A Modest Proposal for a GNU infrastructure license
James Michael DuPont
I am bringing this idea up on gnu.misc.discuss as to
avoid eating up more bandwidth on the dotgnu, gcc and
mono lists where we have had many interesting
discussions about similar topics.
I have a modest proposition :
To create a license for software that only takes GPLed
sourcecode as input and produces GPLed code as output.
Basically, it is to create a licensed persistent
memory model where all data and functions in whatever
form are treated as if they are in the main memory of
a statically linked executable, even if all types of
neat tricks are used.
More to the point, that the inputs and outputs must be
already be source code (in what ever form)
and that must already be under the GPL (in whatever
form). If documentation is produced, then it would be
under the GPL documentation license.
It is a bit like a recusive function f(n>1) that does
not solve the base case, and leave the n=1 and n=0 to
a specific other function.
The GPL is used by many projects to maintain thier
freedom.
The LGPL is a license that allows for the production
of non-free software. This could be seen as a
right-wing GPL, a RGPL that is a restrictive GPL.
One that only allows for the production of more free
software.
Because the GPL is so sucessful, there are many
applications that use it. It is now possible to work
only with free software.
I would like to give those who use free software an
advantage over those who don't.
By making services and software that are only legally
usable by those using free software.
This special type of license will only applies to
software that works on other software, thus
meta-software.
The inputs and outputs of this meta software of any
form would be limited to free software only.
This of course does not affect the input and output of
the free software once it is compiled, so the used
software is not changed. Only the usage of the GNU
Infrastructure License would be affected.
The GPL is carefully designed to allow users to do
anything with the input and output of the files not to
affected by the license.
Some authors decide to publish a version under GPL for
non-profit usage. I want to go a step further.
My question is if I can publish a software that is for
GNU usage only? That is only licensed for usage in
conjunction with free software, down to the user.
This would then be usable to create an entire range of
services for promoting the usage of the GPL and giving
them and advantage and a incentive over non-free
software.
This would overcome the issues non-free software using
XML, File Systems, XML-RPC, SOAP, Dynamic Linking, and
Web Services for interfacing with GNU software and
going around the GPL.
This would give an explicit license to a limited set
of powerful interfaces that are for use for only free
software. This would give people who want to
experiment with interfacing to the GCC via dynamic
linkage a legal basic
Now the protection of the data files are very tricky,
because the data file are sourcecode.
The best would be to treat all data files as if they
are persistent inside the memory of the GPLed program
and that any usage of them is a derived work.
If a data file is produced derived from source code,
this data file would also be under a special form of
the GPL that prevents it from being used by any non
GNU infrastructure licensed tool.
Of course you wont have the freedom to take away this
protection offered.
This would allow the safe storage of internal data
structures without their usage in non-free software.
No non-gpled program would not be allowed to use any
of the internal data structures of this program.
Not to read or write them or process them
electronically. If it is offered as a web service,
only users from free software projects would be
allowed to use the software in conjunction with free
software.
I hope to hear your feedback.
Best Regards,
Mike
=====
James Michael DuPont
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
Michael DuPont
I am bringing this idea up on the nntp gnu.misc.discuss as to
avoid eating up more bandwidth on the dotgnu, gcc and
mono lists where we have had many interesting
discussions about similar topics.
Note that I have posted this to the mailling list, and hope that you
dont get this twice. It seems that not all emails get posted to the
mailling list.
Daniel Carrera
I don't know how I got into this CC list. Could you please take me off of
it? I already receive more email than I can read.
Daniel.
On Mon, 10 Jun 2002, James Michael DuPont wrote:
> Dear All,
>
> I am bringing this idea up on gnu.misc.discuss as to
> avoid eating up more bandwidth on the dotgnu, gcc and
> mono lists where we have had many interesting
> discussions about similar topics.
>
Anthony Ventimiglia
gnu.misc.discuss>>>>>>
> I have a modest proposition :
> To create a license for software that only takes GPLed
> sourcecode as input and produces GPLed code as output.
I pretty much think you won't be able to GPL this program. Two reasons:
First, I think it's against the spirit of the GPL to restrict users in any
way or decide who can and can't use your software.
But that's not much of a valid reason, here's the reason why it just won't
fly -- If you GPL this software, then write it so it will only input and
output GPL'd stuff, under the GPL, anyone can modify and distribute the
source to remove that restriction. So you'd have to use a non-GPL license
for all this.
> The GPL is used by many projects to maintain thier
> freedom.
Yes it is, but what you're proposing is very restrictive, the GPL maintains
freedom, but It's not your place to force it down people's throats.
> The LGPL is a license that allows for the production
> of non-free software. This could be seen as a
> right-wing GPL, a RGPL that is a restrictive GPL.
> One that only allows for the production of more free
> software.
I see you're point, but it's the wrong idea. There are folks out there who
don't want to release Free Software, and their minds won't be changed by
not allowing them to use it. The beauty of the GPL and LGPL is that it can
live side by side with non-free software. You can't have a revolution with
a 100% takeover, by having the ability to have Free software used on and
with non-free software, people can get used to it, realize how powerful it
is, and eventually decide it's better overall.
