sided-ness of the user base, as it
is only considering the first generation of derivative works.

The broken window fallacy is about labor that could have been spent
elsewhere if someone else had done something differently. The only
time that comes into play in my programming life is when I have to
recode something that is nominally available under the GPL, so I'm not
sure this is really making the point you think it is.

Regards,
Pat

 

[Steven D'Aprano]

The thing you GPL fanbois refuse to understand or accept is that, in the
real world, a person or company who doesn't want to open source their
"derivative work" will only rarely be forced to by the GPL. They'll work
around it instead, vast majority of the time. They could:

1. Derive their work from a project with a license that grants the user
more freedom
2. Reimplment the functionality seperately (*cough* PySide)

Yes. So what? In what possible way is this an argument against the GPL?

If I offer to mow your lawn for $20, and you refuse, I don't have to
force my services onto you. You can mow your lawn yourself, or find
somebody who will do it for $10, or find some kind generous soul who will
do it for free under an MIT licence.

If you don't want the obligations of the GPL, nobody is going to force
you to distribute or derive from the GPLed software. If you're not
willing to meet my conditions to use my software, then I don't want you
using my software. Go write your own, or find somebody else who will do
it for a fee, or for free. What you see as a disadvantage of the GPL is a
feature, not a bug.

You seem to be under the impression that the only public good developers
should care about is maximising the number of people who create
derivative works from your project. That's obviously not so in the real
world, where vast numbers of developers release software under closed-
source licences, or don't release it to the public at all, and it's
certainly not the case for people who use the GPL. The FSF even has a FAQ
about that exact point.

3. Ignore the license

This is a crime under copyright law, and there have been many instances
of companies who thought they could ignore the GPL learning different.

 

[Steven D'Aprano]

The broken window fallacy is about labor that could have been spent
elsewhere if someone else had done something differently. The only time
that comes into play in my programming life is when I have to recode
something that is nominally available under the GPL, so I'm not sure
this is really making the point you think it is.

You've never had to recode something because it was nominally available
under a proprietary licence that you (or your client) was unwilling to
use? Lucky you!

The GPL ensures that once software has entered the commons (and therefore
available for all), it can never be removed from the commons. The MIT
licence does not. Now, you might argue that in practice once software is
released under an MIT licence, it is unlikely to ever disappear from the
commons. Well, perhaps, but if so, that's despite and not because of the
licence.

In practice, I believe most MIT-licenced code never even makes it into
the commons in the first place. I'm willing to predict that the majority
of code you've written for paying customers (as opposed to specifically
for open source projects) has disappeared into their code base, never to
be seen by anyone outside of the company. Am I right?

 

[Carl Banks]

Yes. So what? In what possible way is this an argument against the GPL?

[snip a bunch of crap I don't care about]


It's not. It's an argument that the GPL doesn't do much good.

Arguments against the GPL are found elsewhere in this thread, I don't
need to repeat them here.


Carl Banks

 

[Ethan Furman]

You've never had to recode something because it was nominally available
under a proprietary licence that you (or your client) was unwilling to
use? Lucky you!

Steven, did you actually read what he wrote? If you did, why would you
say something so stupid?

~Ethan~

 

[Paul Boddie]

[...]

Just to deal with your Ubuntu "high horse" situation first, you should
take a look at the following for what people regard to be the best
practices around GPL-licensed software distribution:

http://www.softwarefreedom.org/resources/2008/compliance-guide.html

If you still think Ubuntu are violating the GPL or encouraging others
to do so, feel free to contact their lawyers who I'm sure will be very
interested to hear from you.

When the leader of your religion bandies terms like "freedom" and
"evil" about, what do you expect?  Seriously?

I thought you were "done". I guess you are: again, we have the usual
courting of public outrage by labelling stuff you don't like as
"religion" - presumably not the "right one", either - when it is no
such thing.

[...]

My primary agenda is to explain that RMS does, in fact, have an
agenda, and the GPL was designed as a tool in furtherance of that
agenda, and that while the agenda does have some arguably noble goals,
before using the GPL people should understand its consequences both
for good and bad, and make their own determination about whether it's
the right license for their project.

