Desktop options (KDE, Gnome, FVWM-CDE, etc.)
Dara Parsavand
Having learned the small amount of Unix I know on a Sun Desktop, I
guess I've always been partial to the idea of a more integrated
desktop. I was never that satisfied with fvwm (2, 95, or plain -
haven't tried the CDE flavor yet), twm, afterstep and the like. I'm
now running KDE which is pretty well integrated, but still dumps cores
a bit too often (it's still alpha, so I'm not complaining - in fact it
is pretty good alpha mostly).
Today, comp.os.linux.announce had a post on the GNOME project which
I'll quote from briefly:
The GNOME Desktop project
(GNU Network Object Model Environment)
http://bananoid.nuclecu.unam.mx/gnome
* Goals
We want to develop a free and complete set of user friendly
applications and desktop tools, similar to CDE and KDE but based
entirely on free software:
- We want the applications to have a common look and feel, and
to share as many visual elements and UI concepts as possible.
---- (some text deleted) ----
* Some common questions regarding the project
Why don't you just use/contribute to KDE?
KDE is a nice project; they have good hackers working on it and
they have done a very good job. Unfortunately, they selected the
non-free Qt toolkit as the foundation for the project, which poses
legal problems for those desiring to redistribute the software.
<end quote>
I'd be interested in hearing from Linux (free BSD, etc.) advocates on
what they think the future of the free Desktop is. Do you think Qt is
really that much of a problem? Troll tech's license sounds reasonable
to me - Qt libraries are mostly unrestricted if you are developing
freeware and both Qt and KDE were on my Linux Systems Lab Redhat CD
(contrib - but still better than Netscape which I have to waste time
downloading from their site). I have never used CDE so I can't even
compare the free contenders to that one.
dara parsavand | pars...@prony.colorado.edu (do not use header address)
Nathan Hand
pars...@boulder.Colorado.EDU (Dara Parsavand) writes:
> I'd be interested in hearing from Linux (free BSD, etc.) advocates on
> what they think the future of the free Desktop is.
GNUStep.
> Do you think Qt is really that much of a problem?
Yes, the licensing issue is a problem.
> Troll tech's license sounds reasonable
> to me - Qt libraries are mostly unrestricted if you are developing
> freeware and both Qt and KDE were on my Linux Systems Lab Redhat CD
And commercial developers? Not all software is free, and I'd like
all my Linux apps to have the same look and feel. Not one type of
look and feel for free apps, and another for commercial apps.
There are other problems with the QT license anyway. Be very wary
of things like QT: they may not be around tomorrow. Free software
doesn't suffer from company decisions, which is why I prefer it.
--
The idea that an arbitrary naive human should be able to properly use
a given tool without training or understanding is even more wrong for
computing than it is for other tools (eg automobiles, airplanes, guns
or power saws).
o r c e l l . p o r t l a n d . o r . u s
In article <m2hgcia...@stoli.spirits.org.au>,
Nathan Hand <natha...@anu.edu.au> wrote:
>There are other problems with the QT license anyway. Be very wary
>of things like QT: they may not be around tomorrow.
And if Troll Tech rolls belly up and vanishes, that will leave 50,000
people with copies of the source, and no more copyright holder.
Dumb argument, even for the not invented here crowd.
____
david parsons \bi/ Remember, Qt is bad because Troll Tech likes to eat.
\/
Alan Shutko
>>>>> "david" == david parsons <o> writes:
david> In article <m2hgcia...@stoli.spirits.org.au>, Nathan Hand
david> <natha...@anu.edu.au> wrote:
>> There are other problems with the QT license anyway. Be very wary
>> of things like QT: they may not be around tomorrow.
david> And if Troll Tech rolls belly up and vanishes, that will
david> leave 50,000 people with copies of the source, and no more
david> copyright holder.
And no legal way to maintain it or release fixed versions. Unless the
QT license allows that if they go under?
--
Alan Shutko <a...@acm.org> - By consent of the corrupted
Let's organize this thing and take all the fun out of it.
bobf
david parsons <o r c @ p e l l . p o r t l a n d . o r . u s > wrote in
article <5tnesk$p...@pell.pell.portland.or.us>...
> In article <m2hgcia...@stoli.spirits.org.au>,
> Nathan Hand <natha...@anu.edu.au> wrote:
>
> >There are other problems with the QT license anyway. Be very wary
> >of things like QT: they may not be around tomorrow.
>
> And if Troll Tech rolls belly up and vanishes, that will leave 50,000
> people with copies of the source, and no more copyright holder.
>
>
> Dumb argument, even for the not invented here crowd.
I thought the argument was mostly valid... By your reasoning the Commodore
Amiga should be around, as for the Commodore 64, Heathkit, TRS-80 Models 1
and 3, Commodore PET, Apple II and e and gs, and so on. Keyword=Customer
Base!=Source Code.
o r c e l l . p o r t l a n d . o r . u s
In article <m27mdcl...@hubert.wuh.wustl.edu>,>david> In article <m2hgcia...@stoli.spirits.org.au>, Nathan Hand
>david> <natha...@anu.edu.au> wrote:
>
>>> There are other problems with the QT license anyway. Be very wary
>>> of things like QT: they may not be around tomorrow.
>
>david> leave 50,000 people with copies of the source, and no more
>david> copyright holder.
>
>And no legal way to maintain it or release fixed versions. Unless the
>QT license allows that if they go under?
What happens to a copyright when the copyright holder falls over
dead? Either the copyright is given to someone else, at which point
you can pester them for support, or it lapses into the public domain.
Either way, the people who are using Qt won't lose; they have the
source or a copyright owner who can fix any problems.
____
david parsons \bi/ Not invented here is a pretty icky idea.
\/
JEDI
On 25 Aug 1997 11:34:41 +1000, Nathan Hand <natha...@anu.edu.au> wrote:
>o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
>
>> Nathan Hand <natha...@anu.edu.au> wrote:
>>
>> >There are other problems with the QT license anyway. Be very wary
>> >of things like QT: they may not be around tomorrow.
>>
>> people with copies of the source, and no more copyright holder.
>
>No, with a defunct copyright holder, and nobody with the right
>to do anything to the source code. That's the QT license.
With a defunt copyright holder, there is no one left to sue
you over copy rights. We can do as we like if there is no one
left around to complain.
Roberto Alsina
In article <m23env9...@hubert.wuh.wustl.edu>, Alan Shutko wrote:
>>>>>> "D" == david parsons <o> writes:
>
>>> And no legal way to maintain it or release fixed versions. Unless
>>> the QT license allows that if they go under?
>
>D> dead? Either the copyright is given to someone else, at which
>D> point you can pester them for support, or it lapses into the public
>D> domain.
>
>Nope. If an author created a work now, the copyright would be valid
>for 50 years after he died.
It would ve valid, but not assigned to the author, since he is dead.
It would be assigned to his heirs (not sure about the word, the guys he
left his things to :-)
>D> Either way, the people who are using Qt won't lose; they have the
>D> source or a copyright owner who can fix any problems.
>
>So, they'll have the source, but no legal way to change it. Just
>because Troll Tech goes under does not mean that the copyright lapses
>into public domain within a reasonable timeframe. Someone still owns
>it. Perhaps it'll be considered joint custody of the people who had
>been part of TT, or maybe it'll be bought along with whatever assets
>of TT exist.
It's C++, guys!. They can't change the licensing of 1.21, right?. Since it's
C++, you probably don't need to fix anything. Just inherit and replace or
extend.
--
("\''/").__..-''"`-. . Roberto Alsina
`9_ 9 ) `-. ( ).`-._.`) ral...@unl.edu.ar
(_Y_.)' ._ ) `._`. " -.-' Centro de Telematica
_..`-'_..-_/ /-'_.' Universidad Nacional del Litoral
(l)-'' ((i).' ((!.' Santa Fe - Argentina
Nathan Hand
je...@Dementia.sprynet.com (JEDI) writes:
Wrong. Defunct doesn't mean they're not legally active. It just
means they are not developing anymore. As somebody else already
pointed out, the defunct copyright holder could just be sitting
on QT as an asset, but not something they want to develop. This
would kill any QT development, and QT apps would stagnate. Note
the defunct copyright holder doesn't have to be Trolltech.
There are other problems with QT too, like the way I can't take
my free apps and compile them for non-UNIX platforms without me
forking over truckloads of dosh. I want my free apps to be used
on all platforms. Cygnus has expended lots of effort making the
GNU-Win32 subsystem, just to allow easy porting of UNIX apps. I
don't want to choose a widget set that makes it hard again.
Free software doesn't mean UNIX software. It means free for all
people, regardless of chosen platform.
PS: Once again, I have nothing against QT being commercial, and
I thank Trolltech for providing such a mature and useful widget
set for Linux developers to start coding great apps with. But I
don't think QT should become the standard widget set. It is not
very conducive to creating truly free software. QT is better as
a cross-platform widget set for building commercial apps.
Nathan Hand
o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
> Nathan Hand <natha...@anu.edu.au> wrote:
> >je...@Dementia.sprynet.com (JEDI) writes:
>
> >> With a defunt copyright holder, there is no one left to sue
> >> you over copy rights. We can do as we like if there is no one
> >> left around to complain.
>
> >Wrong. Defunct doesn't mean they're not legally active. It just
> >means they are not developing anymore.
>
> Ah. Foolish me, I thought that Australians spoke *english*;
> I was not aware that your language had different meanings
> for common English words.
>
> DE-FUNCT adj. Having ceased to live or exist; extinct; dead.
> [Latin _defunctus_, past participle of _defungi_, to discharge,
> finish, die: _de-_ (intensive) + _fungi_, to discharge (...)]
>
> So, when a native english speaker says something is defunct,
> it means that thing is dead. What exactly does `defunct' mean
> in your language?
Please don't be condescending. Companies can be dead, but very much
a legally active entity. Defunct companies can still own the rights
to a product, and can prevent others from owning or developing such
products. They can have no money, and sell nothing, but own lots.
Remember that just because the company doesn't work anymore (ie, it
is "dead") does not mean the owners are also dead, or that whatever
the company owned is now owned by the public. The company can still
sue you "from the grave", so to speak, if you break the copyright.
> >There are other problems with QT too, like the way I can't take
> >my free apps and compile them for non-UNIX platforms without me
> >forking over truckloads of dosh.
>
> You can't be as stupid as this. It's common practice to WRITE
> YOUR OWN LIBRARY if you really want to port to a platform that
> doesn't have that library.
What? QT is already cross-platform, and I find it damn useless that
any freeware QT apps I write for Linux can't be freeware on Windows
with the Windows version of the QT libs. It means I've got to spend
time porting to a free Windows lib, then maintain two source trees.
It's hardly in the spirit of free software, to permit freeware apps
for UNIX users, but deny it to Windows users. I don't think it's at
all "stupid" to want all users to have easy access to free software
irregardless of platform.
> And I may be naive, but I don't
> see Troll Tech attempting to enforce an interface copyright.
Read the QT license, or their FAQ. It clearly states that you can't
use a freeware QT license to develop freeware apps on Windows. They
suggest you purchase a commercial QT license (for lots of dosh) and
then you can write freeware apps on Windows.
That's money out of your pocket, of course. And nobody but you, and
other people who have paid for the *commercial* QT license, will be
able to compile your "free" software on Windows. It's not very free
(in the bird sense) if you must purchase a cage to keep it in.
> >But I
> >don't think QT should become the standard widget set. It is not
> >very conducive to creating truly free software.
>
> Nonsense. It would be difficult to find a licensing scheme
> which is more conductive to ensuring that the software remain
> free than the one that QT comes with.
Denying non-UNIX users easy access to your QT apps does not help to
spread the concept of free software, in either sense of the word.
> (I can refuse to release
> internal applications linked with GPLed software; but I _have_
> _to_ release the sources to internal applications that are
> linked with the free QT library.)
Really? That is crazy.
> ____
> david parsons \bi/ and the taint of `not invented here' fills the air.
> \/
I don't care where it's invented, but it's not a good library for a
free software community. It isn't free for everyone, but only for a
very select few. I'll refrain from saying the B word, because I can
be fairly certain other people see the problem for themselves.
It's OK that QT is commercial, and it is wonderful that Trolltech's
been kind enough to let us use it for free. However I hate to think
that people see this license as being "just as good as the GPL", as
it's only free when writing free UNIX applications.
Because it *isn't* as good as the GPL.
And please tone down any replies, David. I can get really nasty too
if I want to, but I don't think it's necessary here.
--
The sticker on the side of the box said "Supported Platforms: Windows 95,
Windows NT 4.0, or better", so clearly Linux was a supported platform.
Vlad Petersen
david parsons wrote:
............
> >There are other problems with QT too, like the way I can't take
> >my free apps and compile them for non-UNIX platforms without me
> >forking over truckloads of dosh.
>
> You can't be as stupid as this. It's common practice to WRITE
> YOUR OWN LIBRARY if you really want to port to a platform that
> see Troll Tech attempting to enforce an interface copyright.
>
> >don't think QT should become the standard widget set. It is not
> >very conducive to creating truly free software.
>
> Nonsense. It would be difficult to find a licensing scheme
> which is more conductive to ensuring that the software remain
I do not think that's true. You can pay a fee and only then sell
your qt based software. Otherwise, the QT license denies you the
freedom of selling your application. No wonder Linux vendors
refuse to ship QT libs with their distributions! On the other hand,
GPL does provide me the freedom to sell, let's say, GCC and
Emacs.
> internal applications linked with GPLed software; but I _have_
> _to_ release the sources to internal applications that are
> linked with the free QT library.)
That by no means means that QT is more free than GPL is:
e.g. I do not even have a single QT library of software installed
on my system and it's been working that way ever since. But if I
removed every GNU package from my Linux system (or just some little
library), it would make the system unuseable.
QT developers are motivated by greed, plain and simple. GNU project
is much more motivated by the spirit of free software.
> ____
> david parsons \bi/ and the taint of `not invented here' fills the air.
> \/
--
My real email address is: <vladimip at uniserve dot com>
#include <disclaimer.h> | *Good pings come in small packets*
Vancouver | Windows: for IQs smaller than 95
British Columbia | SIGSIG -- signature too long (core dumped)
o r c e l l . p o r t l a n d . o r . u s
In article <m2vi0mz...@stoli.spirits.org.au>,
Nathan Hand <natha...@anu.edu.au> engages in a bout of criminally
stupid posting:
[I wrote]
>> You can't be as stupid as this. It's common practice to WRITE
>> YOUR OWN LIBRARY if you really want to port to a platform that
>> doesn't have that library.
.
.
.
>> And I may be naive, but I don't
>> see Troll Tech attempting to enforce an interface copyright.
>
>Read the QT license, or their FAQ. It clearly states that you can't
>use a freeware QT license to develop freeware apps on Windows.
You don't get it, do you?
I'll use small words: If I write my own library that looks like QT,
then I don't have to pay Troll Tech any money when I use that
library.
____
david parsons \bi/ I take it back; it was a bad idea to let hoi polloi
\/ onto the net.
Mr K. Nice
Oh Vlad Petersen <vlad...@compuserve.com> badger! Don't you see?
>
> david parsons wrote:
> > Nonsense. It would be difficult to find a licensing scheme
> > which is more conductive to ensuring that the software remain
> > free than the one that QT comes with. (I can refuse to release
>
> I do not think that's true. You can pay a fee and only then sell
> your qt based software. Otherwise, the QT license denies you the
> freedom of selling your application. No wonder Linux vendors
> refuse to ship QT libs with their distributions! On the other hand,
> GPL does provide me the freedom to sell, let's say, GCC and
> Emacs.
>
Unless I've misunderstood the Qt licence, any GPL software produced
using Qt can be sold in the same way as any other GPL software, i.e.,
as long as source code is available.
> > internal applications linked with GPLed software; but I _have_
> > _to_ release the sources to internal applications that are
> > linked with the free QT library.)
>
> That by no means means that QT is more free than GPL is:
> e.g. I do not even have a single QT library of software installed
> on my system and it's been working that way ever since. But if I
> removed every GNU package from my Linux system (or just some little
> library), it would make the system unuseable.
>
Huh? How does that make GPL code more free?
> QT developers are motivated by greed, plain and simple.
No, I think the Qt developers are motivated by the desire to get paid
for their work. Sounds fair enough to me.
> GNU project is much more motivated by the spirit of free software.
Well, that's true, but the Qt developers obviously care to some extent
about free software, otherwise why the free software licencing?