> Because the GPL is so sucessful, there are many
> applications that use it. It is now possible to work
> only with free software.
The reason it's successful is because it's FREE, what you are suggesting is
NOT-FREE. It's saying, "Yes you can use this FREE program, but you aren't
FREE in what you do with it."
> My question is if I can publish a software that is for
> GNU usage only? That is only licensed for usage in
> conjunction with free software, down to the user.
You can but it's not Free Software.
> I hope to hear your feedback.
> Best Regards,
I think this argument makes as much sense as Anit-abortionists who kill
abortion doctors. In their case they are taking life in the defense of
another.
In your case you propose to remove freedom in order to preserve it. No
Thanks.
Rob Ristroph
Michael>
Michael> I have a modest proposition :
Michael> To create a license for software that only takes GPLed
Michael> sourcecode as input and produces GPLed code as output.
...
Michael> I would like to give those who use free software an
Michael> advantage over those who don't.
Michael>
Michael> By making services and software that are only legally
Michael> usable by those using free software.
Michael,
I think that a license that only limits the conditions of
re-distribution is very strong, because Title 17 (The Federal
Copyright Code) explicitly gives the copyright holder control
over copying.
A license that limits _use_ in some way has to depend on some
other mechanism, such as the way the EULA's of the more
restrictive proprietary software work -- claiming some sort of
contract. What would be the trade in the contract ? You get
to re-distribute the software if you only use it as I specify
? If they didn't re-distribute it, could they not enter the
contract and thus use it however they wished ?
I think that, in general, attempts to limit use as opposed to
distribution are legally weaker.
--Rob
Isaac
>
> I think that, in general, attempts to limit use as opposed to
> distribution are legally weaker.
>
shareware rely on being able to allow copying and distribution but
still being able to require payment for usage?
Isaac
Michael DuPont
> >>>>On Monday 10 June 2002 09:30 am, Michael DuPont wrote to
> gnu.misc.discuss>>>>>>
>
> > I have a modest proposition :
> > To create a license for software that only takes GPLed
> > sourcecode as input and produces GPLed code as output.
>
> I pretty much think you won't be able to GPL this program. Two reasons:
> First, I think it's against the spirit of the GPL to restrict users in any
> way or decide who can and can't use your software.
I dont think I can just GPL it, the GPl does not seem to cover it.
That is why we need a more restrictive GPL.
As I said, there are license that restrict the usage to just
non-profit usage,
well why not restrict the usage to only GPL?
If the input is only source code, and the outputted source code is
under the same license, then what is the problem?
> source to remove that restriction. So you'd have to use a non-GPL license
> for all this.
As I said, a RGPL.
> > The GPL is used by many projects to maintain thier
> > freedom.
>
> Yes it is, but what you're proposing is very restrictive, the GPL maintains
> freedom, but It's not your place to force it down people's throats.
>
> > The LGPL is a license that allows for the production
> > of non-free software. This could be seen as a
> > right-wing GPL, a RGPL that is a restrictive GPL.
> > One that only allows for the production of more free
> > software.
>
> I see you're point, but it's the wrong idea. There are folks out there who
> don't want to release Free Software, and their minds won't be changed by
> not allowing them to use it.
Right now they are just creating file interfaces and network
interfaces to go around the GPL. No one can stop them.
> You can't have a revolution with
> a 100% takeover, by having the ability to have Free software used on and
> with non-free software, people can get used to it, realize how powerful it
> is, and eventually decide it's better overall.
Times are changing, and there is a big enough market for GPLed
sourcecode to create a derivitive market that excludes non-free
software.
>
> > Because the GPL is so sucessful, there are many
> > applications that use it. It is now possible to work
> > only with free software.
>
> The reason it's successful is because it's FREE, what you are suggesting is
> NOT-FREE. It's saying, "Yes you can use this FREE program, but you aren't
> FREE in what you do with it."
You can use this free software to work on free software. A software
tool that promotes and helps free software developers only. You are
free to use it on any free software you like, it is just too dangerous
to allow its usage on non-free software becuase it would allow for the
abuse of the free software.
> I think this argument makes as much sense as Anit-abortionists who kill
> abortion doctors. In their case they are taking life in the defense of
> another.
>
> In your case you propose to remove freedom in order to preserve it. No
> Thanks.
I think you are shooting too quickly here.
I am talking about a very specific case here, about meta-software that
operates only on other software. It is not your every day editor.
Take a look at my project introspector.sourceforge.net, the FSF has
requested that I dont continue with it, because it represents a danger
to free software.
That is because we dont have the license to protect free software yet
on that level.
Regards,
Mike
Jeffrey Siegal
> (Actually, I once thought of a way to have a usage restriction in a license.
> IANAL, so I have no idea if this would work. Put a clause in the license
> which says "you are permitted to distribute copies only if the copies are
> distributed to someone who agrees to term X". If you distribute copies to
> someone who doesn't agree, you've distributed a pirated copy, and there are
> no rights to use pirated copies. Of course, this would pretty much rule
> out something like putting copies on a web site.)