Reading through your "translations" of what are effectively honest
summaries, one gets the impression that you have quite a chip on your
shoulder about the FSF and RMS. Referring to the GPL as a "commercial"
licence and stating that it (as opposed to any other licence or even
the word "copyright" followed by a name) is a threat to sue people,
presumably appealing to the libertarian crowd with a judicious mention
of "government" just to fan the flames of supposed injustice, really
does triangulate where you are coming from. So, yes, we're now rather
more aware of what your agenda is, I think.

And I don't think it improves any argument you may have by projecting
notions of "morality" or "immorality" onto what I have written,
especially when I have deliberately chosen to use other terms which
avoid involving such notions, or by equating the copyleft licences
with criminal enterprises ("pyramid scheme"), or by suggesting that I
endorse criminal endeavours. But if that's what you have left to say
at this point, then I think you probably are "done".

Paul

 

[Paul Boddie]

IMO this only makes sense if one agrees that people should not be allowed
to sell software for money.  Absent that agreement, your argument about
freedom seems rather limited.

You'll have to explain this to me because I don't quite follow your
assertion. You can sell copyleft-licensed software, although I accept
that you can't set an arbitrarily high price on the sources for
someone who has already acquired a binary distribution.

Paul

 

[Patrick Maupin]

You've never had to recode something because it was nominally available
under a proprietary licence that you (or your client) was unwilling to
use? Lucky you!

Don't be silly. That's why I started writing open source software in
the first place. But if I start writing stuff to put in the commons
with strings removed, why would I bother with a license that just adds
some strings back?

The GPL ensures that once software has entered the commons (and therefore
available for all), it can never be removed from the commons.

No it doesn't. It just insures that if people actually *distribute*
the software to others, they have to distribute the source. In any
case, for software to remain in the commons, it has to be available
where people can get to it. Somebody has to care enough to maintain a
repository, or it has to be good enough for people to distribute.

The MIT licence does not.

The only difference is that somebody has to care enough to maintain a
repository, or it has to be good enough for people to distribute
*along with source*.

Now, you might argue that in practice once software is
released under an MIT licence, it is unlikely to ever disappear from the
commons.

Depends on the software. See above.

Well, perhaps, but if so, that's despite and not because of the
licence.

Same thing for GPLed software. See above.

In practice, I believe most MIT-licenced code never even makes it into
the commons in the first place.

Interesting assertion.

I'm willing to predict that the majority
of code you've written for paying customers (as opposed to specifically
for open source projects) has disappeared into their code base, never to
be seen by anyone outside of the company. Am I right?

That's true, but what on earth does that have to do with the MIT
license?

Regards,
Pat

 

[Paul Boddie]

Yes. So what? In what possible way is this an argument against the GPL?

[...]

It's not.  It's an argument that the GPL doesn't do much good.

Right. So nobody got the benefit from Qt under the GPL or PyQt under
the GPL? Even the PySide developers seem hell-bent on picking over the
work of the PyQt developers for ideas, although they obviously won't
touch the code. Nokia seem to have accrued tremendous benefit from the
existence of PyQt because I rather doubt that anyone would have
bothered rolling a set of mature, usable Python bindings for Qt now
had some not existed already and proved that dynamic languages are
worth supporting.

Arguments against the GPL are found elsewhere in this thread, I don't
need to repeat them here.

Yes, don't bother. They fit in rather well with the comment you made
above.

Paul

 

[Patrick Maupin]

Just to deal with your Ubuntu "high horse" situation first, you should
take a look at the following for what people regard to be the best
practices around GPL-licensed software distribution:

http://www.softwarefreedom.org/resources/2008/compliance-guide.html

Before, you were busy pointing me at the GPL FAQ as authoritative.
When I show where the FAQ says that you should distribute source if
you give somebody a CD, you point me at a document which is obviously
designed for the Ciscos of the world.

If you still think Ubuntu are violating the GPL or encouraging others
to do so, feel free to contact their lawyers who I'm sure will be very
interested to hear from you.

Did I *ever* say that Ubuntu was violating the GPL. No. Do I believe
that the practices of any binary Linux distribution that fits on a
single CD make it easy for the downloader/burner to violate the GPL.
Yes.