--
/ If you call this a short .sig, you oppose its reality. | Tim Fisken. /
/ If you do not call it a short .sig, you ignore the fact. | That's my /
/ Now say quickly what it is. | name. /
Per Abrahamsen
[ Fut: gnu.misc.discuss ]
mrn...@rose-cottage.demon.co.uk (Mr K. Nice) writes:
> Well, that's true, but the Qt developers obviously care to some extent
> about free software, otherwise why the free software licencing?
Qt is gratis for some users for much the same reasons Netscape is
gratis for some users, and that DEC has donated gratis hardware to
educational institutions.
These examples gratis software and hardware have little in common with
the kind of freedom you get with free software like FreeBSD, Linux GNU
software.
To some, the gratis software is better, and to others the free (in the
sense above) software is better. I wish more people would recognize
this, and stop flaming when someone asks for a free alternative to a
gratis product.
Per Abrahamsen
[ Fut: gnu.misc.discuss ]
mrn...@rose-cottage.demon.co.uk (Mr K. Nice) writes:
> Unless I've misunderstood the Qt licence, any GPL software produced
> using Qt can be sold in the same way as any other GPL software, i.e.,
> as long as source code is available.
The Qt license is deliberately misleading. The Qt license adds
additional restrictions, something which the GPL forbids. This means
that the two licenses are in conflict. The result is that you cannot
legally distribute a program which contains both Qt and GPL'ed code.
Marinos J. Yannikos
In article <5tduap$3...@lace.colorado.edu>, Dara Parsavand wrote:
>[...]
>I'd be interested in hearing from Linux (free BSD, etc.) advocates on
>what they think the future of the free Desktop is.
As long as people keep on fighting over silly licensing details and insist
on cooking their own brew instead of contributing to at least one stable,
usable solution, the world of desktops will stay as much of a mess as
most other issues. If you want the least hassle, stick with something
non-volatile such as twm if you can live with its bugs (have you noticed
any?) and perhaps add a few gimmicks such as tkGoodstuff. Just my opinion.
-nino
--
Please change the last part of my address to "at" if you're replying by mail.
It's nice to be important, but it's more important to be nice.
Vlad Petersen
> > I do not think that's true. You can pay a fee and only then sell
> > your qt based software. Otherwise, the QT license denies you the
> > freedom of selling your application. No wonder Linux vendors
> > refuse to ship QT libs with their distributions! On the other hand,
> > GPL does provide me the freedom to sell, let's say, GCC and
> > Emacs.
> Unless I've misunderstood the Qt licence, any GPL software produced
> using Qt can be sold in the same way as any other GPL software, i.e.,
> as long as source code is available.
Yes, you've misunderstood their license which forbids reselling
QT-based applications unless you pay a big fee to troll.no
........
> > That by no means means that QT is more free than GPL is:
> > e.g. I do not even have a single QT library of software installed
> > on my system and it's been working that way ever since. But if I
> > removed every GNU package from my Linux system (or just some little
> > library), it would make the system unuseable.
> >
> Huh? How does that make GPL code more free?
Now it seems that you never read neither QT license nor GPL.
> > QT developers are motivated by greed, plain and simple.
>
> No, I think the Qt developers are motivated by the desire to get paid
> for their work. Sounds fair enough to me.
There is a number of companies selling/supporting free software,
Cygnus, RedHat, Caldera and others. I would say that they are
equally motivated by both desire to get paid and the free
philosophy. Notice that QT license terms highly contradict
to GPL which makes redistribution of QT and QT based applications
on Linux CDROMs legally impossible: solution providers will not
be able to sell such a CDROM or related services to customers
because of QT license. That was my point.
> > GNU project is much more motivated by the spirit of free software.
>
> Well, that's true, but the Qt developers obviously care to some extent
> about free software, otherwise why the free software licencing?
To get people hooked on QT, plain and simple.
Roberto Alsina
In article <340DA01C...@compuserve.com>, Vlad Petersen wrote:
>Mr K. Nice wrote:
>> Oh Vlad Petersen <vlad...@compuserve.com> badger! Don't you see?
>........
>> > I do not think that's true. You can pay a fee and only then sell
>> > your qt based software. Otherwise, the QT license denies you the
>> > freedom of selling your application. No wonder Linux vendors
>> > refuse to ship QT libs with their distributions! On the other hand,
>> > GPL does provide me the freedom to sell, let's say, GCC and
>> > Emacs.
>......
>> Unless I've misunderstood the Qt licence, any GPL software produced
>> using Qt can be sold in the same way as any other GPL software, i.e.,
>> as long as source code is available.
>
>Yes, you've misunderstood their license which forbids reselling
>QT-based applications unless you pay a big fee to troll.no
Nope. the Qt license does *not* forbid you sell a GPLed app using Qt.
It forbids you sell it and not give away the source under the GPL terms.
You could ask one of the trolls about it.
>........
>> > That by no means means that QT is more free than GPL is:
>> > e.g. I do not even have a single QT library of software installed
>> > on my system and it's been working that way ever since. But if I
>> > removed every GNU package from my Linux system (or just some little
>> > library), it would make the system unuseable.
>> >
>> Huh? How does that make GPL code more free?
>
>Now it seems that you never read neither QT license nor GPL.
Since I develop software using Qt and release it under the GPL, my
software is as free as, say, gcc.
The only restriction is that if somebody was mad enough to try using my
software on Windows, he should get somebody with a commercial Qt license
to compile it for him.
That's no different from having to find somebody with a Motif license to
build a static binary for a GPLed program using Motif.
>> > QT developers are motivated by greed, plain and simple.
>>
>> No, I think the Qt developers are motivated by the desire to get paid
>> for their work. Sounds fair enough to me.
>
>There is a number of companies selling/supporting free software,
>Cygnus, RedHat, Caldera and others. I would say that they are
>equally motivated by both desire to get paid and the free
>philosophy. Notice that QT license terms highly contradict
>to GPL which makes redistribution of QT and QT based applications
>on Linux CDROMs legally impossible: solution providers will not
>be able to sell such a CDROM or related services to customers
>because of QT license. That was my point.
The only distribution that says it won't distribute Qt based software
because of Qt's license is Debian, and they actually put that software in
their non-free directory.
Red Hat has different reasons, their problem is not Qt's runtime license,
but Qt's development license, and that it would confuse their customers.
Caldera, as well as other Linux distributions have no problem. Actually
some of them sponsored a KDE (GPLed project based on Qt) event, called
KDE-ONE.
So, I would say there is quite a bit of evidence that Qt's license
doesn't prevent Qt based software from being placed on a Linux CD.
Looks like you have not read Qt's license.
>> > GNU project is much more motivated by the spirit of free software.
>>
>> Well, that's true, but the Qt developers obviously care to some extent
>> about free software, otherwise why the free software licencing?
>
>To get people hooked on QT, plain and simple.
Now that's funny.
Bruce Stephens
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> Since I develop software using Qt and release it under the GPL, my
> software is as free as, say, gcc. The only restriction is that if
> somebody was mad enough to try using my software on Windows, he
> should get somebody with a commercial Qt license to compile it for
> him.
Erm, but what about the GPL? If your program is called P (which is
covered by GPL), then the program combined with the library that it
requires (P+Qt) is presumably a derived work of P (and of Qt, of
course, but the Qt license is OK about this). But P+Qt then can't be
distributed, since the GPL forbids this, doesn't it?
> That's no different from having to find somebody with a Motif license to
> build a static binary for a GPLed program using Motif.
Pretty much, yes. XEmacs doesn't require Motif (and requires no
source changes to compile without it), so perhaps that's how it's
covered. But DDD certainly does (or did, last I looked); perhaps they
have a slightly modified license to allow this? (For that matter, you
could do that too. Just to make things *really* clear, why don't the
Qt people provide a freeish license, like the GPL but allowing
distribution with Qt?)
To a large extent, since you're presumably the copyright owner, so
long as you don't mind about the way that P+Qt is distributed, then
there's no problem, and you can say up front the view you take, if
it's not obvious. You can surely see why some people feel a bit
unhappy about the situation though? I've no idea how RedHat and
people feel; I assume KDE isn't finished enough to be important enough
yet.
I assume the GTK/Gnome people think their toolkit (which seems to work
fine for the GIMP) is technically as good as Qt, and is designed to be
callable from a variety of languages, so what's wrong with them using
it?
I agree it would be nice for everybody to agree on one free toolkit,
but Qt isn't going to be it, just because it's not free on Windoze and
Macs. I'd hoped that Tk might be it, but it really doesn't seem to be
moving to language independence, although I still hope people get it
there one day.
Nathan Hand
o r c @ p e l l . p o r t l a n d . o r . u s (david parsons) writes:
> Nathan Hand <natha...@anu.edu.au> engages in a bout of criminally
> stupid posting:
Uh-huh, ok...
> nathan wrote
> > orc wrote
> >
> >> You can't be as stupid as this. It's common practice to WRITE
> >> YOUR OWN LIBRARY if you really want to port to a platform that
> >> doesn't have that library.
> .
> .
> .
> >> And I may be naive, but I don't
> >> see Troll Tech attempting to enforce an interface copyright.
> >
> >Read the QT license, or their FAQ. It clearly states that you can't
> >use a freeware QT license to develop freeware apps on Windows.
>
> You don't get it, do you?
>
> I'll use small words: If I write my own library that looks like QT,
> then I don't have to pay Troll Tech any money when I use that
> library.
So all I've got to do is write a library, which works on Windows
and Linux, and has the same look, and compiles the same code, on
both platforms without modification, using the exact same API as
QT, and *then* I should use QT for Linux?
Is this what you're trying to say?
I'll say it again: QT is not helping free software by making the
definition of free software be "anything that costs no money and
runs on UNIX platforms". It ignores non-UNIX platforms, and this
discrimination alone suggests another library is better.
QT doesn't allow for the idea of _free_ software.
> ____
> david parsons \bi/ I take it back; it was a bad idea to let hoi polloi
> \/ onto the net.
You speaking in tongues now?
Matt McLeod
In article <340E33AA...@compuserve.com>,
Vlad Petersen <vlad...@compuserve.com> wrote:
:Eirik Eng wrote this extremely confusing stuff:
:[snip]
:Here you say you can resell:
:
:> No, he has not misunderstood our license. Nowhere does it say that you are
:> not allowed to resell Qt-based applications if you develop them using the
:> Troll Tech Free License. Please read our license:
:......
:> Our license policy is simple:
:>
:> If you want to GPL your software, put it under the FreeBSD license
:
:I don't understand that. If I want to GPL my software, I put it
:under GPL. If I want to license my software under BSD terms, I
:use BSD license. Please don't mix those in one bunch.
I read that as saying it is OK to license your software under GPL, BSD, or
public domain. Not GPL==BSD.
:> or put it in the public domain etc, you can do so without paying a
:> single cent, and you can sell as much of it for as much as you want.
:
:Above you said one can sell.
:
:> If you don't want to make free software that's fine with us, but
:> then you should pay us a fee.
:
:Above you said one can't sell. Please explain how it is not in
:contradiction with two previous statements of yours I quoted.
There is nothing in the GPL which says you can't sell GPL'd software. So if
you write your software using the free Qt license, and release it under GPL,
then you can sell your software (but you have to provide source too).
:Do you folks have three different licenses??
:
:Why not simply use GPL if yours is free as well?
Presumably because they want to encourage free software, but also want to
eat. So they let people use Qt for free software at no cost, but require
payment for non-free software (i.e., where it is not licensed under one of
the "freeware" licenses such as GPL or BSD).
:> > solution providers will not be able to sell such a CDROM or related
:> > services to customers because of QT license. That was my point.
:>
:> This is simply not true, where did you get this information? Have you
:> actually read our license? Please also take a look at our FAQ:
:> http://www.troll.no/faq/faq-freeedition.html
:
:I just did. It is in contradiction to some parts of your post.
I didn't see any contridiction in his post. It made perfect sense to me.
--
Matt McLeod, <mjm(at)attila.apana.org.au>
"Please try to understand before one of us dies".
Paul Seelig
Vlad Petersen <vlad...@compuserve.com> writes:
> Eirik Eng wrote this extremely confusing stuff:
> [snip]
> Here you say you can resell:
>
> > No, he has not misunderstood our license. Nowhere does it say that you are
> > not allowed to resell Qt-based applications if you develop them using the
> > Troll Tech Free License. Please read our license:
> ......
> > Our license policy is simple:
> >
> > If you want to GPL your software, put it under the FreeBSD license
>
> I don't understand that. If I want to GPL my software, I put it
> under GPL. If I want to license my software under BSD terms, I
> use BSD license. Please don't mix those in one bunch.
>
It doesn't help your or other readers understanding very much putting
the line you quoted completely out of context! The whole paragraph
actually states:
If you want to GPL your software, put it under the FreeBSD license
or put it in the public domain etc, you can do so without paying a
single cent, and you can sell as much of it for as much as you want.
This gives you the three options putting your software *either* under
the GPL *or* the FreeBSD license *or* in the public domain. There is
no statement saying that one should put his software under the FreeBSD
license in order to GPL it!? Please take care to read more carefully!
> > or put it in the public domain etc, you can do so without paying a
> > single cent, and you can sell as much of it for as much as you want.
>
> Above you said one can sell.
>
> > If you don't want to make free software that's fine with us, but
> > then you should pay us a fee.
>
> Above you said one can't sell. Please explain how it is not in
> contradiction with two previous statements of yours I quoted.
>
The difference is about free software published with reusable source
code for the user's modification pleasure and strictly commercial
proprietary software published without any reuseable source code at
all. Is this really so hard to understand?
> Do you folks have three different licenses??
>
Do you actually try to understand what has been written here or do you
prefer to be simply provocative? You seem to be really trying very
hard to confuse what Eirik Eng actually wrote very clearly! Not very
fair actually... :-(
Cheers, P. *8^)
--
Paul Seelig pse...@goofy.zdv.uni-mainz.de
African Music Archive - Institute for Ethnology and Africa Studies
Johannes Gutenberg-University - Forum 6 - 55099 Mainz/Germany
My Homepage in the WWW at the URL http://www.uni-mainz.de/~pseelig
Roberto Alsina
In article <m3zppu5...@cenderis.demon.co.uk>, Bruce Stephens wrote:
>ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
>> Since I develop software using Qt and release it under the GPL, my
>> software is as free as, say, gcc. The only restriction is that if
>> somebody was mad enough to try using my software on Windows, he
>> should get somebody with a commercial Qt license to compile it for
>> him.
>
>Erm, but what about the GPL? If your program is called P (which is
>covered by GPL), then the program combined with the library that it
>requires (P+Qt) is presumably a derived work of P (and of Qt, of
>course, but the Qt license is OK about this). But P+Qt then can't be
>distributed, since the GPL forbids this, doesn't it?
That argument is not good: In a Solaris box, P needs libc. P+libc is a
derived work of P, I can't distribute P+libc because the GPL forbids it.
I don't believe a program that uses a pre-existant library somehow turns
that library into part of a derived work.
>> That's no different from having to find somebody with a Motif license to
>> build a static binary for a GPLed program using Motif.
>
>Pretty much, yes. XEmacs doesn't require Motif (and requires no
>source changes to compile without it), so perhaps that's how it's
>covered. But DDD certainly does (or did, last I looked); perhaps they
>have a slightly modified license to allow this? (For that matter, you
>could do that too. Just to make things *really* clear, why don't the
>Qt people provide a freeish license, like the GPL but allowing
>distribution with Qt?)
>
>To a large extent, since you're presumably the copyright owner, so
>long as you don't mind about the way that P+Qt is distributed, then
>there's no problem, and you can say up front the view you take, if
>it's not obvious. You can surely see why some people feel a bit
>unhappy about the situation though? I've no idea how RedHat and
>people feel;
I have read statements from the Red Hat people, and their problem is not
with redistribution of KDE, but that if they include Qt, their buyers
could fall in love with it ;-), and couldn't develop commercial software
using it.
>I assume KDE isn't finished enough to be important enough yet.
Then you haven't tried it lately :-)
>I assume the GTK/Gnome people think their toolkit (which seems to work
>fine for the GIMP) is technically as good as Qt, and is designed to be
>callable from a variety of languages, so what's wrong with them using
>it?
Nothing, obviously. They can use whatever they like. I wouldn't use GTK--
instead of Qt for C++ development, but that's just me.
>I agree it would be nice for everybody to agree on one free toolkit,
>but Qt isn't going to be it, just because it's not free on Windoze and
>Macs. I'd hoped that Tk might be it, but it really doesn't seem to be
>moving to language independence, although I still hope people get it
>there one day.
Well, GTK doesn't even exist on macs and windows, so I guess that's not
it either?