No it wouldn't. Just put the same license terms on the web site.
Jeffrey Siegal
> I think that a license that only limits the conditions of
> re-distribution is very strong, because Title 17 (The Federal
> Copyright Code) explicitly gives the copyright holder control
> over copying.
The courts have consistently ruled that using software almost always
means making copies of it -- on disk, in RAM, etc. -- and that such
copies are often subject to the control of the copyright holder. In
order to exercise this control, however, the copyright holder must
distribute the software using licensing rather than sale of copies in
order to negate section 117 of the act (which gives owners, but not
licensees, of copies of programs the right to use them without permission).
Jeffrey Siegal
> I pretty much think you won't be able to GPL this program. Two reasons:
> First, I think it's against the spirit of the GPL to restrict users in any
> way or decide who can and can't use your software.
Bear in mind that the FSF has already said that the next version of the
GPL will incorporate some use restrictions. In particular, those making
use of a GPLed program accessible on a public web site will be required
to make the source code to the program available as well.
Michael DuPont
> I think that a license that only limits the conditions of
> re-distribution is very strong, because Title 17 (The Federal
> Copyright Code) explicitly gives the copyright holder control
> over copying.
The copyright holder can specify that the software is only usable
under certain conditions, right?
>
> A license that limits _use_ in some way has to depend on some
> other mechanism, such as the way the EULA's of the more
> restrictive proprietary software work -- claiming some sort of
> contract.
So how do the limitations of the GPLed for private and non-profit use
work?
> What would be the trade in the contract ?
You can use this software and get services in return that this
information is used only by free software.
>You get
> to re-distribute the software if you only use it as I specify
> ?
The software can be redistributed, just not used on non free software.
The output files are then source code and there is no license to use
the intermediate files. The are licensed like in memory data.
> If they didn't re-distribute it, could they not enter the
> contract and thus use it however they wished ?
There is no other contract that gives them any other rights to use the
software.
Mike
Jeffrey Siegal
>> I think that a license that only limits the conditions of
>> re-distribution is very strong, because Title 17 (The Federal
>> Copyright Code) explicitly gives the copyright holder control
>> over copying.
>
> The copyright holder can specify that the software is only usable
> under certain conditions, right?
Only if the software is licensed. Not if copies of it are sold (or
given away free).
>> A license that limits _use_ in some way has to depend on some
>> other mechanism, such as the way the EULA's of the more
>> restrictive proprietary software work -- claiming some sort of
>> contract.
>
> So how do the limitations of the GPLed for private and non-profit use
> work?
There are no such limitations.
Michael DuPont
You know what I was just thinking.
If GPL does not protect against the usage of XML and Files to extract
data,
then I could create software the uses this one weakness in the GPL,
and then put it under this new proposed "R-GPL" license.
There would be no recourse for the GPLed author to protect themselves
against this under the GPL, but it would at least allow people to
write software that uses more advanced technology with stronger
protection. Anyone who then used the output files would have to then
put thier code under the RGPL, creating a huge chain reaction.
The software that I write to use these XML files would be completly
isolated from the GCC, so there would be nothing stopping me from
creating a much stronger license for that software?
Problem is that the XML files would then be open for anyone to use! If
I were to link directly into the GCC, then I would not be able to
create a stronger license.
There must be a way to limit the usage of data files under some
special form of the GPL.
This comes back down to the discussion I had on the various mailling
lists, and with many people in private.
Here are the relevant links :
Also on this list :
Here is the google link the discussion here was "Question about
derived works from GCC"
http://groups.google.com/groups?hl=de&lr=&threadm=8118d547.0203062352.7dcacb62%40posting.google.com&rnum=1&prev=/groups%3Fq%3Dmdupont777%2Bissac%26hl%3Dde%26lr%3D%26selm%3D8118d547.0203062352.7dcacb62%2540posting.google.com%26rnum%3D1
On the GCC mailling list :
http://gcc.gnu.org/ml/gcc/2002-02/msg01792.html
http://gcc.gnu.org/ml/gcc/2002-02/msg01808.html
on the inline list :
http://archive.develooper.com/inl...@perl.org/msg01125.html
One the mono list
http://lists.ximian.com/archives/public/mono-list/2002-May/005770.html
on the DOTGNU list :
http://subscribe.dotgnu.org/pipermail/developers/2002-March/002193.html
http://archive.dotgnu.org/pipermail/developers/2002-June/003258.html
on the XMLDEV
http://lists.xml.org/archives/xml-dev/200202/msg01411.html
on the OHS-Talk List:
http://www.bootstrap.org/lists/ba-ohs-talk/0204/msg00104.html
Mike
Michael DuPont
Fine, As I said before (Se my reply to myself), there is nothing
stopping me interfacing to a Gpled program via XML-RPC and then
restricting the license as I see fit.
If the GPL is not strong enough to protect itself, something stronger
is needed.
The affero changes are not going to cut it on this round.
mike
Anthony Ventimiglia
>
>> You can't have a revolution with
>> a 100% takeover, by having the ability to have Free software used on and
>> with non-free software, people can get used to it, realize how powerful
>> it is, and eventually decide it's better overall.