I thought you were "done". I guess you are: again, we have the usual
courting of public outrage by labelling stuff you don't like as
"religion" - presumably not the "right one", either - when it is no
such thing.

Well, you conveniently ignore sections of your bible (for example, the
part of the FAQ where it says you should distribute source with
binary) and reach for more obscure scrolls whenever the real world
gets in the way of your fantasy. Even here, you don't bother to quote
what you wrote, which would show I am just responding to your outrage.

Reading through your "translations" of what are effectively honest
summaries.

There are multiple sides to every discussion, and everybody comes to
the table with biases. If you honestly think that you are not biased,
then you are deluding yourself. If you realize that you are biased,
then you will also come to realize that my translations are equally
honest.

one gets the impression that you have quite a chip on your
shoulder about the FSF and RMS.

I can take them or leave them until they and their followers start
spouting damaging nonsense. Many businesses were scared to death of
FOSS for many years, and I lay the blame squarely on RMS's shoulders.
You see only the good he has done; it is tempered by quite a bit of
bad.

Referring to the GPL as a "commercial"
licence and stating that it (as opposed to any other licence or even
the word "copyright" followed by a name) is a threat to sue people,
presumably appealing to the libertarian crowd with a judicious mention
of "government" just to fan the flames of supposed injustice, really
does triangulate where you are coming from. So, yes, we're now rather
more aware of what your agenda is, I think.

I'm not the one who keeps spouting that it "gives" freedoms (or even
privileges) that copyright would have taken away. That's complete
bullshit. As I said, copyright *allows* the author to control various
aspects of his work ("take away freedoms" in GPL-speak), and all
licenses (including the GPL) explicitly state which aspects the author
plans to control. The only way the author can really exert control is
to sue or credibly threaten to sue. I can actually point to multiple
instances of GPL authors suing, and people like you crowing about how
great it is that the GPL stands up in court, but I don't actually
recall any suits about violations of the MIT or Apache licenses. So,
yes, I firmly believe that when somebody slaps a GPL license on their
software (and especially if they sign the copyrights over to the FSF)
they are trying to signal that they are willing to go to court to
protect their rights. This is no different than when Microsoft sues
an infringer, and is not an evil thing, but it is definitely something
to be aware of. The easiest way to not get tangled in that kind of
lawsuit is to just make sure that you never distribute any software
with a commercial-type license on it (including the GPL).

And I don't think it improves any argument you may have by projecting
notions of "morality" or "immorality" onto what I have written,

But you're arguing from a moral standpoint.

especially when I have deliberately chosen to use other terms which
avoid involving such notions

Yes, but you're making exactly the same arguments as others, just
changing the name.

or by equating the copyleft licences
with criminal enterprises ("pyramid scheme"),

Well, that may be a bit OTT. What I really should say is that the GPL
license has a "selfish gene." It tries really hard to propagate
itself, at the expense of the genes of other licenses.

or by suggesting that I
endorse criminal endeavours.

Hmmm, don't recall doing that. If I did, I certainly apologize.

But if that's what you have left to say
at this point, then I think you probably are "done".

Well, I thought I was before, but then the discussion about
downloading an ISO and burning it and giving it to a friend came up.
This may be technically allowable under the license, but nothing you
or anybody else has written has yet proved that to me.

Regards,
Pat

 

[Paul Boddie]

Before, you were busy pointing me at the GPL FAQ as authoritative.

No, the licence is the authority, although the FAQ would probably be
useful to clarify the licence author's intent in a litigation
environment.

[Fast-forward through the usual tirade, this time featuring words like
"bible", "moral", "evil"...]

Well, I thought I was before, but then the discussion about
downloading an ISO and burning it and giving it to a friend came up.
This may be technically allowable under the license, but nothing you
or anybody else has written has yet proved that to me.

Section 3 of GPLv2 (and section 6(d) of GPLv3 reads similarly): "If
distribution of executable or object code is made by offering access
to copy from a designated place, then offering equivalent access to
copy the source code from the same place counts as distribution of the
source code, even though third parties are not compelled to copy the
source along with the object code."