And... you *can* build free software for windows95/NT using the X version of
Qt, cygwin32, the X client libraries, a free X server for '95 like
microimage's, and a few other thingies (But I wouldn't do it except for a
hefty amount of money, so I can pay my shrink ;-).
Something like that is done for OS/2, too.
Best regards.
Roberto Alsina
In article <340E33AA...@compuserve.com>, Vlad Petersen wrote:
>Eirik Eng wrote this extremely confusing stuff:
>[snip]
>Here you say you can resell:
>
>> No, he has not misunderstood our license. Nowhere does it say that you are
>> not allowed to resell Qt-based applications if you develop them using the
>> Troll Tech Free License. Please read our license:
>......
>> Our license policy is simple:
>>
>> If you want to GPL your software, put it under the FreeBSD license
>
>I don't understand that. If I want to GPL my software, I put it
>under GPL. If I want to license my software under BSD terms, I
>use BSD license. Please don't mix those in one bunch.
He meant "If you want to GPL your software, *or* put it under the FreeBSD
license", etc.
>
>> or put it in the public domain etc, you can do so without paying a
>> single cent, and you can sell as much of it for as much as you want.
>
>Above you said one can sell.
>
>> If you don't want to make free software that's fine with us, but
>> then you should pay us a fee.
>
>Above you said one can't sell. Please explain how it is not in
>contradiction with two previous statements of yours I quoted.
>
>Do you folks have three different licenses??
>
>Why not simply use GPL if yours is free as well?
Because they want commercial software to pay them!
You can write GPL software using Qt, and then sell it, and not pay anything.
You can write commercial software using Qt, then sell it, and pay the trolls.
It's not that hard :-)
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>> Erm, but what about the GPL? If your program is called P (which is
>> covered by GPL), then the program combined with the library that it
>> requires (P+Qt) is presumably a derived work of P (and of Qt, of
>> course, but the Qt license is OK about this). But P+Qt then can't
>> be distributed, since the GPL forbids this, doesn't it?
ra> That argument is not good: In a Solaris box, P needs libc. P+libc
ra> is a derived work of P, I can't distribute P+libc because the GPL
ra> forbids it.
Not true, the GPL has an exception that covers libc, libm, etc. The FSF
has also ruled that it covers Motif, because Motif is a very commonly
supplied as a standard part of UNIX systems these days.
Troll argues that this means the exception also applies to Qt in the
same way. However, I, for one, am not 100% convinced of this. The
language in the part of the GPL defining the exception contains numerous
fuzzy terms, and the availability, etc. of Motif and Qt differ in
important and seemingly relevant ways.
If Qt is actually not allowed under the exception, then it would be
illegal to distribute GPL'd code that links with Qt, since the licenses
would conflict (the same way as the GPL and BSD licenses conflict).
Troll needs to contact the FSF directly and ask them. Only then will
Qt's status be completely clear.
The last time this thread came up, someone at Troll emailed me saying
that RMS had in fact approved their usage, but they would not agree to
show me his message to that effect until they checked with him. That
was over a month ago but I haven't heard anything else. So perhaps this
is already resolved, but I think Troll should put the FSF's statement up
on their WWW site so all can see it.
Until that happens, regardless of what Troll writes on their WWW site
FAQ, legitimate questions remain about Qt's precise status.
ra> I don't believe a program that uses a pre-existant library somehow
ra> turns that library into part of a derived work.
Sure it does. But more importantly for this argument, using the
pre-existent library turns the _program_ into a derived work.
At least that's the FSF's claim. AFAIK it hasn't been tested in court
yet, but it's not at all obvious that they're wrong.
If they _are_ wrong, then the GPL itself is for the most part moot.
--
-------------------------------------------------------------------------------
Paul D. Smith <psm...@baynetworks.com> Network Management Development
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
-------------------------------------------------------------------------------
These are my opinions--Bay Networks takes no responsibility for them.
Roger Espel Llima
In article <340e58b3...@news.rhrz.uni-bonn.de>,
Matthias Moeller <uzs...@uni-bonn.de> wrote:
>So, where is the problem ? If you want to distribute free software you
>dont have to pay for your QT-License. If you sell software which is
>not free you want to earn money. And in this case it is at least
>understandable that troll wants to get a fee because you are proftting
>from their work.
The problem is that TrollTech does not allow redistribution of modified
Qt libraries, so if one day TrolTech folds or decides to stop
distributing a free editing, the free libraries cannot evolve anymore,
or be ported, and all programs that use them are guaranteed a slow death
unless someone makes a free equivalent to Qt. This obviously does not
happen with programs that use free libraries.
Suggestion: distribute a version of Qt's source under the GPL, in
addition to the existing free and commercial licenses. It doesn't
hamper the position of the commercial library, since any programs using
the GPL version must be GPL too, and it would reassure free software
developpers.
--
Roger Espel Llima
es...@llaic.univ-bpclermont.fr, es...@unix.bigots.org
http://www.eleves.ens.fr:8080/home/espel/index.html
Roberto Alsina
In article <p590xc3...@baynetworks.com>, Paul D. Smith wrote:
>%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> >> Erm, but what about the GPL? If your program is called P (which is
> >> covered by GPL), then the program combined with the library that it
> >> requires (P+Qt) is presumably a derived work of P (and of Qt, of
> >> course, but the Qt license is OK about this). But P+Qt then can't
> >> be distributed, since the GPL forbids this, doesn't it?
>
> ra> That argument is not good: In a Solaris box, P needs libc. P+libc
> ra> is a derived work of P, I can't distribute P+libc because the GPL
> ra> forbids it.
>
>Not true, the GPL has an exception that covers libc, libm, etc. The FSF
>has also ruled that it covers Motif, because Motif is a very commonly
>supplied as a standard part of UNIX systems these days.
But not on Linux, not on Windows sytems, not on FreeBSD, NetBSD, OpenBSD,
not on the HURD, and is not part of the proposed GNU system the FSF so
much likes to talk about.
This argument would mean that a app developed using Motif would be
distributable under the GPL in some systems and not on others. That is
directly against the GPL isn't it?
I mean, if I say: this is GPL only on Solaris, you would claim that
restriction is incompatible with the GPL, right?
>Troll argues that this means the exception also applies to Qt in the
>same way. However, I, for one, am not 100% convinced of this. The
>language in the part of the GPL defining the exception contains numerous
>fuzzy terms, and the availability, etc. of Motif and Qt differ in
>important and seemingly relevant ways.
Qt is available for no cost for almost every Unix variant, Motif isn't.
I guess that alone would make Qt more acceptable than Motif as a platform
for GPLed software.
>If Qt is actually not allowed under the exception, then it would be
>illegal to distribute GPL'd code that links with Qt, since the licenses
>would conflict (the same way as the GPL and BSD licenses conflict).
Well, I wrote the code, and I distribute it, so I am the judge as to if
Qt falls under the exception, not the FSF. It's *my* code.
>Troll needs to contact the FSF directly and ask them. Only then will
>Qt's status be completely clear.
The FSF is not the owner of the GPL, nor its guardian, just the authors.
>The last time this thread came up, someone at Troll emailed me saying
>that RMS had in fact approved their usage, but they would not agree to
>show me his message to that effect until they checked with him. That
>was over a month ago but I haven't heard anything else. So perhaps this
>is already resolved, but I think Troll should put the FSF's statement up
>on their WWW site so all can see it.
>
>Until that happens, regardless of what Troll writes on their WWW site
>FAQ, legitimate questions remain about Qt's precise status.
>
> ra> I don't believe a program that uses a pre-existant library somehow
> ra> turns that library into part of a derived work.
>
>Sure it does.
I can't turn something I didn't write or own into a derived work!
I didn't even touch it!
> But more importantly for this argument, using the
>pre-existent library turns the _program_ into a derived work.
I don't agree with this either (perhaps I'm just stubborn :-)
If the program doesn't include code from the library, it's not a derived
work of the library.
That distinction is actually made by the FSF in the LGPL.
That's why linking to a LGPLed library does *not* make my app GPLed.
>At least that's the FSF's claim. AFAIK it hasn't been tested in court
>yet, but it's not at all obvious that they're wrong.
It's not so obvious that they are right either ;-)
>If they _are_ wrong, then the GPL itself is for the most part moot.
Well, every lawyer I have shown it to says that :-(
Bruce Stephens
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> That argument is not good: In a Solaris box, P needs libc. P+libc is a
> derived work of P, I can't distribute P+libc because the GPL forbids it.
There's an explicit exemption for things normally part of the
compiler or operating system, so standard libraries are OK.
Roger Espel Llima
In article <5um2u6$o4f$1...@dungeon.inka.de>,
Andreas Jellinghaus <a...@dungeon.inka.de> wrote:
>i still think, that troll people could do the best deal with this
>licence, and till now nobody showed me a way they could do better.
>(don't say "put it under lgpl", they still have to earn money).
so put it under GPL (not LGPL), *and* maintain their existing two
licenses (or at least the commercial one). since GPL software cannot be
combined and distributed with non-GPL, that would not take any advantage
off their commercial offer, and we'd have a truly free Qt for the free
software world.
Arnt Gulbrandsen
psm...@baynetworks.com (Paul D. Smith)
> The last time this thread came up, someone at Troll emailed me saying
> that RMS had in fact approved their usage, but they would not agree to
> show me his message to that effect until they checked with him. That
> was over a month ago but I haven't heard anything else. So perhaps this
> is already resolved, but I think Troll should put the FSF's statement up
> on their WWW site so all can see it.
I found the one message I've sent you (and archived), and nobody else
here can remember having written anything to you, but the message I
found doesn't resemble your description very much... so, I'll repeat
what it said, but hopefully more clearly:
- Qt is written entirely by Troll Tech AS. It does not contain any
GPL'd code.
- Therefore, Qt is not covered by the GPL, and nothing Troll Tech AS
does with respect to Qt can be in conflict with the GPL.
Btw, I did not claim to have been in contact with rms.
--Arnt
Roger Espel Llima
In article <m3zppsk...@lupinella.troll.no>,
Arnt Gulbrandsen <agu...@troll.no> wrote:
>> Either way, what would *you* lose from a GPL'd version of Qt?
>
>I used to answer that question and some others like it. I got tired
>of people ignoring the answers. I stopped. Sorry.
>
>For one thing, we believe that using the GPL would not let us recoup
>the heavy investments required to develop the library.
Even while keeping the commercial version, which the GPL allows you to
do without a problem?
Anyway, you've probably been through exactly these arguments quite a few
times, so we might as well close the discussion :)
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>> Not true, the GPL has an exception that covers libc, libm, etc. The FSF
>> has also ruled that it covers Motif, because Motif is a very commonly
>> supplied as a standard part of UNIX systems these days.
ra> But not on Linux, not on Windows sytems, not on FreeBSD, NetBSD,
ra> OpenBSD, not on the HURD, and is not part of the proposed GNU
ra> system the FSF so much likes to talk about.
As I said, it's very difficult to see how to apply this exception clause
in specific cases. But the FSF says that Motif passes, so... there you
go.
ra> Qt is available for no cost for almost every Unix variant, Motif
ra> isn't. I guess that alone would make Qt more acceptable than
ra> Motif as a platform for GPLed software.
Again, the FSF doesn't care about cost, in terms of $$. You can charge
as much as you can manage to convince someone to pay for any bit of
GPL'd software you like.
What they care about is _freedom_. And while Qt is a lot freer in some
respects than Motif, it's still not nearly as free (by the FSF's lights)
as a GPL'd library.
The question is not how free Qt is or isn't, the question is very basic:
is the Qt free license compatible or incompatible with the GPL?
If it's compatible, great. If it's not compatible, then GPL'd code
can't be combined with it. Now, if you're the copyright holder on some
bit of code and you want to put it under the GPL and you want to add a
clause that says Qt is OK, that's fine. Or you could just use a
completely different license than the GPL. It's your code, who cares?
We're talking here about combining general GPL'd code and Qt code,
neither of which you "own".
>> If Qt is actually not allowed under the exception, then it would be
>> illegal to distribute GPL'd code that links with Qt, since the licenses
>> would conflict (the same way as the GPL and BSD licenses conflict).
ra> Well, I wrote the code, and I distribute it, so I am the judge as to if
ra> Qt falls under the exception, not the FSF. It's *my* code.
If it's your code you don't even have to distribute it under the GPL at
all, so the entire argument is uninteresting.
>> Troll needs to contact the FSF directly and ask them. Only then will
>> Qt's status be completely clear.
ra> The FSF is not the owner of the GPL, nor its guardian, just the
ra> authors.
They are the owner and gaurdians of the GPL; the GPL is copyright too,
you know :). They aren't the owner of all code placed under the GPL
license, of course.
But they are the people most responsible for defending it in court. If
they do so, and the upshot is that your interpretation of the same legal
text (the GPL) is not legally valid, then that puts you in a weaker
position. For what, I don't know :)
ra> I don't believe a program that uses a pre-existant library somehow
ra> turns that library into part of a derived work.
>> Sure it does.
ra> I can't turn something I didn't write or own into a derived work!
ra> I didn't even touch it!
The combination of the two is a work derived from the original.
>> But more importantly for this argument, using the pre-existent
>> library turns the _program_ into a derived work.
ra> I don't agree with this either (perhaps I'm just stubborn :-) If
ra> the program doesn't include code from the library, it's not a
ra> derived work of the library.
Copyright law is often not nearly as clearcut as you might think. I'm
no lawyer, but I've heard stories... :)
ra> That distinction is actually made by the FSF in the LGPL. That's
ra> why linking to a LGPLed library does *not* make my app GPLed.
Exactly. But the LGPL is a completely different license. If the
distinction weren't real, as you claim, then there would be no need for
the FSF to have created the LGPL, since they could just use the GPL.
The FSF thinks the distinction _is_ relevant.
>> At least that's the FSF's claim. AFAIK it hasn't been tested in court
>> yet, but it's not at all obvious that they're wrong.
ra> It's not so obvious that they are right either ;-)
True. Wanna take 'em to court about it?
No? Well, then, you'll either have to take the most conservative
approach to avoid the possibility of legal action, or avoid using GPL'd
code, or walk the tight-rope and hope no one sues you.
>> If they _are_ wrong, then the GPL itself is for the most part moot.
ra> Well, every lawyer I have shown it to says that :-(
I seriously doubt you'll find _any_ lawyers who say that. If this were
true, people would be taking GPL'd code and imbedding it proprietary
apps right and left.
Perhaps you didn't fully understand my meaning.
Paul D. Smith
%% Arnt Gulbrandsen <agu...@troll.no> writes:
ag> - Qt is written entirely by Troll Tech AS. It does not contain any
ag> GPL'd code.
ag> - Therefore, Qt is not covered by the GPL, and nothing Troll Tech AS
ag> does with respect to Qt can be in conflict with the GPL.
No one has suggested anything of the sort. You are missing the entire
thrust of the argument.
No one is arguing that Troll doesn't have the right to distribute Qt
under any license it sees fit, including "something almost but not quite
GPL".
The _argument_ is over Troll's claim that it is perfectly legal to link
Qt with GPL'd code and release the entirety under the GPL. You claim it
is. I claim that that is not as clearcut as you seem to think.
I hope we can agree that it is not allowed to link GPL'd and BSD'd code
together, because the BSD license places extra restrictions that GPL
doesn't allow; the two licenses are incompatible and there's no way to
resolve them, so you can't combine them.
Right? I don't think there is much debate over this.
Now, Qt is -also- released under a license which adds restrictions that
are not compatible with the GPL. I hope we can also agree to that... it
seems quite obvious.
Yet, Troll claims that you are allowed to link GPL'd code and Qt code
together and distribute the bundle under the GPL.
As I understand it, the way Troll gets around the seeming impasse is
that they claim Qt falls under the "system component" exception of the
GPL, which is how libc, libm, and even Motif pass muster. The FSF has
published their position on Motif, that it complies with the exception.
As far as I know, they have published no such position on Qt.
I feel that the "system component" exception is murky at best. It uses
many terms which aren't clearly defined, and in the case of distributing
source which can be compiled on many platforms, some of which have
different sets of system components it's not clear at all how to apply
it.
If it's applied in its most broad fashion then the GPL has no force at
all. All you'd have to do is create one Linux or FreeBSD distribution
somewhere that contained any given bit of GPL'd code as "standard" and
voila! Now you can write code under any license you like and use that
GPL'd code all you want, and claim it's valid because of the exception.
I'm sure that's not the intent.
So, until this situation is clarified WRT Qt, by a statement from the
FSF, I think the claim by Troll that Qt can be distributed with GPL'd
code must remain suspect.
ag> Btw, I did not claim to have been in contact with rms.