> Times are changing, and there is a big enough market for GPLed
> sourcecode to create a derivitive market that excludes non-free
> software.
The reason times were able to change is because of the Freedom of GPL'd
software. If the GPL was written with restrictions silmiar to what you're
proposing, I don't think it would exist today.
By Exclusion, you remove freedom. That's my very point. Naming it the RGPL
would be misleading, because it would NOT be a general Public License, it
would be aimed at restricting a certain group. When you put the word
exclude in there, regardless of context, you remove freedom.
What would happen if apache was made so it wouldn't serve pages to non-free
browsers? Or if Apache was made so it could not run on Non-Free OS's ? Do
you think Apache would be the number one server on the net right now if It
had been given those restrictions from day one ?
Or what if you were told that vim (or emacs) couldn't be used for non-free
software. For instance if you tried writing a DOS batch file you wouldn't
be able to save it ? How do you think that bird would fly ? Would you still
consider that software free ?
Now I know Apache and Vim aren't under the GPL, so that might be a bad
example, but let's take gcc. What if gcc were written so it would
automatically upload the source for everything it compiles to a public
site. Essentially making it impossible to write proprietary code ? I sure
as hell wouldn't use anything like that. Do you think I'm the only one who
would feel that way?
>> I think this argument makes as much sense as Anit-abortionists who kill
>> abortion doctors. In their case they are taking life in the defense of
>> another.
>>
>> In your case you propose to remove freedom in order to preserve it. No
>> Thanks.
> I think you are shooting too quickly here.
No not at all. That's exactly what you're proposing.
> I am talking about a very specific case here, about meta-software that
> operates only on other software. It is not your every day editor.
> Take a look at my project introspector.sourceforge.net, the FSF has
> requested that I dont continue with it, because it represents a danger
> to free software.
> That is because we dont have the license to protect free software yet
> on that level.
I realize it's not your everyday editor, regardless, you cannot possibly
restrict a piece of software in this way and keep it free. So you are using
a restrictive non-free license to defend free-software. And Like I said I'm
totally against that.
I would love to see a copy of the exact letter from the FSF asking you to
stop this, so I can see their reasoning behind it. From the way I
understand the GNU policy, they would not be able to endorse anything
written with a restrictive license like this, because it goes against the
freedom of free software.
To top it off, this license would be impossible to enforce unless the
source code was totally proprietary.
Anthony Ventimiglia
What you just mentioned has been a grey area in the GPL for some time, and
it's not restrictive as much as protective. It's not really a use
restriction, it's just a more specific definition of distribution.
What is being proposed here is a completely different thing.
Anthony Ventimiglia
These qoutes are all coming right off the GNU site:
These are all quotes from RMS stating why he started the GNU project:
http://www.gnu.org/gnu/thegnuproject.html
-quote-
"The easy choice was to join the proprietary software world, signing
nondisclosure agreements and promising not to help my fellow hacker. Most
likely I would also be developing software that was released under a
nondisclosure agreements, thus adding to the pressure on other people
to betray their fellows too.
I could have made money this way, and perhaps amused myself writing code.
But I knew that at the end of my career, I would look back on years of
building walls to divide people, and feel I had spent my life making
the world a worse place."
-------
"Building walls" is what the GNU project is aimed at stopping, you are
proposing doing exectly that.
-qoute-
"I chose to make the system compatible with Unix so that it would be
portable, and so that Unix users could easily switch to it. The name
GNU was chosen following a hacker tradition, as a recursive acronym for
"GNU's Not Unix." "
-------
Right there he states that he chose to make GNU Unix compatible to be an
attractive alternative to non-free software and OS's
-quote-
The GNU C library uses a special kind of copyleft called the GNU Library
General Public License, which gives permission to link proprietary
software with the library. Why make this exception?
It is not a matter of principle; there is no principle that says
proprietary software products are entitled to include our code. (Why
contribute to a project predicated on refusing to share with us?) Using
the LGPL for the C library, or for any library, is a matter of strategy.
The C library does a generic job; every proprietary system or compiler
comes with a C library. Therefore, to make our C library available only
to free software would not have given free software any advantage--it
would only have discouraged use of our library.
One system is an exception to this: on the GNU system (and this includes
GNU/Linux), the GNU C library is the only C library. So the distribution
terms of the GNU C library determine whether it is possible to compile a
proprietary program for the GNU system. There is no ethical reason to
allow proprietary applications on the GNU system, but strategically it
seems that disallowing them would do more to discourage use of the GNU
system than to encourage development of free applications.
--------
Until everyone is using and writing free-software this is still valid. Even
if there comes a day when Proprietary, non-free software ceases to exist,
removing this philosophy would render the GNU project non-free.
John Hasler
> I realize it's not your everyday editor, regardless, you cannot possibly
> restrict a piece of software in this way and keep it free.
You also cannot possibly restrict a piece of software in this way and
expect anyone to use it. Your program will end up in the same place as the
1980s stuff that had licenses saying things like "This program may not be
used by the South African police".