And here's that FAQ entry which clarifies the intent:

http://www.gnu.org/licenses/gpl-faq.html#DistributeWithSourceOnInternet

Like I said, if you really have a problem with Ubuntu shipping CDs and
exposing others to copyright infringement litigation - or even
themselves, since they (and all major distributions) are actively
distributing binaries but not necessarily sources in the very same
download or on the very same disc - then maybe you should take it up
with them.

Paul

 

[Patrick Maupin]

No, the licence is the authority, although the FAQ would probably be
useful to clarify the licence author's intent in a litigation
environment.
Agreed.

Section 3 of GPLv2 (and section 6(d) of GPLv3 reads similarly): "If
distribution of executable or object code is made by offering access
to copy from a designated place, then offering equivalent access to
copy the source code from the same place counts as distribution of the
source code, even though third parties are not compelled to copy the
source along with the object code."

And here's that FAQ entry which clarifies the intent:

http://www.gnu.org/licenses/gpl-faq.html#DistributeWithSourceOnInternet

That entry, along with the written offer, certainly covers Ubuntu when
they distribute a CD.

But if I *download* an ISO, burn it on a CD, and give it away, *I* am
the one distributing the physical copy, not Ubuntu, and I am not going
to put up an FTP server just so my friend can get source from it. And
as section 6 of GPL v3 makes clear, I am not allowed to piggyback on
Ubuntu's source offer. My situation *really is* covered by the FAQ
entry I referred you to:

http://www.gnu.org/licenses/gpl-faq.html#UnchangedJustBinary

Like I said, if you really have a problem with Ubuntu shipping CDs and
exposing others to copyright infringement litigation.

So, deliberately or not, you're trying to change the discussion
again. I *never* discussed Ubuntu shipping a physical CD, and never
intimated that that was a problem. My discussion was *always* about
an individual *downloading* an ISO and *burning* a CD himself, then
*distributing* the CD to someone else.

- or even
themselves, since they (and all major distributions) are actively
distributing binaries but not necessarily sources in the very same
download or on the very same disc - then maybe you should take it up
with them.

Again, I never intimated this. Please read more carefully in the
future before you reply, and then perhaps you will actually make
cogent replies that address my points, and then I won't be so
frustrated that I make snide comments you take offense at, OK? This
has happened on at least 4 separate occasions in this thread, and
sometimes a single misunderstanding goes on for quite a few posts, so
I'm starting to wonder if it's deliberate.

Regards,
Pat

 

[Albert van der Horst]

Come on, 99% of the projects released under GPL did so because they
don't want to learn much about the law; they just need to release it
under a certain license so their users have some legal certainty. Most
programmers are not lawyers and don't care about the law and don't care
about the GPL; if a commercial programmer want to use the GPL-code in an
incompatible licensed program, and he comes up asking, many would just
be happy to say yes.

This is a big reason for me to release everything (see my website,
it is a *lot*) under GPL. If someone wants to use it they can,
if someone wants to use it commercially, they can too, as long
as they pay me a little bit too. Really, I'm reasonable.

Groetjes Albert

 

[Paul Boddie]

Section 3 of GPLv2 (and section 6(d) of GPLv3 reads similarly): "If
distribution of executable or object code is made by offering access
to copy from a designated place, then offering equivalent access to
copy the source code from the same place counts as distribution of the
source code, even though third parties are not compelled to copy the
source along with the object code."

And here's that FAQ entry which clarifies the intent:

[...]

http://www.gnu.org/licenses/gpl-faq.html#UnchangedJustBinary

We're all aware of the obligation to provide source code. You've spent
the last few days complaining about it.

So, deliberately or not, you're trying to change the discussion
again.  I *never* discussed Ubuntu shipping a physical CD, and never
intimated that that was a problem.  My discussion was *always* about
an individual *downloading* an ISO and *burning* a CD himself, then
*distributing* the CD to someone else.

I am not changing the discussion at all. You are describing a
situation where someone gets the binaries but not the sources, but
according to the licence they should get both of those things
(ignoring written offers and the like), and this does apply to Ubuntu
since precisely this act of distribution (to use the older term) is
performed by them. That you then pass on the binaries without the
sources is an equivalent situation, ignoring for the moment that you
do not yourself have the sources either.