It wasn't you, although I can't recall who it was. And it was via
private email, so there would be no record of it via Dejanews or
anything.
Further, the thread was on one of the comp.os.linux newsgroups, not the
gnu.misc.discuss newsgroups.
But, I could definitely have imagined it or misunderstood it.
Daniel R Risacher
>>>>> "ps" == Paul D Smith <psm...@baynetworks.com> writes:
ps> I hope we can agree that it is not allowed to link GPL'd and
ps> BSD'd code together, because the BSD license places extra
ps> restrictions that GPL doesn't allow; the two licenses are
ps> incompatible and there's no way to resolve them, so you can't
ps> combine them.
ps> Right? I don't think there is much debate over this.
I'm not sure this is clear, either. The linux kernel appears to
contains code that is under BSD license. (not held by BSD, mind you,
but under BSD-style terms)
As I understand, BSD license typically means: "do what you want, but
preserve my copyright, and my copyright notices".
Since the GPL also requires maintenance of copyright notices, where is
the conflict? Please educate me.
Dan Risacher
Roberto Alsina
In article <m2vi0fm...@stoli.spirits.org.au>, Nathan Hand wrote:
>ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
>> That several KDE developers are talking about changing their programs to
>> a BSD license. I certainly am starting to think about it.
>
>Is that a thinly veiled threat?
Of course not!
It's just that when people say the FSF has the final word on what I can
do with code I wrote, I start to think if the GPL is the right license.
>What next? People including
>
> This software may not be used by anyone I don't like, and that
> includes the following list of people...
I have actually seen this :-)
>Grow up! The reason there is a debate about the QT license is because
>some people are concerned about your rights, and the rights of others
>to freely use and propogate your software. It's hardly mature to just
>fob these people off when you get tired of the debate!
fob? I'm sorry, I don't know what that word means.
Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
500.000 lines of code, and put it under the GPL" and have people come and
say "You can't do that because you link with Qt".
Everybody has every right to freely use and propagate my software. That's
why I placed it under the GPL. Now, some people is saying I *can't* put
it under the GPL. How is that going to help using and propagating it?
The sources *I* wrote are there, anybody can use them, modify them, give
them away or sell them. What else am I supposed to do? Give my heart to
the vultures?
I mean, at my current fees, I have donated many thousand dollars of
software to the free software community. I am not so naive as to expect
Troll Tech to do the same thing. I think they are doing more than they
had to do, and what they do is more than enough.
>And just to add some fuel to the fire, don't use the BSD license! The
>BSD license explicitly permits people to take your code, and sell any
>modified versions without releasing code back as freeware. This helps
>commercial efforts, and it doesn't help free software.
Then what's the alternative? I should not make it GPL, I should not make
it BSD.... How can I make the people in this group happy? Go commercial?
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
ra> /usr/src/linux/drivers/net/sunlance.c
As I said, I think there's a special exception added to the license for
the Linux kernel that excepts drivers from the GPL restrictions. I
can't remember offhand, but it's all described in the kernel source
distribution so it's not hard to check (if I was at home...)
ra> You know what's the net outcome this far of all this Qt license
ra> talk? That several KDE developers are talking about changing
ra> their programs to a BSD license. I certainly am starting to think
ra> about it.
It's your code; release it however you like.
If you're happy with the BSD license, by all means use it. Or, add a
clause to the GPL for your program that explicitly allows it to be used
with Qt. Or use the Artistic License.
Or, *gasp*!, _ASK Stallman_! He doesn't read USENET, not even the gnu.*
hierarchy, so discussing it here won't get an answer. Email to
r...@gnu.ai.mit.edu. Once there's an official statement we can all hang
up our keyboards and go home.
I'll say it again: we're talking here about whether it is legal to
bundle pure GPL'd code with Qt. If you're the copyright holder you have
all kinds of options to avoid the question entirely. Use them.
Roberto Alsina
In article <m2afhrt...@stoli.spirits.org.au>, Nathan Hand wrote:
>ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
>> In article <m2vi0fm...@stoli.spirits.org.au>, Nathan Hand wrote:
>> >ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>> >
>> >> You know what's the net outcome this far of all this Qt license talk?
>> >> That several KDE developers are talking about changing their programs to
>> >> a BSD license. I certainly am starting to think about it.
>> >
>> >Is that a thinly veiled threat?
>>
>> Of course not!
>> It's just that when people say the FSF has the final word on what I can
>> do with code I wrote, I start to think if the GPL is the right license.
>
>and, more importantly, your code.
I'm a big boy, they don't have to ;-)
>> >What next? People including
>> >
>> > This software may not be used by anyone I don't like, and that
>> > includes the following list of people...
>>
>> I have actually seen this :-)
>
>Sheesh! :-(
>
>> >Grow up! The reason there is a debate about the QT license is because
>> >some people are concerned about your rights, and the rights of others
>> >to freely use and propogate your software. It's hardly mature to just
>> >fob these people off when you get tired of the debate!
>>
>> fob? I'm sorry, I don't know what that word means.
>
>Fobbing someone off is like dismissing them out of hand.
Oh! thanks. Another day, another word :-)
>> Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
>> 500.000 lines of code, and put it under the GPL" and have people come and
>> say "You can't do that because you link with Qt".
>
>Well it's true. Whether you like it or not. The law doesn't listen
>to you. You have to listen to, and follow, the law.
But the GPL is not a law. And the FSF has no word in how to read the GPL
either, that's the court's job. And the FSF has no saying about my code,
or Qt, or how my code relates to Qt, or how the GPL relates to my code
and Qt. They are just a 3rd party on this.
>And that includes making sure that any intellectualy copyright you
>make is *legal* and *binding*.
Legal it sure is. I can't make an illegal license agreement if I don't
take other people's code. And for the binding part, I have asked
copyright lawyers, and they are not really sure the GPL could be used in
court.
>Failure to do so is catastrophic. Maybe not for you but definitely
>for your code, and that is the *real* issue.
>
>> Everybody has every right to freely use and propagate my software. That's
>> why I placed it under the GPL. Now, some people is saying I *can't* put
>> it under the GPL. How is that going to help using and propagating it?
>
>You *shouldn't* because by doing so you create a legal paradox. QT
>and GPL licenses are at odds to each other.
Well, that's the whole point isn't it? I don't think they are. The only
reason I have been given is that linking with Qt "imposes aditional
restrictions to my code" and that's simply not true, since my code is
there, and you can do anything the GPL allows with it.
Consider this: FSF says linking with Motif is ok, because it's "usually
part of the system". Of what system? Qt comes with several Linux
distributions. Certainly with more than come with Motif. Doesn't that
make Qt fall under the same exception?
>There is a difference between the FSF's advice, and the FSF trying
>to exert control over you. They have only given advice, so far.
>
>I've never seen the FSF exert control over anyone, and I doubt you
>have ever had the FSF heavies bang on your door, either.
Of course not. I don't think they have "heavies", if you mean what I
think you mean :-)
>> The sources *I* wrote are there, anybody can use them, modify them, give
>> them away or sell them. What else am I supposed to do? Give my heart to
>> the vultures?
>
>The FSF isn't trying to get your heart so don't be so melodramatic
>about the whole thing.
Come on, it's just a touch of colour ;-)
I didn't mean it to be taken as the FSF are the vultures, just that this
is as far as my altruism goes.
>The FSF is simply trying to promote *free* software. Software that
>is of benefit to *everyone*, not just the select few.
So am I. Why not let me?
>And because this is such a complicated legal topic, they have gone
>to a lot of effort to make the really neat GPL licensing scheme.
I agree it's neat, I don't think an interpretation that forbids me
releasing my code is neat, nor helpful.
>So when you break the GPL by, for example, linking against QT, the
>FSF is going to tell you there are problems here.
I *can't* break the GPL, because I *own* the code. I am not even subject
to the GPL with respect to it.
>How is this in any way shape or form related to "giving your heart
>to the vultures"?
Read above for my intended meaning.
>> I mean, at my current fees, I have donated many thousand dollars of
>> software to the free software community.
>
>Well jolly good for you.
Thank you. I feel good doing it :-)
>> I am not so naive as to expect
>> Troll Tech to do the same thing. I think they are doing more than they
>> had to do, and what they do is more than enough.
>
>I agree! But it's still improper to make lies up about the FSF, or
>to claim they are "vultures".
I didn't mean to call them vultures. Perhaps my choice of words was not too
good. And where have I lied about the FSF??????
>The FSF has free software in interest, not your heart, and despite
>popular myth, RMS doesn't cackle evilly while destroying your life.
Hehe. You mean he doesn't fly at night to drink blood of unsuspecting
developers? Another myth falls ;-)
>> >And just to add some fuel to the fire, don't use the BSD license! The
>> >BSD license explicitly permits people to take your code, and sell any
>> >modified versions without releasing code back as freeware. This helps
>> >commercial efforts, and it doesn't help free software.
>>
>> Then what's the alternative? I should not make it GPL, I should not make
>> it BSD.... How can I make the people in this group happy? Go commercial?
>
>Recognise there's a problem with the QT license, as it stands, and
>stop being so defensive of it. That's all.
I don't think there is one. The only argument I have been given doesn't
make much sense (at least to me), and there are previous decisions by the
very own FSF (about Motif, for example) that would contradict that argument.
I guess we will have to agree to disagree.
>And a little less of the melodramatic "vulture" speak would not go
>astray, either. And while I'm at it...
Well, I am a little melodramatic in general. That's how I'm built :-)
><rant>
>What is it with people these days?! RMS and his band of do-gooders
>have worked damn hard, for over a decade, to develop free software
>of extraordinary quality. And now, just as people are reaping the
>rewards of the FSF's hard work, they're turning against the FSF in
>droves, claiming they are "vultures" and "meddlers" and "attention
>seekers". Are people so entirely stupid to think that we should do
>anything except thank RMS and his band of do-gooders? Wouldn't the
>tiniest bit of respect be due, eh?
></rant>
1) I didn't mean to call them vultures.
2) I respect the FSF
3) I like the GPL (why would I go through all this if I didn't?)
4) I thank them a lot.
PS: Thanks for calling me a liar and a stupid. I'll try to keep that in mind.
Wouter Scholten
Nathan Hand wrote:
>
> ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> > You know what's the net outcome this far of all this Qt license talk?
> > That several KDE developers are talking about changing their programs to
> > a BSD license. I certainly am starting to think about it.
>
> Is that a thinly veiled threat?
[snip]
> And just to add some fuel to the fire, don't use the BSD license! The
> BSD license explicitly permits people to take your code, and sell any
> modified versions without releasing code back as freeware. This helps
> commercial efforts, and it doesn't help free software.
No, but it would bring down the cost of producing commerial apps, which
in the end does help everyone.
Furthermore, LGPL lib's are also helping commercial developers (via less
specific
features, but still..)
Wouter
--
Windrugs: Just say no
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
ra> It's just that when people say the FSF has the final word on what
ra> I can do with code I wrote, I start to think if the GPL is the
ra> right license.
You continually miss the point. _You_ are stating what people can do
with the code you write, when you put the code under the GPL.
The GPL is a legal document and it has a particular legal meaning. We
are discussing here what that meaning is. If you don't agree with that
meaning, and you want your code to have a different set of licensing
restrictions, then don't put your code under the GPL. Or, put it under
the GPL but add some extra text at the bottom granting exceptions or
exclusions.
> It's hardly mature to just fob these people off when you get tired
> of the debate!
ra> fob? I'm sorry, I don't know what that word means.
http://www.m-w.com/cgi-bin/mweb?book=Dictionary&va=fob+off
ra> Anyway, it's not a maturity problem. It's just: "Hey, we wrote
ra> this nice 500.000 lines of code, and put it under the GPL" and
ra> have people come and say "You can't do that because you link with
ra> Qt".
Yes. I don't see the problem. It's a fact that if there are two bits
of code and they have different licensing terms, you may not be able to
use them together because those terms may conflict.
You can _wish_ that this weren't true, but that doesn't make it not
true.
Since you hold the copyright on one of the bits, your dilemma is
actually pointless. Simply change the licensing terms on the bit you
own so that you're sure it doesn't conflict with the other bit.
ra> The sources *I* wrote are there, anybody can use them, modify
ra> them, give them away or sell them. What else am I supposed to do?
ra> Give my heart to the vultures?
Use a license for your code that accurately represents what you want to
do with it. Obviously there's the possibility the GPL doesn't do this,
so don't use it or modify it.
The GPL is a document, it contains certain words in a certain order and
they have a certain meaning. When you apply those same words in that
same order to your code, they have the same meaning (surprise!) You
can't say ``Oh by the way, my license says "blue" but I really meant
"red"''.
I just don't see how else I can explain it.
Per Abrahamsen
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> And I *want* to make it GPL. I like the GPL!
Have you considered making your application LGPL instead? The LGPL
can also be used for applications, and it is specifically designed to
allow linking with non-free software (like Qt).
Roberto Alsina
In article <p5iuwf1...@baynetworks.com>, Paul D. Smith wrote:
>%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> ra> It's just that when people say the FSF has the final word on what
> ra> I can do with code I wrote, I start to think if the GPL is the
> ra> right license.
>
>You continually miss the point. _You_ are stating what people can do
>with the code you write, when you put the code under the GPL.
Yes...
>The GPL is a legal document and it has a particular legal meaning. We
>are discussing here what that meaning is. If you don't agree with that
>meaning, and you want your code to have a different set of licensing
>restrictions, then don't put your code under the GPL. Or, put it under
>the GPL but add some extra text at the bottom granting exceptions or
>exclusions.
>
> > It's hardly mature to just fob these people off when you get tired
> > of the debate!
>
> ra> fob? I'm sorry, I don't know what that word means.
>
>http://www.m-w.com/cgi-bin/mweb?book=Dictionary&va=fob+off
>
> ra> Anyway, it's not a maturity problem. It's just: "Hey, we wrote
> ra> this nice 500.000 lines of code, and put it under the GPL" and
> ra> have people come and say "You can't do that because you link with
> ra> Qt".
>
>Yes. I don't see the problem. It's a fact that if there are two bits
>of code and they have different licensing terms, you may not be able to
>use them together because those terms may conflict.
The fact that there are two independent bits of code makes it hard for me
to understand how the licensing on one affects the other, independent,
bit of code.
>You can _wish_ that this weren't true, but that doesn't make it not
>true.
You can say it's true, but that doesn't make it true either ;-)
>Since you hold the copyright on one of the bits, your dilemma is
>actually pointless. Simply change the licensing terms on the bit you
>own so that you're sure it doesn't conflict with the other bit.
But I believe it doesn't conflict. The basic point for me is that people
here is saying I am somehow breaking the GPL by putting code I am the
sole author of under it.
I have only received *one* reason for it, being that the Qt license
imposes an aditional restriction on my code. Of course nobody mentioned
precisely *what* restriction it imposes.
Imposibility of modifying Qt can't be one, since Qt is not under the GPL,
and is not part of my application, and therefore we don't have the right to
do so.
> ra> The sources *I* wrote are there, anybody can use them, modify
> ra> them, give them away or sell them. What else am I supposed to do?
> ra> Give my heart to the vultures?
>
>Use a license for your code that accurately represents what you want to
>do with it. Obviously there's the possibility the GPL doesn't do this,
>so don't use it or modify it.
I believe the GPL as it's written represents what I want to do. I don't
think some people's interpretation of it does.
>The GPL is a document, it contains certain words in a certain order and
>they have a certain meaning. When you apply those same words in that
>same order to your code, they have the same meaning (surprise!) You
>can't say ``Oh by the way, my license says "blue" but I really meant
>"red"''.
The problem is that the GPL says "it's red, unless in this case, which is
blue". I think my particular case falls in the blue zone. You think it
falls on the red part (or viceversa ;-).
>I just don't see how else I can explain it.
I just don't see anything explained.
I'll just ask this:
Tell me, in short words ;-) what exactly is the additional restriction
linking to Qt imposes on my code, which is released under the GPL, and
hasn't been previously granted by the FSF as acceptable?
Once I am told this, I will happily pack and leave the GPL land. I think
it should be obvious that I *do* like the GPL, and I *do* want to use it.
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>> Yes. I don't see the problem. It's a fact that if there are two bits
>> of code and they have different licensing terms, you may not be able to
>> use them together because those terms may conflict.
ra> The fact that there are two independent bits of code makes it hard
ra> for me to understand how the licensing on one affects the other,
ra> independent, bit of code.
I never said "independent", _you_ said "independent".
The fact is, the FSF's (and Troll Tech's!!) position is that they're not
independent, since one of them is useless (can't be compiled, linked, or
run) without the other.