--
John Hasler
jo...@dhh.gt.org (John Hasler)
Dancing Horse Hill
Elmwood, WI
Isaac
<shep...@pack.0f.wo1v35> wrote:
>
> What you just mentioned has been a grey area in the GPL for some time, and
> it's not restrictive as much as protective. It's not really a use
> restriction, it's just a more specific definition of distribution.
It's a defintion of distribution that doesn't match the one in copyright
law. It seems much closer to a usage restriction than a distribution
restriction.
Isaac
Anthony Ventimiglia
> Anthony Ventimiglia writes:
>> I realize it's not your everyday editor, regardless, you cannot possibly
>> restrict a piece of software in this way and keep it free.
>
> You also cannot possibly restrict a piece of software in this way and
> expect anyone to use it. Your program will end up in the same place as
> the 1980s stuff that had licenses saying things like "This program may not
> be used by the South African police".
Exactly
Michael DuPont
That is crazy. Up to know there is not any support of the GCC group to
provide any dynamic linking interfaces or file exchange formats.
Becuase of thier worries.
RMS said on that list I will dig up the mail, that if anyone wants to
link the gcc dynamically, to please contact him privatly to discuss
it.
That removes all scripting language support for linking into the gcc,
you cannot embed the parser into an IDE, you cannot build a program
database.
If the only limitation for using these tools was that you cannot use
the program database from non-free software, which you cannot do now
anyway, then what is the difference?
Right now you can only access the parse trees of the gcc if you are a
blessed gcc developer with a good understanding of the compiler
internal.
There are not any C++, java, perl, or any other language interfaces
supported or wished for.
The situation is now that the gcc group is asking you to not use your
freedom to protect them. This new license would cover the dynamic
intefaces and libs, and file format only. A very small part of the
spectrum, only that which is not covered by the GPL.
That way you can then safely us this data without having anyone fear
abuse.
May people would willling only use this only by GPLed code, and for
GPled projects. There is a huge market for work on gpled code.
Let me try and explain again :
1. lots of gpled code to maintain.
2. hardly any tools to do code transformations.
3. the tools available dont have good parsers or dont support a raplid
development, ide and scripting environment.
4. the gcc provides all the information the ide would need, but you
cannot use your freedom to do so without getting baned from the list
and getting absolutly no support.
5. the IDEs roll thier own bad parser, and free software looks schizo.
6. Microsoft create lots of hidden interfaces and protected file
formats, limit the access to the program database to only thier
programs. But the have a great IDE and people are productive because
they have embedded the compiler and dont have any problem with using
it.
So you tell me if noone would use this software, the IDE can be under
GPLed, but the interface to the gcc can only be used by GPLed
software.
The Parse trees would not be licensed for use by anyone else, so you
wont have non-free software hooking in and extracting them, the data
is licensed as if it is in the memory of the gcc, so you cannot access
it.
The only reason why the gcc has been sucessfull is becuase the network
technology for exchanging data between programs has been too slow and
hard to use. Now it is easy and standard.
The only protection the gcc has is that it is monolithic, and none
uses thier freedom to make somthing more flexible. That is also
because the GPL does not support enough interfaces to protect that
flexibility.
My proposed R-GPL would be PRO GPL, it would allow for the creation of
many interesting programs and web services, it would promote free
software and foster the creation of many more programs. The only
restriction is the restrict that is in place by convention in the
gcc, that you dont give the data to thier parties.
So you ask yourself who is restricting freedom here?
It is ok to create aggreements that restrict peoples freedom in the
name of protecting the gpl, but is not ok to just put down into a
license, however restrictive excatly what is being aggreed. All of you
seem to be shocked about what I am proposing, it is just what is
happening right now, but not codified.
Mike
mdupont
RMS said that v3 will not cover it and not to hold my breath on any version
of the GPL covering it.
mike
mdupont
mike
"Isaac" <is...@latveria.castledoom.org> wrote in message
news:slrnagd9h...@latveria.castledoom.org...
mdupont
> "Building walls" is what the GNU project is aimed at stopping, you are
> proposing doing exectly that.
OK, so you are saying that I should not care how the data is used and if it
is harming the GNU project.
I am just trying to add in features and build a safety net.
> Right there he states that he chose to make GNU Unix compatible to be an
> attractive alternative to non-free software and OS's
Only if a big name goes first, we follow? What about defining and creating a
interface in an act of creativity?
>> One system is an exception to this: on the GNU system (and this
includes
>> GNU/Linux), the GNU C library is the only C library. So the
distribution
>> terms of the GNU C library determine whether it is possible to compile
a
>> proprietary program for the GNU system. There is no ethical reason to
>> allow proprietary applications on the GNU system, but strategically it
>> seems that disallowing them would do more to discourage use of the GNU
>> system than to encourage development of free applications.
> Until everyone is using and writing free-software this is still valid.
Even
> if there comes a day when Proprietary, non-free software ceases to exist,
> removing this philosophy would render the GNU project non-free.
The problem is that the GPL does not cover webservices or any form of non
standard linking.