So, what are you supposed to do when the recipient "calls" you on the
lack of sources? (And, yes, clearly the FSF anticipates that not
everyone will request the sources because it is written in that very
excerpt I provide above.) If the recipient is strict about exact
compliance, you will have to provide the sources on CD to them. And
this makes sense: if they can only make use of the binaries if
provided on CD (and not, say, on an FTP site because they don't have
an Internet connection, for example), then they will need to receive
the sources in the same manner. Of course, the recipient may only
demand certain sources, not wishing to avail themself of the sources
for all copyleft-licensed packages in the binary distribution.

Now we return to the matter of getting the Ubuntu sources. If you
ordered a CD from Ubuntu via their ShipIt service, it is at this point
that you can demand a CD of corresponding sources. If they cannot
provide one, then obviously it poses a problem for your compliance
(and theirs, and you should see once again why Ubuntu's activities do
matter), but naturally Ubuntu provide parallel binary and source
repositories for all their packages. So, even if they were found not
to be in compliance according to the strictest interpretation of the
licence, it is technically possible for you to acquire the
corresponding sources and make them available to the person who was
given the CD. If you downloaded an ISO file, Ubuntu could (and do)
obviously provide source packages from the same location: their Web
site and various mirrors.

Really, if at this point you think I'm playing games with you, then
you really need to stop taking score and formulate the exact problem
you have with the distribution of Ubuntu-style media, because I'm
starting to think that the only real problem here is the one you have
with people using copyleft-style licences for their works. Since we've
had to hear about that over several days, I don't think that
articulating that particular problem once again really brings anything
more to the discussion.

Paul

 

[Patrick Maupin]

Section 3 of GPLv2 (and section 6(d) of GPLv3 reads similarly): "If
distribution of executable or object code is made by offering access
to copy from a designated place, then offering equivalent access to
copy the source code from the same place counts as distribution of the
source code, even though third parties are not compelled to copy the
source along with the object code."
And here's that FAQ entry which clarifies the intent:
http://www.gnu.org/licenses/gpl-faq.html#DistributeWithSourceOnInternet

[...]

http://www.gnu.org/licenses/gpl-faq.html#UnchangedJustBinary

We're all aware of the obligation to provide source code. You've spent
the last few days complaining about it.

So, deliberately or not, you're trying to change the discussion
again.  I *never* discussed Ubuntu shipping a physical CD, and never
intimated that that was a problem.  My discussion was *always* about
an individual *downloading* an ISO and *burning* a CD himself, then
*distributing* the CD to someone else.

I am not changing the discussion at all. You are describing a
situation where someone gets the binaries but not the sources, but
according to the licence they should get both of those things
(ignoring written offers and the like), and this does apply to Ubuntu
since precisely this act of distribution (to use the older term) is
performed by them. That you then pass on the binaries without the
sources is an equivalent situation, ignoring for the moment that you
do not yourself have the sources either.

If Joe downloads and burns a CD for his friend, he may not have the
sources and may not have any intention of getting them, and probably
didn't provide a "written offer." What you're "ignoring for the
moment" is my whole point, that unlike Ubuntu, Joe is now in violation
of the GPL license, because he provided neither a written offer nor
source on CD, nor his own download site.

So, what are you supposed to do when the recipient "calls" you on the
lack of sources?

There is possibly no "calling". Since no source and no written offer
was delivered, Joe's friend may not know about the issue. Even Joe
himself just saw it was "free software" and didn't read the fine
print, so he may not have a clue how to get the source.

(And, yes, clearly the FSF anticipates that not
everyone will request the sources because it is written in that very
excerpt I provide above.)

Which Joe doesn't know about and didn't adhere to in any case.

If the recipient is strict about exact
compliance, you will have to provide the sources on CD to them.

That could be a year later, and Joe, who doesn't really even know
anything about source, is really going to have a hard time figuring
out exactly which sources went into the CD he downloaded that long
ago.