>> It's a fact that if there are two bits of code and they have
>> different licensing terms, you may not be able to use them together
>> because those terms may conflict.
>> You can _wish_ that this weren't true, but that doesn't make it not
>> true.
ra> You can say it's true, but that doesn't make it true either ;-)
What I said is definitely true. Whether or not the terms actually
conflict in _this_ case, is another question.
ra> But I believe it doesn't conflict. The basic point for me is that
ra> people here is saying I am somehow breaking the GPL by putting
ra> code I am the sole author of under it.
You can put the code under any license you like. Why not put your code
under a shareware license which doesn't allow anyone to use it unless
they pay you $50?
What? You say you can't, because Qt's license doesn't let you? But
your code is completely independent of Qt, right, so why does it matter
what Qt's license says? Oh no! Troll Tech is trying to tell you what
you can do with your own code!
*grin*
ra> I have only received *one* reason for it, being that the Qt license
ra> imposes an aditional restriction on my code. Of course nobody mentioned
ra> precisely *what* restriction it imposes.
ra> Imposibility of modifying Qt can't be one, since Qt is not under
ra> the GPL, and is not part of my application, and therefore we don't
ra> have the right to do so.
Qt _IS_ part of your application, since your application is useless
without it.
You must be reading this on comp.os.linux.advocacy, since we just went
through this exact same thing last week on gnu.misc.discuss. I suggest
you visit DejaNews or something and review the arguments made here about
this exact subject; repitition is boring.
ra> Tell me, in short words ;-) what exactly is the additional
ra> restriction linking to Qt imposes on my code, which is released
ra> under the GPL, and hasn't been previously granted by the FSF as
ra> acceptable?
Qt is not free, in the GPL'd sense. Your code depends on Qt. The GPL
says that you can't combine GPL'd code with any other licensed code
unless you can distribute _the entirety_ under the GPL. That's want it
says. Go read it. You can't distribute Qt under the terms of the GPL,
so there you go. Impasse.
The only "escape hatch" is the system component exception. Troll Tech
claims that Qt falls under that exception, since there're some Linux
distributions that include it. Maybe they're right. But maybe they're
not.
The FSF has not granted anything about Qt specifically acceptable. As
much as you want to draw parallels between Motif and Qt, there are
significant differences. For example, Motif is available as part of the
system on a majority of "kinds" of UNIX systems available: HP-UX, AIX,
DG-UX, Irix, Solaris, etc. Qt is available as part of the system on
only one "kind" of UNIX system: Linux. And not even the majority of
variants of that.
Is that a valid distinction? It seems _reasonable_, anyway... who
knows? That's my point. The FSF has ruled on Motif. They haven't
ruled on Qt. There are enough differences between them that it's not a
no-brainer that one implies the other.
You can do anything you like. If you believe the GPL allows you to use
Qt under the system component exception, go ahead. I'm just pointing out
that it's not 100% sure that this is legal, so you do so at your own
risk. That risk is probably very small. But it's there.
Klaus Schilling
Nathan Hand <natha...@anu.edu.au> writes:
>
> ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> > You know what's the net outcome this far of all this Qt license talk?
> > That several KDE developers are talking about changing their programs to
> > a BSD license. I certainly am starting to think about it.
>
> Is that a thinly veiled threat?
>
> What next? People including
>
> This software may not be used by anyone I don't like, and that
> includes the following list of people...
The list : Billy Gate$, Billy Gate$, Billy Gate$,...
But free for anyone also means free for B.Gate$, besides I don't think
Billygoat would know what to do with something like Emacs in a constructive
way.
>
> Grow up! The reason there is a debate about the QT license is because
> some people are concerned about your rights, and the rights of others
> to freely use and propogate your software. It's hardly mature to just
> fob these people off when you get tired of the debate!
>
> And just to add some fuel to the fire, don't use the BSD license! The
> BSD license explicitly permits people to take your code, and sell any
> modified versions without releasing code back as freeware. This helps
> commercial efforts, and it doesn't help free software.
Yeah. I wouldn't use anything than GPL for my software ( If I get ever done
with coding my Fantasy-RPG for Linux-consoles anyways, let alone my yet
futile attempts to write certain modules for Guile to access less well-used
programming libs like ncurses). Software should be free!
Klaus Schilling the gnu-bull
Bruce Stephens
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> In article <m3pvqoa...@cenderis.demon.co.uk>, Bruce Stephens wrote:
> >There's an explicit exemption for things normally part of the
> >compiler or operating system, so standard libraries are OK.
>
> That part of the GPL contradicts the GPL itself, since it would mean a
> program is under the GPL in some systems and not in others.
No, a program is covered by the GPL. Whether it's usable on some
system or not may vary. There's no guarantee that there's a
compiler/interpreter for whatever language the program requires, for
example, and many GPLd programs require X, which will cause problems
for many Windows users. I don't see this as a problem. GPL was
written to maximise the freedoms people have in a certain direction
(whilst not helping people who want to make products which don't give
users these freedoms). I agree Motif is a bit of an oddity, but then
the FSF has always encouraged people not to use Motif, or to use one
of the free Motif-like toolkits, and Motif is standard on many
workstation OSs.
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
This is growing old, and no new ground is being covered...
ra> But the GPL is not a law.
The GPL is a legally binding contract (if you accept it--and if you
don't accept it, you can do very little with GPL'd code since it's
copyright).
ra> And the FSF has no word in how to read the GPL either, that's the
ra> court's job.
The FSF and their lawyers wrote the GPL. In the absence of a court
decision, _they_ are the best arbiters of what it means. Certainly
they're the best arbiters of what they _meant_ it to mean, and thus who
they may decide is violating it and take to court.
If they go to court and the GPL is ruled on, either for or against, you
can be damn sure that will also affect _your_ code, under the exact same
GPL license.
In the absence of a court decision, you ignore the FSF's opinion on what
the license means at your own peril, since they're the experts.
ra> And the FSF has no saying about my code, or Qt, or how my code
ra> relates to Qt, or how the GPL relates to my code and Qt. They are
ra> just a 3rd party on this.
True, but no one has claimed they do. As I said before, _YOU_ said how
your code is licensed when you used the GPL. The GPL means what it
means. If you put your code under the GPL, it still means the same
thing for your code as it does for everyone else's code under the GPL.
That includes the FSF's.
ra> Legal it sure is. I can't make an illegal license agreement if I don't
ra> take other people's code. And for the binding part, I have asked
ra> copyright lawyers, and they are not really sure the GPL could be used in
ra> court.
"Not really sure" is not at all the same as "definitely can't be".
You obviously _can_ ignore any license or law you like. If you do,
though, you run a risk. Whether or not it's worth it is a decision only
you can make.
ra> Well, that's the whole point isn't it? I don't think they are. The
ra> only reason I have been given is that linking with Qt "imposes
ra> aditional restrictions to my code" and that's simply not true,
ra> since my code is there, and you can do anything the GPL allows
ra> with it.
The additional restriction is that it is not totally free since it needs
Qt to work, and Qt is not totally free (in the GPL sense of things).
This has been discussed over and over already.
ra> Consider this: FSF says linking with Motif is ok, because it's
ra> "usually part of the system". Of what system? Qt comes with
ra> several Linux distributions. Certainly with more than come with
ra> Motif. Doesn't that make Qt fall under the same exception?
Exactly. Maybe it does; the exception clause is "exception"ally murky.
But then, why doesn't any bit of code anyone bothers to stick on a CD or
FTP site as part of some Linux distribution they made up count as an
exception?
That's why I've said a hundred times, "we don't know". You have to ask
the FSF. Or, ignore the issue and run the risk of legal action, however
small that might be. Or, not use or modify the GPL to remove the risk.
Roberto Alsina
In article <p5bu271...@baynetworks.com>, Paul D. Smith wrote:
>%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> >> of code and they have different licensing terms, you may not be able to
> >> use them together because those terms may conflict.
>
> ra> The fact that there are two independent bits of code makes it hard
> ra> for me to understand how the licensing on one affects the other,
> ra> independent, bit of code.
>
>I never said "independent", _you_ said "independent".
>
>The fact is, the FSF's (and Troll Tech's!!) position is that they're not
>independent, since one of them is useless (can't be compiled, linked, or
>run) without the other.
Well, I have to admit you are right on that these two pieces are not
independent in that sense. They are independent in the sense that they
have different copyright owners, and different licensing terms.
According to your deffinition of independent, Emacs on a platform that
doesn't have gnu binutils is dependent on that platform's ld, since you
can't link without it.
Not even the GPL demands this degree of independency, since it accepts
the "part of the system" clause.
I honestly believe you are demanding too much to declare something
"independent", but of course that's just me.
> >> It's a fact that if there are two bits of code and they have
> >> different licensing terms, you may not be able to use them together
> >> because those terms may conflict.
>
> >> You can _wish_ that this weren't true, but that doesn't make it not
> >> true.
>
> ra> You can say it's true, but that doesn't make it true either ;-)
>
>What I said is definitely true. Whether or not the terms actually
>conflict in _this_ case, is another question.
I meant "you can say it's true in this case", sorry.
> ra> But I believe it doesn't conflict. The basic point for me is that
> ra> people here is saying I am somehow breaking the GPL by putting
> ra> code I am the sole author of under it.
>
>You can put the code under any license you like. Why not put your code
>under a shareware license which doesn't allow anyone to use it unless
>they pay you $50?
>
>What? You say you can't, because Qt's license doesn't let you? But
>your code is completely independent of Qt, right, so why does it matter
>what Qt's license says? Oh no! Troll Tech is trying to tell you what
>you can do with your own code!
>
>*grin*
Low blow! ;-) The difference is that while I have agreed to Qt's license.
I haven't agreed to anybody's GPL on this particular case. I can't be
breaking a license I haven't agreed to.
> ra> I have only received *one* reason for it, being that the Qt license
> ra> imposes an aditional restriction on my code. Of course nobody mentioned
> ra> precisely *what* restriction it imposes.
>
> ra> Imposibility of modifying Qt can't be one, since Qt is not under
> ra> the GPL, and is not part of my application, and therefore we don't
> ra> have the right to do so.
>
>Qt _IS_ part of your application, since your application is useless
>without it.
In that case, most of the operating system is part of my application.
Actually gdbm should be part of my app too, since it's GPL (not LGPL).
This means I have to distribute gdbm too?
I have a distinct feeling that it doesn't make much sense. Unless somehow
the FSF has redefined the meaning of application, that is.
As an example: suppose my program calls system(), and uses a feature of
the korn shell. There are no free korn shells that have that particular
feature. Suddenly the korn shell is part of my application? (the korn
shell is not part of the OS I'm using, either).
Also, I *can* distribute non-functional or useless code under the GPL,
since it does not require that the program does anything at all. It
doesn't even require that the program compiles in anything other than one
given platform, and not even that, since the program is provided as is.
>You must be reading this on comp.os.linux.advocacy, since we just went
>through this exact same thing last week on gnu.misc.discuss.
I thought that you guys did that every week on gnu.misc.discuss? ;-)
> I suggest you visit DejaNews or something and review the arguments made
> here about this exact subject; repitition is boring.
>
> ra> Tell me, in short words ;-) what exactly is the additional
> ra> restriction linking to Qt imposes on my code, which is released
> ra> under the GPL, and hasn't been previously granted by the FSF as
> ra> acceptable?
>
>Qt is not free, in the GPL'd sense.
Right.
> Your code depends on Qt.
Right.
>The GPL says that you can't combine GPL'd code with any other licensed code
>unless you can distribute _the entirety_ under the GPL.
I am not combining the code. I am just using the API and linking to one
implementation of it.
I think you refer to
"You must cause any work that you distribute or publish, that in whole or
in part contains or is derived from the Program or any part thereof, to be
licensed as a whole at no charge to all third parties under the terms of
this License."
The program I publish contains no code whatsoever derived from any
program or part thereof that was published under the GPL, so I don't
believe this applies.
The only part that seems to apply is this:
"For an executable work, complete source code means all the source code
for all modules it contains, plus any associated interface definition files"
I guess this means the header files for Qt should be made available under
a license that permits its modification, but certainly it implies nothing
about the rest of Qt's source.
I'm fairly confident that if this is all, Troll Tech could accept it.
It would also mean, by the "part of the system" clause, that if those
headers are part of "the major components (compiler, kernel, and so on)
of the operating system on which the executable runs", I don't have to.
Now, this is tricky.
Qt is part of the system for at least some systems, but not all.
If I forbid building and running the Qt dependent program in systems
where Qt is not part of the system, I am contradicting the GPL,
specifically the part that says "... to be licensed ... to *all* third
parties under the terms of this license".
It also goes against the GPL in that the GPL only covers "only copying,
distribution and modification". It doesn't cover compiling, linking, or
running.
> That's want it says.
Yes.
>Go read it.
I have, doesn't it show? ;-).
> You can't distribute Qt under the terms of the GPL, so there you go. Impasse.
I can't, but I don't need to.
Qt is part of at least one system.
My program can then be released under the GPL for that platform.
Since the GPL doesn't cover anything related to the platform where my
program is compiled or runned, that's where the GPL's "full source code"
requirement doesn't imply I have to distribute Qt.
>The only "escape hatch" is the system component exception. Troll Tech
>claims that Qt falls under that exception, since there're some Linux
>distributions that include it. Maybe they're right. But maybe they're
>not.
>The FSF has not granted anything about Qt specifically acceptable. As
>much as you want to draw parallels between Motif and Qt, there are
>significant differences. For example, Motif is available as part of the
>system on a majority of "kinds" of UNIX systems available: HP-UX, AIX,
>DG-UX, Irix, Solaris, etc. Qt is available as part of the system on
>only one "kind" of UNIX system: Linux. And not even the majority of
>variants of that.
That is true now. A few years ago, SunOs still didn't ship with Motif,
and it accounted for quite a big part of UNIX sales.
The GNU system does not ship (or even include in their specification) Motif.
Most commercial Unixes ship with Motif 1.2. If I require Motif 2.0 does
that leave me out?
All windows and Macs do not include Motif.
Why make artificial distinctions between UNIX and non UNIX?
Motif is part of only a tiny fraction of systems.
>Is that a valid distinction? It seems _reasonable_, anyway... who
>knows? That's my point. The FSF has ruled on Motif. They haven't
>ruled on Qt. There are enough differences between them that it's not a
>no-brainer that one implies the other.
>
>You can do anything you like. If you believe the GPL allows you to use
>Qt under the system component exception, go ahead. I'm just pointing out
>that it's not 100% sure that this is legal, so you do so at your own
>risk. That risk is probably very small. But it's there.
Well, neither of us are lawyers, I guess. And then again, nothing is 100%
legal, if a lawyer thinks he can get a few bucks out of it ;-)
Bruce Stephens
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
> 500.000 lines of code, and put it under the GPL" and have people come and
> say "You can't do that because you link with Qt".
You can certainly do this: Qt explicitly allows you to do it, and
otherwise you're the copyright owner. The problem is that it's not
entirely clear that anybody else can distribute your code. On the
other hand, it is a bit nitpicking, since you're making it clear what
your intent is, so presumably you aren't going to chase people if they
act according to your wishes. Indeed, I'd like to think courts would
take a dim view of you trying to take such action.
Roberto Alsina
Oh, but you can't use ncurses either!
You see, ncurses is not part of the system on most unixes, just like Qt.
So you will have to distribute ncurses with your program, and under the
terms of the GPL.
Since ncurses demands you to that "this notice ... is not removed from any
of its header files", you can't modify it in all the extent the GPL
allows you (the GPL says you can replace copyright notices with another
"conspicuous and appropiate copyright notice").
Just kidding (I hope ;-)
Isaac
In article <m2afhrt...@stoli.spirits.org.au>, Nathan Hand wrote:
>> Everybody has every right to freely use and propagate my software. That's
>> why I placed it under the GPL. Now, some people is saying I *can't* put
>> it under the GPL. How is that going to help using and propagating it?
>
>You *shouldn't* because by doing so you create a legal paradox. QT
>and GPL licenses are at odds to each other.
>
>to exert control over you. They have only given advice, so far.
I'm not familiar with the code in question, but if the poster is
talking about their own code and not the FSF's or anyone else's then
the license means whatever the author says it means as long as it's
clear to the users. If there are any ambiguities, the author can
clear it up with some amplifying information.
Of course if he combines it with other GPL'd code he'll have to
respect the wishes of the other authors.