These is no protection against any form of usage, so the FSF/GCC tries to
stamp out anyone creating such interfaces to the gcc. If you limit them to
use by free software, which is what is done by keeping them in memory
anyway, then you can build better tools.
Otherwise our freedom is being limited because the GPL license in not good
enough to protect it.
mike
mdupont
what if I put the introspector output into the OpenCyc Projects?
They dont seem to be blasted for locking people in:
http://www.opencyc.org/license
>The OpenCyc Knowledge Base consists of code, written in the declarative
language CycL, that represents or supports the >representation of facts and
rules pertaining to consensus reality.
>OpenCyc is licensed using the GNU Lesser General Public License, whose text
can also be found below. The OpenCyc >CycL code base is the "library"
referred to in the LGPL license. The terms of this license equally apply to
renamings and >other logically equivalent reformulations of the Knowledge
Base (or portions thereof) in any natural or formal language.
From the FAQ:
>Q: What does Cycorp mean by saying that the knowledge base will be "open
source"? Will it be publicly available? Will it be free?
>A: Yes, OpenCyc may be freely copied, distributed and used for commercial
or non-commercial purposes according to the terms of the >OpenCyc license,
which is similar to the GNU Library or "Lesser" Public License (LGPL).
Qualified parties can obtain a free license to a >substantially larger
subset of the Cyc Knowledge Base known as ResearchCyc, which is for R&D use
only. The complete Cyc Knowledge >Base can be licensed from Cycorp, Inc. for
commercial use. Terms for licensing the complete Cyc KB are negotiated on an
individual basis. >Year by year, each assertion in the latest version of Cyc
will migrate to a subsequent release of ResearchCyc, and each assertion in
>ResearchCyc will migrate to a later release of OpenCyc.
"Isaac" <is...@latveria.castledoom.org> wrote in message
news:slrnagd9h...@latveria.castledoom.org...
Christopher Browne
>> "Building walls" is what the GNU project is aimed at stopping, you
>> are proposing doing exectly that.
>
> OK, so you are saying that I should not care how the data is used
> and if it is harming the GNU project.
A "safety net" that people refuse to use is not a useful safety net.
And you're proposing building a "safety net" that many will refuse to
use.
--
(concatenate 'string "cbbrowne" "@acm.org")
http://www.cbbrowne.com/info/linuxxian.html
"We should forget about small efficiencies, say about 97% of the time:
premature optimization is the root of all evil." -- Donald Knuth
Anthony Ventimiglia
> That is crazy. Up to know there is not any support of the GCC group to
> provide any dynamic linking interfaces or file exchange formats.
> Becuase of thier worries.
> RMS said on that list I will dig up the mail, that if anyone wants to
> link the gcc dynamically, to please contact him privatly to discuss
> it.
Well have you sent him this proposal? Or contacted him at all?
> That way you can then safely us this data without having anyone fear
> abuse.
I have no fear of abuse.
> May people would willling only use this only by GPLed code, and for
> GPled projects. There is a huge market for work on gpled code.
> Let me try and explain again :
>
>
> 1. lots of gpled code to maintain.
> 2. hardly any tools to do code transformations.
> 3. the tools available dont have good parsers or dont support a raplid
> development, ide and scripting environment.
> 4. the gcc provides all the information the ide would need, but you
> cannot use your freedom to do so without getting baned from the list
> and getting absolutly no support.
> 5. the IDEs roll thier own bad parser, and free software looks schizo.
> 6. Microsoft create lots of hidden interfaces and protected file
> formats, limit the access to the program database to only thier
> programs. But the have a great IDE and people are productive because
> they have embedded the compiler and dont have any problem with using
> it.
>
> So you tell me if noone would use this software, the IDE can be under
> GPLed, but the interface to the gcc can only be used by GPLed
> software.
I wouldn't use it, because I want to have the right to put any copyright I
desire on my software, I want my right to freedom. But I probably wouldn't
ever use an IDE, because vim is all the IDE I need.
> My proposed R-GPL would be PRO GPL, it would allow for the creation of
> many interesting programs and web services, it would promote free
> software and foster the creation of many more programs. The only
> restriction is the restrict that is in place by convention in the
> gcc, that you dont give the data to thier parties.
Non-free software cannot promote free software.
>
> So you ask yourself who is restricting freedom here?
> It is ok to create aggreements that restrict peoples freedom in the
> name of protecting the gpl, but is not ok to just put down into a
> license, however restrictive excatly what is being aggreed. All of you
> seem to be shocked about what I am proposing, it is just what is
> happening right now, but not codified.
Like I said before -- it's not ok for a pro lifer to kill abortion doctors.
Anthony Ventimiglia
> OK, so you are saying that I should not care how the data is used and if
> it is harming the GNU project.
> I am just trying to add in features and build a safety net.
I'm saying that you are building walls. Sure you might be against people
writing non-free software with your tool, but you can't stop them without
releasing non-free software yourself. How would you release the source code
to anyone without stopping them from removing your "non-free" blocks ?