And
this makes sense: if they can only make use of the binaries if
provided on CD (and not, say, on an FTP site because they don't have
an Internet connection, for example), then they will need to receive
the sources in the same manner.

To an extent it makes sense. That's why I explained that I thought it
would be nice of Ubuntu to put a warning to Joe on their site
explaining the consequences of helping his friend out. Of course,
since the warning would only serve to decrease object downloads, and
since Joe's friend doesn't really want the source anyway, there is no
real point. That doesn't alter the fact that Joe is immediately in
violation of the GPL once he delivers the CD to his friend without the
written offer.

Of course, the recipient may only
demand certain sources, not wishing to avail themself of the sources
for all copyleft-licensed packages in the binary distribution.

[ Stuff about ShipIt snipped because I was never discussing that.]

Really, if at this point you think I'm playing games with you.

I don't know what to think about that. Even after I've explicitly
said multiple times I'm not discussing when Ubuntu ships a CD, you
still felt compelled to include a big paragraph about ShipIt. Is it
to confuse? Or because have OCD? I don't really know.

then
you really need to stop taking score and formulate the exact problem
you have with the distribution of Ubuntu-style media,

I explained it fully multiple times.

because I'm
starting to think that the only real problem here is the one you have
with people using copyleft-style licences for their works.

Well, as I have tried to explain, there are tradeoffs with any
license, including the GPL. With the GPL, you can easily adhere to
the letter of the license by shipping source with object. But
sometimes the source is so huge, people take shortcuts to get around
that. The license allows this, via a requirement to provide source
later, that could actually fall on people who really don't understand
that that is required of them (perhaps not very often in practice, but
certainly in theory).

Since we've
had to hear about that over several days, I don't think that
articulating that particular problem once again really brings anything
more to the discussion.

Probably not, but you've still never addressed how easy it is for a
complete neophyte to go to, e.g. the Ubuntu site, be told do a
download, try it out, get excited, burn CDs for his friends, and then
be in violation of the GPL. As I have said, I don't view this as a
*practical* issue, but it is an example of how those in the GPL
community turn a blind eye to innocent infringement, just like
Microsoft.

After all, the Ubuntu download page says "When the CD is ready, simply
put it in your CD drive, restart your computer and follow the
instructions that will appear on your screen. Don't forget that you
can create more copies and pass the CD to as many people as you like."
without mentioning *anything* about source code.

Regards,
Pat

 

[Paul Boddie]

If Joe downloads and burns a CD for his friend, he may not have the
sources and may not have any intention of getting them, and probably
didn't provide a "written offer."  What you're "ignoring for the
moment" is my whole point, that unlike Ubuntu, Joe is now in violation
of the GPL license, because he provided neither a written offer nor
source on CD, nor his own download site.

Now, wait a moment! Your point is that just by giving the binary CD to
someone, you are now in violation of the licence. What I tried to
explain is that this situation is anticipated - that the FSF
acknowledges that the recipient won't have received the sources at the
same time in all situations - and that the same distributor is
responsible for providing the sources. As long as they don't deny the
recipient access to the sources, by the same means, they are not
violating the licence.

You have a point about recipients not being immediately and obviously
informed of the things they are entitled to, but that is a matter for
the distributing parties to remedy: that is arguably what happens
when, upon loud squealing about matters of "ideology", distributors
decide to de-emphasise the Free Software aspect of their
distributions. Nevertheless, it is my understanding that anyone
attempting to use or install such distributions do get to see a
summary of the licences; only people who pass on the software without
inspecting it (which would involve actually inserting the CD and
booting from it) will be unaware of its contents, and they could only
be held responsible as reasonably as one's Internet service provider
if that party were asked to provide source packages for "that Linux
distro I downloaded last year".

You also have a point about whether people are able to provide sources
at a later date, which might be troublesome if someone gave someone
else a CD with an old version of Ubuntu on it and then were asked to
provide the source packages. Naturally, the FSF have attempted to
address these points in version 3 of the GPL. I would be interested to
hear the opinion of the FSF and distributors on this matter, but I
think it's absurd to accuse the FSF as operating as you allege
Microsoft do, especially as the distributors are the ones who
encourage the sharing of the installation media.