Isaac
Tim Smith
Paul D. Smith <psm...@baynetworks.com> wrote:
>Or, *gasp*!, _ASK Stallman_! He doesn't read USENET, not even the gnu.*
>hierarchy, so discussing it here won't get an answer. Email to
>r...@gnu.ai.mit.edu. Once there's an official statement we can all hang
>up our keyboards and go home.
No, we can't hang up our keyboards then, because RMS's official statement
only tells you what will happen if you use his or FSF's code. There are
other people who have released GPL'ed code, and they might not agree with
RMS's ruling. You've got to ask everyone whose GPL'ed code you plan to
use.
--Tim Smith
Tim Smith
Vlad Petersen <vlad...@compuserve.com> wrote:
>> If you want to GPL your software, put it under the FreeBSD license
>> or put it in the public domain etc, you can do so without paying a
>> single cent, and you can sell as much of it for as much as you want.
>
>I don't understand that. If I want to GPL my software, I put it
>under GPL. If I want to license my software under BSD terms, I
>use BSD license. Please don't mix those in one bunch.
You read that wrong. It's like this:
If ( (you want to GPL your software) ||
(you want to put it under the FreeBSD license) ||
(you want to put it in the public domain) )
{
Troll Tech won't charge you;
Troll Tech won't prevent you from charging;
}
You seem to have read it as
If ( you want to GPL your software )
put it under the FreeBSD license;
...
--Tim Smith
Roberto Alsina
In article <3410E407...@compuserve.com>, Vlad Petersen wrote:
>Roberto Alsina wrote:
>..........
>> Consider this: FSF says linking with Motif is ok, because it's "usually
>> part of the system". Of what system? Qt comes with several Linux
>> distributions.
>
>Not with RedHat - that's what I am using so I know. Not with
>Debian - at least not officially: it is somewhere under .../non-free/
>directory. Don't know about Slackware but having taken a quick
>glance through an Infomagic CD bundle, I haven't located any QT
>stuff.
It is in Debian's CD, it's on Red Hat's web site. It comes with SUSE, it
comes with McAfee's WebShield, it will soon come with others.
>> Certainly with more than come with Motif. Doesn't that
>> make Qt fall under the same exception?
>
>Not really: think SCO, Solaris, AIX, IRIX and other commercial
>guys. Their market share is bigger then that of Linux, if they come
>with X Window System, they all come with Motif and even CDE
>nowadays.
Think Mac, think Windows, think *BSD, and other commercial guys. Their
market share is bigger then that of what you mention, some don't come
with X Window System, none of them comes with Motif or CDE.
>.......
>> >The FSF is simply trying to promote *free* software. Software that
>> >is of benefit to *everyone*, not just the select few.
>>
>> So am I. Why not let me?
>
>Because you're promoting a software which is semi-free. :) The
>benefits of it are doubtful.
So we should not promote semi-free software. Perhaps hide it under the
carpet? ;-)
>> >> Then what's the alternative? I should not make it GPL, I should not make
>> >> it BSD.... How can I make the people in this group happy? Go commercial?
>
>Use other library?
Too late. Counting the things that are not available on other libraries,
and the extra code to overcome the lack of the signal/slot system, I
would have to port about 20.000 lines of code. I'm too tired for that.
If we consider all of KDE, it's over 500.000 lines of code. That's a lot.
Literally a pile of money.
>> ><rant>
>> >What is it with people these days?! RMS and his band of do-gooders
>> >have worked damn hard, for over a decade, to develop free software
>> >of extraordinary quality. And now, just as people are reaping the
>> >rewards of the FSF's hard work, they're turning against the FSF in
>> >droves, claiming they are "vultures" and "meddlers" and "attention
>> >seekers". Are people so entirely stupid to think that we should do
>> >anything except thank RMS and his band of do-gooders? Wouldn't the
>> >tiniest bit of respect be due, eh?
>> ></rant>
>>
>> 1) I didn't mean to call them vultures.
>> 2) I respect the FSF
>> 3) I like the GPL (why would I go through all this if I didn't?)
>> 4) I thank them a lot.
>
>[carnivorous snipped]
Carnivorous? The part where I thank the previous poster for calling me
stupid and a liar? That's not carnivorous ;-)
Vlad Petersen
Roberto Alsina wrote:
..........
> Consider this: FSF says linking with Motif is ok, because it's "usually
> part of the system". Of what system? Qt comes with several Linux
> distributions.
Does it come with all The Big Three: Debian, RedHat, Slackware?
Not with RedHat - that's what I am using so I know. Not with
Debian - at least not officially: it is somewhere under .../non-free/
directory. Don't know about Slackware but having taken a quick
glance through an Infomagic CD bundle, I haven't located any QT
stuff.
> Certainly with more than come with Motif. Doesn't that
> make Qt fall under the same exception?
Not really: think SCO, Solaris, AIX, IRIX and other commercial
guys. Their market share is bigger then that of Linux, if they come
with X Window System, they all come with Motif and even CDE
nowadays.
.......
> >The FSF is simply trying to promote *free* software. Software that
> >is of benefit to *everyone*, not just the select few.
>
> So am I. Why not let me?
Because you're promoting a software which is semi-free. :) The
benefits of it are doubtful.
........
> >> Then what's the alternative? I should not make it GPL, I should not make
> >> it BSD.... How can I make the people in this group happy? Go commercial?
Use other library?
> ><rant>
> >What is it with people these days?! RMS and his band of do-gooders
> >have worked damn hard, for over a decade, to develop free software
> >of extraordinary quality. And now, just as people are reaping the
> >rewards of the FSF's hard work, they're turning against the FSF in
> >droves, claiming they are "vultures" and "meddlers" and "attention
> >seekers". Are people so entirely stupid to think that we should do
> >anything except thank RMS and his band of do-gooders? Wouldn't the
> >tiniest bit of respect be due, eh?
> ></rant>
>
> 1) I didn't mean to call them vultures.
> 2) I respect the FSF
> 3) I like the GPL (why would I go through all this if I didn't?)
> 4) I thank them a lot.
[carnivorous snipped]
--
<vladimip at uniserve com> | Good pings come in small packets
Vancouver, BC | SIGSIG: signature too long (core dumped)
Things are more like they are today than they ever were before -
(From a crosspost between alt.suicide and alt.destroy.microsoft)
Nathan Hand
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> You know what's the net outcome this far of all this Qt license talk?
> That several KDE developers are talking about changing their programs to
> a BSD license. I certainly am starting to think about it.
Is that a thinly veiled threat?
What next? People including
This software may not be used by anyone I don't like, and that
includes the following list of people...
Grow up! The reason there is a debate about the QT license is because
some people are concerned about your rights, and the rights of others
to freely use and propogate your software. It's hardly mature to just
fob these people off when you get tired of the debate!
And just to add some fuel to the fire, don't use the BSD license! The
BSD license explicitly permits people to take your code, and sell any
modified versions without releasing code back as freeware. This helps
commercial efforts, and it doesn't help free software.
--
The sticker on the side of the box said "Supported Platforms: Windows 95,
Windows NT 4.0, or better", so clearly Linux was a supported platform.
Nathan Hand
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> In article <m2vi0fm...@stoli.spirits.org.au>, Nathan Hand wrote:
> >ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> >
> >> You know what's the net outcome this far of all this Qt license talk?
> >> That several KDE developers are talking about changing their programs to
> >> a BSD license. I certainly am starting to think about it.
> >
> >Is that a thinly veiled threat?
>
> Of course not!
> It's just that when people say the FSF has the final word on what I can
> do with code I wrote, I start to think if the GPL is the right license.
The FSF isn't trying to control you. They're trying to protect you
and, more importantly, your code.
> >What next? People including
> >
> > This software may not be used by anyone I don't like, and that
> > includes the following list of people...
>
> I have actually seen this :-)
Sheesh! :-(
> >Grow up! The reason there is a debate about the QT license is because
> >some people are concerned about your rights, and the rights of others
> >to freely use and propogate your software. It's hardly mature to just
> >fob these people off when you get tired of the debate!
>
> fob? I'm sorry, I don't know what that word means.
Fobbing someone off is like dismissing them out of hand.
> Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
> 500.000 lines of code, and put it under the GPL" and have people come and
> say "You can't do that because you link with Qt".
Well it's true. Whether you like it or not. The law doesn't listen
to you. You have to listen to, and follow, the law.
And that includes making sure that any intellectualy copyright you
make is *legal* and *binding*.
Failure to do so is catastrophic. Maybe not for you but definitely
for your code, and that is the *real* issue.
> Everybody has every right to freely use and propagate my software. That's
> why I placed it under the GPL. Now, some people is saying I *can't* put
> it under the GPL. How is that going to help using and propagating it?
You *shouldn't* because by doing so you create a legal paradox. QT
and GPL licenses are at odds to each other.
There is a difference between the FSF's advice, and the FSF trying
to exert control over you. They have only given advice, so far.
I've never seen the FSF exert control over anyone, and I doubt you
have ever had the FSF heavies bang on your door, either.
> The sources *I* wrote are there, anybody can use them, modify them, give
> them away or sell them. What else am I supposed to do? Give my heart to
> the vultures?
The FSF isn't trying to get your heart so don't be so melodramatic
about the whole thing.
The FSF is simply trying to promote *free* software. Software that
is of benefit to *everyone*, not just the select few.
And because this is such a complicated legal topic, they have gone
to a lot of effort to make the really neat GPL licensing scheme.
So when you break the GPL by, for example, linking against QT, the
FSF is going to tell you there are problems here.
How is this in any way shape or form related to "giving your heart
to the vultures"?
> I mean, at my current fees, I have donated many thousand dollars of
> software to the free software community.
Well jolly good for you.
> I am not so naive as to expect
> Troll Tech to do the same thing. I think they are doing more than they
> had to do, and what they do is more than enough.
I agree! But it's still improper to make lies up about the FSF, or
to claim they are "vultures".
The FSF has free software in interest, not your heart, and despite
popular myth, RMS doesn't cackle evilly while destroying your life.
> >And just to add some fuel to the fire, don't use the BSD license! The
> >BSD license explicitly permits people to take your code, and sell any
> >modified versions without releasing code back as freeware. This helps
> >commercial efforts, and it doesn't help free software.
>
> Then what's the alternative? I should not make it GPL, I should not make
> it BSD.... How can I make the people in this group happy? Go commercial?
Recognise there's a problem with the QT license, as it stands, and
stop being so defensive of it. That's all.
And a little less of the melodramatic "vulture" speak would not go
astray, either. And while I'm at it...
<rant>
What is it with people these days?! RMS and his band of do-gooders
have worked damn hard, for over a decade, to develop free software
of extraordinary quality. And now, just as people are reaping the
rewards of the FSF's hard work, they're turning against the FSF in
droves, claiming they are "vultures" and "meddlers" and "attention
seekers". Are people so entirely stupid to think that we should do
anything except thank RMS and his band of do-gooders? Wouldn't the
tiniest bit of respect be due, eh?
</rant>
--
Roger Espel Llima
In article <p5bu271...@baynetworks.com>,
>Qt under the system component exception, go ahead. I'm just pointing out
>that it's not 100% sure that this is legal, so you do so at your own
>risk. That risk is probably very small. But it's there.
At your own risk _from whom_ specifically? It's your code, so *you* are
certainly not going to get sued for distributing it under whichever
license you wish, including the GPL. And you are obviously not going
to sue someone else for distributing your own GPL software which links
to Qt, because you did it yourself to begin with and you'd look
ridiculous in court, and you woulnd't want to anyway. So where's the
problem, in practice?
Klaus Schilling
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
>
> Oh, but you can't use ncurses either!
>
> You see, ncurses is not part of the system on most unixes, just like Qt.
> So you will have to distribute ncurses with your program, and under the
> terms of the GPL.
>
> Since ncurses demands you to that "this notice ... is not removed from any
> of its header files", you can't modify it in all the extent the GPL
> allows you (the GPL says you can replace copyright notices with another
> "conspicuous and appropiate copyright notice").
>
> Just kidding (I hope ;-)
>
> --
Ncurses is possibly put under LGPL somewhen next year and then maintained
by Florian Laroche, the copyright being shifted to the FSF. At least that's
what Stallman told to the ncurses-mailing-list a few months ago.
Besides there are GPL-programs that link against ncurses, for example Lynx ,
Midnight Commander, Emacs, Oleo, dialog and Minicom.
Some of them can use alternatively other libs than ncurses, for example
slangcurses (in the case of mid.com. and lynx) or a custom termcap-based
feature (emacs) instead, but Dialog doesn't intend to be linked with slang.
Klaus Schilling
Hrvoje Niksic
es...@solar.sky.net (Roger Espel Llima) writes:
> In article <p5bu271...@baynetworks.com>,
> Paul D. Smith <psm...@baynetworks.com> wrote:
> >You can do anything you like. If you believe the GPL allows you to use
> >Qt under the system component exception, go ahead. I'm just pointing out
> >that it's not 100% sure that this is legal, so you do so at your own
> >risk. That risk is probably very small. But it's there.
>
> At your own risk _from whom_ specifically?
From people whose GPL-ed code you might use. For instance, in your
GPL-ed code you may use lots of other people's GPL-ed code (which is
one of the main points of the GPL). If these other people consider
your interpretation of the GPL invalid, they might conceivably take
action.
If all the code in the program is yours only, then the risk really is
negligible. However, most GPL-ed programs share their code with other
such software.
--
Hrvoje Niksic <hni...@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
I'm sure they'll listen to reason! -- Neal Stevenson, _Snow Crash_
Nathan Hand
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> In article <m2afhrt...@stoli.spirits.org.au>, Nathan Hand wrote:
> >ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> >
> >> It's just that when people say the FSF has the final word on what I can
> >> do with code I wrote, I start to think if the GPL is the right license.
> >
> >The FSF isn't trying to control you. They're trying to protect you
> >and, more importantly, your code.
>
> I'm a big boy, they don't have to ;-)
Nonono, it's your code that needs the protection, as I just said.
Why would you think the FSF cares about you :-)
> >> Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
> >> 500.000 lines of code, and put it under the GPL" and have people come and
> >> say "You can't do that because you link with Qt".
> >
> >Well it's true. Whether you like it or not. The law doesn't listen
> >to you. You have to listen to, and follow, the law.
>
> But the GPL is not a law. And the FSF has no word in how to read the GPL
> either, that's the court's job. And the FSF has no saying about my code,
> or Qt, or how my code relates to Qt, or how the GPL relates to my code
> and Qt. They are just a 3rd party on this.
But the GPL is a license, and therefore is subjected to the law. And
in the absence of anyone who actually understands the GPL, you would
be pretty safe in betting that the FSF gives you better advice, than
say the kid down the road, or your friend on the net.
> >And that includes making sure that any intellectualy copyright you
> >make is *legal* and *binding*.
>
> Legal it sure is. I can't make an illegal license agreement if I don't
> take other people's code. And for the binding part, I have asked
> copyright lawyers, and they are not really sure the GPL could be used in
> court.
You should tell the FSF this! They are depending on it holding up in
court, if it ever came to that.
> [...]
>
> ><rant>
> >What is it with people these days?! RMS and his band of do-gooders
> >have worked damn hard, for over a decade, to develop free software
> >of extraordinary quality. And now, just as people are reaping the
> >rewards of the FSF's hard work, they're turning against the FSF in
> >droves, claiming they are "vultures" and "meddlers" and "attention
> >seekers". Are people so entirely stupid to think that we should do
> >anything except thank RMS and his band of do-gooders? Wouldn't the
> >tiniest bit of respect be due, eh?
> ></rant>
>
> 1) I didn't mean to call them vultures.
> 2) I respect the FSF
> 3) I like the GPL (why would I go through all this if I didn't?)
> 4) I thank them a lot.
Okie, then I guess we're both saying pretty much the same thing. And
as I know I've repeated myself at least twice already in this thread
I guess there's not much more to say on this matter. I think the FSF
was doing OK by pointing out the problems. You're free to disagree.
> PS: Thanks for calling me a liar and a stupid. I'll try to keep that in mind.
If I said either of these things, I didn't mean to, so please do not
take offence. I hardly know you well enough to make either claim!
k...@hub.cs.umb.edu
Suppose someone distributes an executable made by linking Qt and GPL'd
source. If people who get that executable can
(a) get the source for it (all of it, including Qt); and
(b) redistribute this source and/or the binary; and
(c) modify the source (all of it); and
(d) redistribute their modifications under the same terms (thus allowing
others to acquire/modify/redistribute recursively); and
(e) allow money to be charged for the act of redistribution,
then I doubt that the FSF or other GPL authors would object to the Qt
license. If the Qt license does not permit these things (and I gather
it does not, or there wouldn't be any debate), then I think it is very
likely the FSF would object, as it would seem to contravene the spirit
of the GPL.