>> Right there he states that he chose to make GNU Unix compatible to be an
>> attractive alternative to non-free software and OS's
> Only if a big name goes first, we follow? What about defining and creating
> a interface in an act of creativity?
Go right ahead you seem bent on writing something very few people want to
use, make the interface unfamiliar too, I'm sure that will attract more
users.
> The problem is that the GPL does not cover webservices or any form of non
> standard linking.
> These is no protection against any form of usage, so the FSF/GCC tries to
> stamp out anyone creating such interfaces to the gcc. If you limit them to
> use by free software, which is what is done by keeping them in memory
> anyway, then you can build better tools.
Ok listen carefully to what you just said
-- "There is no protection against any form of usage" -- That fits in to my
definition of freedom, I don't know how you define it.
Do you see the whole point at what I'm getting at ? You want to limit
freedom, and that goes against the spirit of the GPL. And the only way you
can write the software you are proposing is by making it non-free.
> Otherwise our freedom is being limited because the GPL license in not good
> enough to protect it.
How is my freedom not being protected ?
How does the GPL fall short ?
True freedom respects the beliefs of others, even if they are against
freedom. If we truly had a world of "Free Software" There would be no need
for the GPL, but because "non-free" software, music, literature, etc ... is
the norm, the Copyleft is needed to play the non-free game.
Michael DuPont
> >>>>On Wednesday 12 June 2002 05:04 am, mdupont boldly stated:
>
> > OK, so you are saying that I should not care how the data is used and if
> > it is harming the GNU project.
> > I am just trying to add in features and build a safety net.
>
> I'm saying that you are building walls. Sure you might be against people
> writing non-free software with your tool, but you can't stop them without
> releasing non-free software yourself. How would you release the source code
> to anyone without stopping them from removing your "non-free" blocks ?
>
> >> Right there he states that he chose to make GNU Unix compatible to be an
> >> attractive alternative to non-free software and OS's
> > Only if a big name goes first, we follow? What about defining and creating
> > a interface in an act of creativity?
>
> Go right ahead you seem bent on writing something very few people want to
> use, make the interface unfamiliar too, I'm sure that will attract more
> users.
not limited.
And there is not other way to do this and link to the GPL.
>
>
> > The problem is that the GPL does not cover webservices or any form of non
> > standard linking.
> > These is no protection against any form of usage, so the FSF/GCC tries to
> > stamp out anyone creating such interfaces to the gcc. If you limit them to
> > use by free software, which is what is done by keeping them in memory
> > anyway, then you can build better tools.
>
> Ok listen carefully to what you just said
>
> -- "There is no protection against any form of usage" -- That fits in to my
> definition of freedom, I don't know how you define it.
That is right, The gpl protects against usage of the software to
create derived forms via linking, but not via file formats. The
protection is only good in main memory, and old fashioned notion.
>
> Do you see the whole point at what I'm getting at ? You want to limit
> freedom, and that goes against the spirit of the GPL. And the only way you
> can write the software you are proposing is by making it non-free.
Sure I do, I see it all the way. I am just taking the arguments from
the FSF and the GCC and taking them to thier logical conclusion. My
first version of the introspector was Dual Licensed, Perl Artistic for
the Perl and GPL for the C. Then I changed to GPL. I am thinking about
the RGPL just to offer the protection to allow for me to link to the
system like I need to. The GCC team and FSF want to limit the freedom
of the users, not me. I am just trying to argue thier point and
failing.
There seems to be no other way to provide a program database without
taking away the users freedom or allowing third parties to access it.
>
> > Otherwise our freedom is being limited because the GPL license in not good
> > enough to protect it.
>
> How is my freedom not being protected ?
If you allow third parties to create interfaces to the gcc, then they
take your freedom to look at the code.
> How does the GPL fall short ?
The GPL does not protect agains RPC,XMLRPC,SOAP,CORBA, or plain old
file level interfaces being non-GPL and non-LGPL.
>
> True freedom respects the beliefs of others, even if they are against
> freedom.
Ok, so you think I can just put this interface via XML under GPL and
not care if it used by third parties to create non-free gcc backends?
>If we truly had a world of "Free Software" There would be no need
> for the GPL, but because "non-free" software, music, literature, etc ... is
> the norm, the Copyleft is needed to play the non-free game.
The copyleft does not cover the interfaces that are needed to create a
program database.
Miek
Anthony Ventimiglia
>> How is my freedom not being protected ?
>
> If you allow third parties to create interfaces to the gcc, then they
> take your freedom to look at the code.
It is anyone's right to release non-free proprietary code, I don't nor do I
expect anyone to have the divine right to access anyone's code. Sure I wish
it was all free, but it's not and I can't force it on anyone.
>> True freedom respects the beliefs of others, even if they are against
>> freedom.
> Ok, so you think I can just put this interface via XML under GPL and
> not care if it used by third parties to create non-free gcc backends?
If you want to promote free software, you have to write free-software.
Michael DuPont
> >>>>On Wednesday 12 June 2002 02:33 am, Michael DuPont boldly stated:
>
>
> > That is crazy. Up to know there is not any support of the GCC group to
> > provide any dynamic linking interfaces or file exchange formats.