Really, if you think distributions should do a better job at educating
their users and helping them uphold any obligations that may apply to
them, you should talk to them about it. But when I attempt to work
though the issues in a thorough manner in order to thrash out what it
is you really object to - and in practice, the only objections you can
seriously have lie in those two points I mention above (not this
"instant violation" situation, discussed in more detail elsewhere [*])
- and all you can do is suggest that other people are trying to
mislead you, I struggle to feel inclined to indulge you further.

And suggesting that people have behavioural disorders ("Or because
have OCD?") might be a source of amusement to you, or may be a neat
debating trick in certain circles you admire, but rest assured that I
am neither amused nor impressed, nor are others likely to be.

Paul

[*] http://www.mail-archive.com/[email protected]/msg31466.html

 

[Terry Reedy]

The following lines from

http://www.softwarefreedom.org/resources/2008/compliance-guide.html

seem to cover the case of someone who casually redistributes, for free,
Ubuntu or whatever. Such can refer people back to the Ubuntu site. They
should, perhaps, be familiar with the url, but I would expect that the
binary Ubuntu distribution CDs have the appropriate offer and details on
that disk. Someone who casually distributes, for free, a subset should
also be covered. Under 4.1.2 Option (b): The Offer,

"The option to provide an offer for source rather than direct source
distribution is a special benefit to companies equipped to handle a
fulfillment process. GPLv2 § 3(c) and GPLv3 § 6(c) avoid burdening
noncommercial, occasional redistributors with fulfillment request
obligations by allowing them to pass along the offer for source as they
received it.

Note that commercial redistributors cannot avail themselves of the
option (c) exception, and so while your offer for source must be good to
anyone who receives the offer (under v2) or the object code (under v3),
it cannot extinguish the obligations of anyone who commercially
redistributes your product. The license terms apply to anyone who
distributes GPL’d software, "

Terry Jan Reedy

 

[Patrick Maupin]

"The option to provide an offer for source rather than direct source
distribution is a special benefit to companies equipped to handle a
fulfillment process. GPLv2 § 3(c) and GPLv3 § 6(c) avoid burdening
noncommercial, occasional redistributors with fulfillment request
obligations by allowing them to pass along the offer for source as they
received it.

Paul Boddie already pointed out that document. As I explained, that
document was written for the Ciscos of the world. The FAQ, which was
written for you and me states very clearly "The general rule is, if
you distribute binaries, you must distribute the complete
corresponding source code too. The exception for the case where you
received a written offer for source code is quite limited." in answer
to the question "I downloaded just the binary from the net. If I
distribute copies, do I have to get the source and distribute that
too?"

As I have pointed out on at least 3 posts by now, this FAQ
interpretation derives directly from the actual license terms and
appears to reflect the terms correctly. If you actually *read* GPLv3 §
6(c), it *only* applies if you received the object code in accordance
with GPLv3 § 6(b). But if you download an ISO from Ubuntu, that
happens under GPLv3 § 6(d), *not* GPLv3 § 6(b).

However, the distribution to your friend when you give him the CD that
you burned for him is under 6(b), so not only do you not have an
upstream to rely on, you are actually in violation of the license once
you give him the CD without your own written offer! (At one level,
this makes sense -- if the 3 year window for source is to have any
teeth, then you can't give the poor guy a CD 2 years after you
downloaded it and expect Ubuntu to make good on the source 5 years
after you downloaded it.)

Now maybe there is some *other* way (besides the obvious ways I've
mentioned such as fair use and the fact that nobody's going to sue
because of the PR fallout from bothering some grandma for sharing a CD
that was advertised as "free") that this is not an issue, but nobody
on this thread has yet shown any credible evidence that the act of
just handing somebody a freshly burned Ubuntu CD with no written offer
is not a violation of the license.

As I have made clear, I do not view this as a direct practical
problem. But I do view it as a huge problem that the license is so
complex that in a couple of days of conversing about it, several
people have asserted that there is no way my reading of the license is
correct, yet nobody has shown solid evidence that would back up an
alternate reading, and I also view it as the tip of the iceberg as far
as the issue of license compliance goes.

Regards,
Pat