But of course I do not speak for the FSF or anyone else but myself :-).
Nathan Hand
t...@halcyon.com (Tim Smith) writes:
Now wouldn't it be great if everyone could structure their posts in
a sensible way like this :-)
Nice and clear and easy to read. Perhaps we should do away with the
Queen's English, and start using C-English :-)
Per Abrahamsen
ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> Too late. Counting the things that are not available on other libraries,
> and the extra code to overcome the lack of the signal/slot system, I
> would have to port about 20.000 lines of code. I'm too tired for that.
Put your application under the LGPL instead of the GPL, and there will
be no longer be any legal fear, uncertainty, or doubt with regard to
the use of Qt.
Vlad Petersen
Roberto Alsina wrote:
........
> >
> >[carnivorous snipped]
>
> Carnivorous? The part where I thank the previous poster for calling me
> stupid and a liar? That's not carnivorous ;-)
No, I meant the animal in your sig. :)
Klaus Schilling
Nathan Hand <natha...@anu.edu.au> writes:
>
> t...@halcyon.com (Tim Smith) writes:
>
> > Vlad Petersen <vlad...@compuserve.com> wrote:
> > >> If you want to GPL your software, put it under the FreeBSD license
> > >> or put it in the public domain etc, you can do so without paying a
> > >> single cent, and you can sell as much of it for as much as you want.
> > >
> > >I don't understand that. If I want to GPL my software, I put it
> > >under GPL. If I want to license my software under BSD terms, I
> > >use BSD license. Please don't mix those in one bunch.
> >
> > You read that wrong. It's like this:
> >
> > If ( (you want to GPL your software) ||
> > (you want to put it under the FreeBSD license) ||
And what is with L.Wall's Artistic License?
> > (you want to put it in the public domain) )
> > {
> > Troll Tech won't charge you;
> > Troll Tech won't prevent you from charging;
> > }
Trolls are a myth anyways.
> >
> > You seem to have read it as
> >
> > If ( you want to GPL your software )
> > put it under the FreeBSD license;
> > ...
>
> Now wouldn't it be great if everyone could structure their posts in
> a sensible way like this :-)
>
> Nice and clear and easy to read. Perhaps we should do away with the
> Queen's English, and start using C-English :-)
Why not Guile-Scheme-English, or Perl-English, or Emacs-Lisp-English?
Or Python-English ? (Or is that reserved to fans of Monty Python???)
And embed an interpreter of the language into your newsreader, and let
it evaluate the code on site.
Klaus Schilling
Marinos J. Yannikos
>Grow up! The reason there is a debate about the QT license is because
>some people are concerned about your rights, and the rights of others
>to freely use and propogate your software. It's hardly mature to just
>fob these people off when you get tired of the debate!
It's fine to be concerned about a programmer's rights, but dealing with
the rights or non-rights of others to "freely use and propogate"
someone's software is that person's business only, unless she/he chooses
to employ others with that task.
>And just to add some fuel to the fire, don't use the BSD license! The
>BSD license explicitly permits people to take your code, and sell any
>modified versions without releasing code back as freeware. This helps
>commercial efforts, and it doesn't help free software.
So why do you think you may know better what the programmer wants than
him/herself? Perhaps someone prefers true, honest "freedom" in licensing,
including the right to make money with someone else's efforts, rather than
the kind of "freedom" you get when you want to include GPL'd code in a large
project and end up having to re-write that code just to be able to sell
yours.
Some people may think that there are more important characteristics of a
program than whether it is free or not, like e.g. quality, while others
seem to just prefer any free software over any commercial counterpart.
-nino
--
Please change the last part of my address to "at" if you're replying by mail.
It's nice to be important, but it's more important to be nice.
Nathan Hand
ni...@complang.tuwien.ac.look-in-sig (Marinos J. Yannikos) writes:
> In article <m2vi0fm...@stoli.spirits.org.au>, Nathan Hand wrote:
> >[...]
> >Grow up! The reason there is a debate about the QT license is because
> >some people are concerned about your rights, and the rights of others
> >to freely use and propogate your software. It's hardly mature to just
> >fob these people off when you get tired of the debate!
>
> It's fine to be concerned about a programmer's rights, but dealing with
> the rights or non-rights of others to "freely use and propogate"
> someone's software is that person's business only, unless she/he chooses
> to employ others with that task.
Of course it is. And when the person makes a legal mistake, it does
make a lot of sense to tell them about it!
Do you prefer a world where people don't tell other people of their
mistakes, perhaps for fear of hurting their feelings?
> >And just to add some fuel to the fire, don't use the BSD license! The
> >BSD license explicitly permits people to take your code, and sell any
> >modified versions without releasing code back as freeware. This helps
> >commercial efforts, and it doesn't help free software.
>
> So why do you think you may know better what the programmer wants than
> him/herself?
Where did I claim I did?
Here, I'm going to write some words, and I'd prefer it if you stick
the ones *I* write in my mouth, here goes...
> Perhaps someone prefers true, honest "freedom" in licensing,
> including the right to make money with someone else's efforts, rather than
> the kind of "freedom" you get when you want to include GPL'd code in a large
> project and end up having to re-write that code just to be able to sell
> yours.
What does this have to do with what I wrote? I just pointed out the
BSD license doesn't help to keep software free.
I'm not exactly hounding the author to not use the BSD license, now
am I? It's just some advice.
Sheesh. Loosen up.
> Some people may think that there are more important characteristics of a
> program than whether it is free or not, like e.g. quality, while others
> seem to just prefer any free software over any commercial counterpart.
And some people apparently prefer to read what they want to read. I
prefer to read what is actually written.
Roberto Alsina
On Sun, 7 Sep 1997, Anand Kumria wrote:
> In article <slrn6107gb....@ultra7.unl.edu.ar> you wrote:
> : In article <m2vi0fm...@stoli.spirits.org.au>, Nathan Hand wrote:
> : >ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
> : >
> : >> You know what's the net outcome this far of all this Qt license talk?
> : >> That several KDE developers are talking about changing their programs to
> : >> a BSD license. I certainly am starting to think about it.
> : >
> : >Is that a thinly veiled threat?
> :
> : Of course not!
> : It's just that when people say the FSF has the final word on what I can
> : do with code I wrote, I start to think if the GPL is the right license.
>
> Irrespective of what license you put your code, as the copyright holder
> you are the only person who decideds what can and can't be done with your work.
Well, not exactly. After I license it to someone, then it's a mattero of
legal interpretation of the license, but you are basically right on this
case.
>
> : Anyway, it's not a maturity problem. It's just: "Hey, we wrote this nice
> : 500.000 lines of code, and put it under the GPL" and have people come and
> : say "You can't do that because you link with Qt".
>
> People who are telling you that are wrong. If you want definitive advise why not
> ask RMS or speak to a lwayer.
>
> : Then what's the alternative? I should not make it GPL, I should not make
> : it BSD.... How can I make the people in this group happy? Go commercial?
>
> The think to do is to ignore the people who don't know what they are on abuot,
> and get advise from those who do.
Ok, thanks a lot. I think I'll just let this thread die :-)
Isaac
In article <axzppq9...@hub.cs.umb.edu>, k...@hub.cs.umb.edu wrote:
[ snipped list of conditions required by the GPL ]
>then I doubt that the FSF or other GPL authors would object to the Qt
>license. If the Qt license does not permit these things (and I gather
>it does not, or there wouldn't be any debate), then I think it is very
>likely the FSF would object, as it would seem to contravene the spirit
>of the GPL.
>
>But of course I do not speak for the FSF or anyone else but myself :-).
>
I think your post was accurate but does not speak directly to the issue
under discussion. In the particular instance at hand, the FSF's code
was not involved and the only GPL author was the originator of the code.
The FSF's only potential complaint would be potential misunderstanding of
the license concerning their own code. I don't see how that gives them
any standing to pursue suing someone. They can copyright the license but
it's legal interpretation has to include the author's intent (which is
quite clear).
Isaac
Isaac
In article <kigyb5a...@jagor.srce.hr>, Hrvoje Niksic wrote:
>
>If all the code in the program is yours only, then the risk really is
>negligible. However, most GPL-ed programs share their code with other
>such software.
>
The risk in this case is not negligible, but zero! There's literially
no involved party to sue you.
Isaac
Hrvoje Niksic
Is...@yellow.submarine.pla (Isaac) writes:
If the authors of GPL-ed programs consider that you use their code in
a program not distributed by the GPL, they might sue.
--
Hrvoje Niksic <hni...@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
"Silence!" cries Freydag. "I did not call thee in for a consultation!"
"They are my innards! I will not have them misread by a poseur!"
Isaac
In article <kigzppp...@jagor.srce.hr>, Hrvoje Niksic wrote:
>
>If the authors of GPL-ed programs consider that you use their code in
>a program not distributed by the GPL, they might sue.
>
>--
Where does the assumption come from that all GPL'd code belongs to
someone else?
Now I agree that even if there are other authors, there is only a
small motivation to pursue the case, but we're talking about whether
someone can apply the GPL to code they wrote and not stuff they
downloaded from prep. I suspect this point is missed because
once it's understood, the whole argument seems silly.
Isaac
Hrvoje Niksic
Is...@yellow.submarine.pla (Isaac) writes:
> You missed the point that in this case there are no other authors.
> Where does the assumption come from that all GPL'd code belongs to
> someone else?
It comes from my message you answered to.
> In article <kigyb5a...@jagor.srce.hr>, Hrvoje Niksic wrote:
me:
> >If all the code in the program is yours only, then the risk really is
> >negligible. However, most GPL-ed programs share their code with other
> >such software.
you:
> The risk in this case is not negligible, but zero! There's literially
> no involved party to sue you.
> Now I agree that even if there are other authors, there is only a
> small motivation to pursue the case, (...)
Then we agree.
--
Hrvoje Niksic <hni...@srce.hr> | Student at FER Zagreb, Croatia
--------------------------------+--------------------------------
"Psychos _do not_ explode when sunlight hits them."
Eugene O'Neil
In article <p5bu271...@baynetworks.com>, psm...@baynetworks.com wrote:
>Qt is not free, in the GPL'd sense. Your code depends on Qt. The GPL
>says that you can't combine GPL'd code with any other licensed code
>unless you can distribute _the entirety_ under the GPL. That's want it
>says. Go read it. You can't distribute Qt under the terms of the GPL,
>so there you go. Impasse.
Huh! Okay, I didn't know that. You learn something new every day...
>The only "escape hatch" is the system component exception. Troll Tech
>claims that Qt falls under that exception, since there're some Linux
>distributions that include it. Maybe they're right. But maybe they're
>not.
System component on WHAT system? I suppose Motif got by because it was a system
component on someone's machine, but it isn't a system component of most Linux
distributions. For that matter, it is certianly not available in Windows, which
can hardly be described as uncommon platform. :-\
If a library does not have to be on every system or even most systems to be a
system component, then it seems to me that making any library into a system
component is a simple matter of creating a system in which the library is a
component. In this day and age, when anyone with access to an FTP server can
create their own Linux distribution, this is a trivial undertaking. I am sure
there are several distributions already that include QT as a component. How many
people, using how many different distributions, is enough to qualify?
I would like to hear what the FSF thinks about this issue, not only the case of
the QT library, but the general case as well.
-Eugene
Steve Peltz
In article <5uqutg$1...@slinky.sky.net>,
Roger Espel Llima <es...@solar.sky.net> wrote:
>ridiculous in court, and you woulnd't want to anyway. So where's the
>problem, in practice?
The only problem in this case is for someone who, looking for the source
code and not finding it, concludes that they can't redistribute the
binary due to it being under the GPL. This makes it less useful for the
people you want to be using it.
Why not just add, after saying that the program may be licensed under the
GPL, that as a specific exeception, executable binaries may be distributed
that are linked with the Qt libraries under whatever terms those libraries
require. As long as you don't require that exception be propagated to any
derived sources, then it shouldn't matter. If you do require that, then the
GPL really isn't what you want, since you're negating it, and no one could
ever mix your code with other "true" GPL code.
I mean, nothing prevents you from releasing something with terms like:
"This program is licensed under the GPL. However, I do not grant anyone
the right to modify it or distribute binary images". You'd be really
stretching things to call your program GPLed, though. You'd be better off
writing down exactly what you DO want, rather than mixing things up with
the GPL.
I still haven't seen exactly what the Qt library terms are.
Paul D. Smith
%% eug...@cs.umb.edu (Eugene O'Neil) writes:
>> The only "escape hatch" is the system component exception. Troll Tech
>> claims that Qt falls under that exception, since there're some Linux
>> distributions that include it. Maybe they're right. But maybe they're
>> not.
eo> System component on WHAT system?
Exactly.
This is why I say the system component clause is too murky to rely on
without an explicit ruling from the FSF. So far we have obvious system
components, like libc, libm, and stuff like libsocket where necessary.
We also have an explicit ruling on Motif.
Anything else, IMO, is questionable.
eo> I suppose Motif got by because it was a system component on
eo> someone's machine, but it isn't a system component of most Linux
eo> distributions. For that matter, it is certianly not available in
eo> Windows, which can hardly be described as uncommon platform. :-\
Well, Windows is a very "uncommon platform" when you're talking about
GPL'd code, esp. code released by the FSF :)
As I mentioned in another article, Motif actually _is_ a system
component on quite a number of flavors of UNIX: AIX, HP-UX, Irix, DG-UX,
Solaris these days, Alpha OSF, etc.
Qt is a "system component" on maybe one or two distributions of Linux;
not even one whole "flavor" of UNIX.
eo> I would like to hear what the FSF thinks about this issue, not
eo> only the case of the QT library, but the general case as well.
So would I, which is why I keep urging people to ask RMS. Remember
that RMS doesn't read USENET, not even the gnu.* hierarchy including
gnu.misc.discuss.
Someone will have to email him directly.
--
-------------------------------------------------------------------------------
Paul D. Smith <psm...@baynetworks.com> Network Management Development
"Please remain calm...I may be mad, but I am a professional." --Mad Scientist
-------------------------------------------------------------------------------
These are my opinions--Bay Networks takes no responsibility for them.
Timothy Watson
Hrvoje Niksic <hni...@srce.hr> writes:
> Is...@yellow.submarine.pla (Isaac) writes:
>
> > In article <kigyb5a...@jagor.srce.hr>, Hrvoje Niksic wrote:
> > >
> > >If all the code in the program is yours only, then the risk really is
> > >negligible. However, most GPL-ed programs share their code with other
> > >such software.
> > >
> >
> > no involved party to sue you.
>
> a program not distributed by the GPL, they might sue.
As in the FSF vs. NeXT
--
________________________________________________________________________
T i m o t h y W a t s o n
tmwa...@junkmail.umich.edu (get rid of junkmail!)
__/| Something there is that doesn't love a wall, that wants it down
Iain Lea
In comp.os.linux.advocacy Vlad Petersen <vlad...@compuserve.com> wrote:
> [ various pros & cons of Qt and other GUI libs etc etc etc ]
You know what I do when these long and winding threads start unfolding ?
I go and get the software package(s), install them and critic them. I know
for some people in this thread (always lots of the same people - ermmm ;|)
that may seem strange but I do it because if any package gets a critical
discussion then it maybe worth looking at (rarely been wrong yet).
Now to KDE, Qt and the free (if I can download it and pay 0 money then its
free for me - YMMV) GUI for *nix. I downloaded KDE+Qt+PerlQt last week and
installed it on my PC (Linux) & Sun at home. Compared to any other GUI and
its underlying toolkit this duo is neat. Coupled with the PerlQt package I
can prototype GUI interfaces to my old/new toolchest apps in a very short
time. For me and not surprisingly business, time is money so PerlQt is going
to come in very handy.
Instead of all these going round in circle threads why does'nt someone who
feels the need to discuss present a nicely round argument to Troll Tech
stating why it would be a good idea to get Qt on Unix KDE released under a
more tolerant licence which would incourage more software development and
give it more market exposure (note I'm not that person as stated above time
is something I'm short of).
--
Iain Lea ia...@ecrc.de <http://www.ecrc.de/>
"We all enter this world in the same way: naked; screaming; soaked
in blood. But if you live your life right, that kind of thing doesn't
have to stop there." -- Dana Gould
Richard Saue
Paul D. Smith wrote in article ...
[lots of stuff deleted]
>You can do anything you like. If you believe the GPL allows you to use
>Qt under the system component exception, go ahead. I'm just pointing out
>that it's not 100% sure that this is legal, so you do so at your own
>risk. That risk is probably very small. But it's there.
>
Have beeing following several discussions over KDE, QT and GPL-license,
my impression that the discussion in its core has more to do with political
than legal issues, even that the discussion in the terms of the words used
seems to
indicate some legal reality.
But I cannot say that I have seen any posting beeing precisely about what
legal
risks an software author or a program user takes using QT dependent programs
released under GPL.
1. Do the author or the user risk legal actions from Troll Tech AS,
who in their
licence tells us that it is perfectly ok for them to release the
code as GPL ?
2. Do the author or the user risk legal actions from FSF who wrote
the GPL licence ?
3. Do the author or the user risk legal action from any third party ?
4. If not, what legal risks are mr Paul D Smith then talking about ?
Excuse me If all this has been explained before. I would then be happy with
a pointer to
the appropriate article.
Regards
Richard Saue
Miguel de Icaza
> >Erm, but what about the GPL? If your program is called P (which is
> >covered by GPL), then the program combined with the library that it
> >requires (P+Qt) is presumably a derived work of P (and of Qt, of
> >course, but the Qt license is OK about this). But P+Qt then can't be
> >distributed, since the GPL forbids this, doesn't it?
>
> That argument is not good: In a Solaris box, P needs libc. P+libc is a
> derived work of P, I can't distribute P+libc because the GPL forbids it.
The GPL explicitly allows this case, see point 3 in the GPL.
> I have read statements from the Red Hat people, and their problem is not
> with redistribution of KDE, but that if they include Qt, their buyers
> could fall in love with it ;-), and couldn't develop commercial software
> using it.
Please do not twist the facts, this was explained by Marc on the Gnome
mailing list and I am sure you read this.
This is what Marc Ewing <ma...@redhat.com> said on the Gnome mailing
list on this specific issue:
Marc> The reason we don't distribute Qt is that it introduces into our
Marc> product a piece of software that 3rd party software vendors have
Marc> to be careful with if they are developing apps. We don't want any
Marc> vendors to have to think about the individual licenses in Red Hat
Marc> Linux. We want them to know they can just develop apps and
Marc> distribute them as they please.
Marc>
Marc> Even Arnt at Troll thinks this is a fine policy.
> >I agree it would be nice for everybody to agree on one free toolkit,
> >but Qt isn't going to be it, just because it's not free on Windoze and
> >Macs. I'd hoped that Tk might be it, but it really doesn't seem to be
> >moving to language independence, although I still hope people get it
> >there one day.
>
> Well, GTK doesn't even exist on macs and windows, so I guess that's not
> it either?
Well, at least with Gtk you have the freedom to port it to the Mac and
to Windows and distribute your own port. The Qt license does not
grant you any right on this regard.
The Gtk ought to be pretty portable. It uses a layer of code for
accessing the windowing system called Gdk.
Miguel.
Paul D. Smith
%% "Richard Saue" <Richar...@jur.uib.no> writes:
rs> Have beeing following several discussions over KDE, QT and
rs> GPL-license, my impression that the discussion in its core has
rs> more to do with political than legal issues, even that the
rs> discussion in the terms of the words used seems to indicate some
rs> legal reality.
Since none of these licenses has ever actually been tested in court and
upheld or struck down, much of the discussion is in fact hypothetical.
However, if someone were to take it to court there'd definitely be legal
costs associated with it, regardless of how it came out.
rs> But I cannot say that I have seen any posting beeing precisely
rs> about what legal risks an software author or a program user takes
rs> using QT dependent programs released under GPL.
The only real risks are these:
* If you use GPL'd code you _don't_ own copyright to in your app,
then whomever held that copyright could very well take action or at
least threaten action.
For example, if you decided your app was going to use both Qt and
GDBM, then you'd possibly have a problem.
Normally when writing code under the GPL, you wouldn't think twice
about this.
* If you own all the copyrights, nevertheless your distribution could
become invalid. It's hard to imagine who could take you to court
over it, since it doesn't seem anyone has any basis for doing so.
Practically, this is probably not a problem.
* People see the GPL on code and they assume certain things. Even if
you can't be sued, you're still giving perhaps a false impression to
others about what your code can and can't be used for. They might
turn around and "accidentally" violate the GPL by introducing 3rd
party GPL'd code, where that author is more strict in his/her
interpretation of the GPL. This isn't the author's problem per se,
but it's misleading.
* As I've said before, my main problem is that Troll Tech has FAQs and
other documentation on their site saying their license is 100%
compatible with the GPL. _They_ make no distinctions about "3rd
party code" or anything. IMO they only tell part of the story.
And they could be right, and all this USENET bandwidth could be a
complete waste of time. All I'm saying is I think it's
irresponsible of them to make blanket statements when, in fact,
they're _not_ totally sure (or shouldn't be :).
rs> 1. Do the author or the user risk legal actions from Troll
rs> Tech AS, who in their licence tells us that it is perfectly ok for
rs> them to release the code as GPL ?
No.
rs> 2. Do the author or the user risk legal actions from FSF
rs> who wrote the GPL licence ?
The user doesn't; the user can do anything he/she likes under the GPL.
It's not until you try to distribute it that you run into the
possibility of contravening the GPL.
If the distributor uses FSF copyrighted code, a real possibility: the
FSF has proved in the past that they're not shy about enforcing the GPL,
even if they've so far not been required to do it in court.
If they felt that Qt didn't meet the requirements and someone brought an
instance of violation they could act on to their attention, they'd
_definitely_ be in touch.
rs> 3. Do the author or the user risk legal action from any
rs> third party ?
If the author uses code copyrighted by the third party, as above.
Miguel de Icaza
> So, where is the problem ? If you want to distribute free software you
> dont have to pay for your QT-License. If you sell software which is
> not free you want to earn money. And in this case it is at least
> understandable that troll wants to get a fee because you are proftting
> from their work.
Troll could dual license their Qt library: they can release Qt under
the GPL and they can release it as well under the commercial terms.
Those people who are making non-free software can get the commercial
license (since their code anyways could not be linked with the GPL
released Qt).
Miguel.
Miguel de Icaza
>Ncurses is possibly put under LGPL somewhen next year and then maintained
>by Florian Laroche, the copyright being shifted to the FSF. At least that's
>what Stallman told to the ncurses-mailing-list a few months ago.
>
>Besides there are GPL-programs that link against ncurses, for example Lynx ,
> Midnight Commander, Emacs, Oleo, dialog and Minicom.
Midnight Commander uses Slang by default. It can be linked with ncurses
if you want to.
Roberto Alsina
In article <s83enel...@luthien.nuclecu.unam.mx>, Miguel de Icaza wrote:
>I had written:
>> I have read statements from the Red Hat people, and their problem is not
>> with redistribution of KDE, but that if they include Qt, their buyers
>> could fall in love with it ;-), and couldn't develop commercial software
>> using it.
>
>Please do not twist the facts, this was explained by Marc on the Gnome
>mailing list and I am sure you read this.
>
>This is what Marc Ewing <ma...@redhat.com> said on the Gnome mailing
>list on this specific issue:
>
>Marc> The reason we don't distribute Qt is that it introduces into our
>Marc> product a piece of software that 3rd party software vendors have
>Marc> to be careful with if they are developing apps. We don't want any
>Marc> vendors to have to think about the individual licenses in Red Hat
>Marc> Linux. We want them to know they can just develop apps and
>Marc> distribute them as they please.
>Marc>
>Marc> Even Arnt at Troll thinks this is a fine policy.
Sounds fairly similar to what I said, that buyers of redhat would not be
able to develop commercial software using Qt (because of the license
problem, of course). I don't see where am I twisting the facts, really.
I used slightly bad wording, perhaps, but certainly was not trying to
twist any facts.
--
Miguel de Icaza
>Consider this: FSF says linking with Motif is ok, because it's "usually
>part of the system". Of what system? Qt comes with several Linux
>distributions. Certainly with more than come with Motif. Doesn't that
>make Qt fall under the same exception?
Well, distributions of Linux that include Motif and/or Qt as part of the
standard system are not completely free systems.
On a completely free system I can not find Motif nor Qt, so on those systems
I can not use any application that requires those libraries.
I am still from those that are happy having a system that is completely free,
without the kind of restrictions found on the Qt license. If I did not want
a free system I would be using Solaris with their CDE instead of Linux.
>>And a little less of the melodramatic "vulture" speak would not go
>>astray, either. And while I'm at it...
>
>Well, I am a little melodramatic in general. That's how I'm built :-)
That hardly will beneffit this discussion.
Miguel.
Roberto Alsina
But that wouldn't allow people to write software under a BSD or artistic
license linked to Qt, alienating a sizable portion of free software
developers.
Roger Espel Llima
In article <slrn61b9vn....@ultra7.unl.edu.ar>,
Roberto Alsina <ral...@ultra7.unl.edu.ar> wrote:
>In article <s8202yl...@luthien.nuclecu.unam.mx>, Miguel de Icaza wrote:
>>
>>Troll could dual license their Qt library: they can release Qt under
>>the GPL and they can release it as well under the commercial terms.
>>
>>Those people who are making non-free software can get the commercial
>>license (since their code anyways could not be linked with the GPL
>>released Qt).
I've suggested that before, they said that they'd lose with it. I don't
quite see how, since linking against a GPL library makes you put your
own program under the GPL, and anyone doing that currently with Qt is
going to use their free license anyway...
>But that wouldn't allow people to write software under a BSD or artistic
>license linked to Qt, alienating a sizable portion of free software
>developers.
That's definitely not the problem. Troll could keep offering Qt's
existing free license, and adding the GPL option. But they've said they
won't, so it's not much use to keep arguing about it.
--
Roger Espel Llima
es...@llaic.univ-bpclermont.fr, es...@unix.bigots.org
http://www.eleves.ens.fr:8080/home/espel/index.html
Roberto Alsina
Miguel de Icaza wrote:
>Roberto Alsina wrote:
>>Consider this: FSF says linking with Motif is ok, because it's "usually
>>part of the system". Of what system? Qt comes with several Linux
>>distributions. Certainly with more than come with Motif. Doesn't that
>>make Qt fall under the same exception?
>
>Well, distributions of Linux that include Motif and/or Qt as part of the
>standard system are not completely free systems.
But that's beyond the point. The GPL doesn't say that the part-of-the-system
exception is about free systems, and the FSF has declared they regard
Motif as a valid case of that exception.
>On a completely free system I can not find Motif nor Qt, so on those systems
>I can not use any application that requires those libraries.
But the GPL doesn't cover compiling, linking, or running a program. It
only covers copying and distributing it. The GPL can't be legal for a
program on one system and invalid on another, since there isn't a concept
of a particular system associated with the program.
>I am still from those that are happy having a system that is completely free,
>without the kind of restrictions found on the Qt license. If I did not want
>a free system I would be using Solaris with their CDE instead of Linux.
I prefer to think this isn't a black/white situation. There are degrees
of freedom, and it's better to take the most free system that will
satisfy the need for the system.
>>>And a little less of the melodramatic "vulture" speak would not go
>>>astray, either. And while I'm at it...
>>
>>Well, I am a little melodramatic in general. That's how I'm built :-)
>
>That hardly will beneffit this discussion.
Ok, I'll try to avoid melodrama while on this thread.
OTOH, being realistic does help.
Paul D. Smith
%% Is...@yellow.submarine.pla (Isaac) writes:
i> In article <p5yb56g...@baynetworks.com>, Paul D. Smith wrote:
>> * If you own all the copyrights, nevertheless your distribution could
>> become invalid. It's hard to imagine who could take you to court
>> over it, since it doesn't seem anyone has any basis for doing so.
>> Practically, this is probably not a problem.
i> Now this is just plain silly. Invalid in what sense?
Invalid in the sense that that same combination (GPL + Qt) may be proven
to be legally inconsistent through other court cases (say something the
FSF might bring about).
i> You seem to feel that the distributor should be sued, but that
i> there is no offended party.
I never said I thought the distributor _should_ be sued, and I
explicitly said "it doesn't seem that anyone has any basis for doing
so", which sounds a lot like what you said... or at least I meant it to :)
i> You are free to distribute any code you own, any way you like,
i> period.
True, but others also distribute it (that is, after all, the entire
point). I suppose in some weird universe it would be OK for the
copyright holder to distribute directly, since he can do whatever he
likes, but that no one _else_ could distribute it since it's illegal.
But this whole area is pretty silly. As I said, practically, this is
not a problem.
i> The only third party I can see being remotely interested in pursuing this
i> would be the FSF. If they won the result would be less GPL'd software,
i> and there would be no affect on the party really responsible for
i> creating thee confusion.
I can easily imagine the FSF pursuing this to make a philosophical point
that perhaps has little to do with Qt but might bother them enough to
make clear in general:
That you can't escape the GPL through the "system component" clause
simply by creating "Joe's Linux Distribution" with the non-free code you
want to use with the GPL as a standard part, and saying it's a "system
component" somewhere, on some system, and so it qualifies.
Troll Tech had released Qt under this license before _any_ distributions
were shipping with it, and claimed the Motif exemption. Now that there
are a couple of distributions at least they have a bit more firm ground,
but I still don't think they're being totally honest with people in
their FAQs.
Of course the FSF could probably accomplish the same thing by tightening
up that section of the GPL, if not in the license itself then at least
via some official statement on how they interpret it.
I think we're basically jumping up and down on the powdered remnants of
the bones of a dead horse at this point, so unless someone has something
new...
Paul D. Smith
%% ral...@ultra7.unl.edu.ar (Roberto Alsina) writes:
ra> But that wouldn't allow people to write software under a BSD or artistic
ra> license linked to Qt, alienating a sizable portion of free software
ra> developers.
The Artistic License doesn't conflict with the GPL.
Isaac
In article <p5yb56g...@baynetworks.com>, Paul D. Smith wrote:
> * If you own all the copyrights, nevertheless your distribution could
> become invalid. It's hard to imagine who could take you to court
> over it, since it doesn't seem anyone has any basis for doing so.
> Practically, this is probably not a problem.
>
that the distributor should be sued, but that there is no offended party.
You are free to distribute any code you own, any way you like, period.
> * People see the GPL on code and they assume certain things. Even if
> you can't be sued, you're still giving perhaps a false impression to
> others about what your code can and can't be used for. They might
> turn around and "accidentally" violate the GPL by introducing 3rd
> party GPL'd code, where that author is more strict in his/her
> interpretation of the GPL. This isn't the author's problem per se,
> but it's misleading.
>
And this is not so silly. Once you grab someone else's code, that
person has the right to object to your interpretation of their
license. I place the blame for this on Troll Tech.
Nonetheless there are several reasons why the question is not worth the
trouble we've spent on this.
1. A GPL'd library with the same API would make the whole question moot
(at least with dynamically linked binaries)
2. Generally speaking there would be very low motivation to pursue
violations of the type you've described because the program will be
distributed in source form. Presumably if borrowed code is GPL'd, the
program will be distributed with GPL copyrights all over it.
The only third party I can see being remotely interested in pursuing this
would be the FSF. If they won the result would be less GPL'd software,
and there would be no affect on the party really responsible for
creating thee confusion. If it were left alone, I guess there would
be a reduced desire for a GPL'd equivalent, but isn't there one in the
works already?
3. The more popular Qt becomes, the more likely that it will get a
Motif style exception granted by FSF (although how this affects code
owned by someone other than FSF, I don't understand even for Motif)
Isaac
o r c e l l . p o r t l a n d . o r . u s
In article <m23env9...@hubert.wuh.wustl.edu>,
Alan Shutko <a...@acm.org> wrote:
>>>>>> "D" == david parsons <o> writes:
>
>>> And no legal way to maintain it or release fixed versions. Unless
>>> the QT license allows that if they go under?
>
>D> What happens to a copyright when the copyright holder falls over
>D> dead? Either the copyright is given to someone else, at which
>D> point you can pester them for support, or it lapses into the public
>D> domain.
>
>Nope. If an author created a work now, the copyright would be valid
>for 50 years after he died.
If nobody takes possession of the copyright, the matter is moot. If
Troll Tech rolls belly up and does not assign (or have the court
assign) the copyright to another agency, the massed governments of
the world will not obligingly step in and protect the Qt license as
a matter of National Defense.
>In other words, Qt users could be screwed.
Nope. The people who have free licenses now will still have them,
and one of the (only) spiffy things about C++ is that it's really
easy to repair defects by building scaffolding around the parts that
are broken.
The only people who would be screwed are the people who believe the
FUD going around here, and they've already gathered their forces behind
one or another of the other 99 million toolkits that are pulling X to
the bottom of the sea.
____
david parsons \bi/ ... at least Qt developers will be able to flee to
\/ other platforms.