> > Becuase of thier worries.
> > RMS said on that list I will dig up the mail, that if anyone wants to
> > link the gcc dynamically, to please contact him privatly to discuss
> > it.
>
> Well have you sent him this proposal? Or contacted him at all?
represents my last stand on this issue.
Why is it that OpenCyc can license data under the LGPL as a
declarative language and all derived facts? How can knowledge bases
and prolog fact tables be licensed? DAML, RDF files? What is it that
we cannot put the intermediate data into such a programming language
and license all derived works? Or is the OpenCyc license invalid?
>
> > That way you can then safely us this data without having anyone fear
> > abuse.
>
> I have no fear of abuse.
The gcc team does.
Read these (many) mails
http://gcc.gnu.org/ml/gcc-patches/2002-04/msg01021.html
-----------------------------------------------------------------------------
>I think there's no chance that this will ever be accepted.
>The only use for this I can see is to have a proprietary
>compiler use the GNU front ends to (1) parse languages that
>the proprietary compiler can't handle or (2) parse GNU C
>extensions common in free software or (3) avoid developing
>a proper GCC machine description for a new processor or
>(4) avoid contributing significant optimizations already
>present in a vendor compiler.
>All of which are goals we neither support nor facilitate.
--------------------------------------
http://gcc.gnu.org/ml/gcc/2002-02/msg01823.html
>> The GCC list has had many people who are interested in
>> getting access to the compiler data, and many of them
>> are turned away in a similar fashion.
>I know. It is a problem, and eventually people will do it, and the
>result will be that people will subvert RMS's intent. If they'd
>figured out how to do this years ago, we'd now have no GNU C++ and
>no GNU Objective-C. In the long run, it's a losing battle, I agree.
-----------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00103.html
>I work on static analisys tool. And would like to use parse tree
produced by
>C++ gcc frontend. The question is did someone did some movements to
store
>parse tree in a file and read it back in a program which is outside
gcc tree?
>I know about possibility to dump into text file, but do not want to
parse it.
------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00166.html
>Isn't this the same discussion which has gotten repeatedly squashed
for
>making it too easy to leverage GCC with a non-GPL'd project via a
>defined interface?
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2002-02/msg01823.html
From Joe Buck
>Yes, that is a concern: proprietary compilers that use the GCC front
ends.
>This is a especially serious risk for languages that are hard to
parse and
>analyze, like C++. There are commercial license-able front ends
>available, like the EDG front end, so any proprietary developers
sniffing
>around would do better to go in that direction, though it will cost.
>That doesn't mean that it would necessarily be ruled out, but we'd
need
>good arguments for how free software can benefit from the capability.
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00198.html
>The organization that owns gcc fears nothing more than making gcc's
front
>ends available as an API. The reason is that gcc is the most
demonstrable
>example of the "GNU public virus" in action: organizations that
originally
>did not want to make their code free made it free because there
wasn't
>another practical choice. That's why GNU has a free C++ and
Objective-C
>compiler today, and that's also why a number of the free back ends
exist.
>RMS feels that if we do what you want, we will quickly see
proprietary
>processor-specific backends that outperform gcc on specific
processors
>(not hard if you make a non-portable compiler) using the GNU front
ends.
>But you are right, there can be advantages.
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00487.html
>There might be a few reasons for that (some of them highly
inflamatory):
>1/ no-one really knows how to tweak the code anyway, and saying it's
not
>a good idea for political reasons let us save face. :-)
>2/ people have started on the work, but gotten so much political
hate-mail
>(`you're going to KILL THE GPL. I hate your guts.') that they
stopped.
>3/ nobody has thought of another way to do that.
>4/ somebody has thought a ludicrously complicated and long and absurd
process
>to do it, that requires gnu-make, and gnu-m4, and libtool, and
gnu-ADA to
>do things, and then got his senses back and thought `Nah, even the
FSF
>wouldn't fall for it'.
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00489.html
>Or people have started the work but refrain from publishing their in
>fear of getting such mails -- which in effect is equivalent to having
>nothing.
>It is an ironical situation that a reading of "free software
>philosophy" prevents free software.
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2000-10/msg00289.html
>> I'm very pleased that someone else here has a clue about freedom.
Its
>> not freedom if some are more "free" than others.
>I am against any notion of trying to encumber file formats, and I can
>assure you that RMS would also reject it, because he's about the most
>consistent person I know.
>As for the rest of your piece, I'd rather not argue political
philosophy
>on a development list. Perhaps we can extend Godwin's law, and also
kill
>discussions when someone writes "statist" rather than "Nazi".
---------------------------------------------------------------------------
http://gcc.gnu.org/ml/gcc/2002-02/msg01808.html
>James Michael DuPont asks lots of questions indicating that he is
>attempting to find a legal way around the GPL, so that he can attach
>proprietary backends to GCC.
>I would urge people not to give him any assistance, even if he is
>only asking the questions out of intellectual curiousity.
---------------------------------------------------------------------------
There are many People Interested in this :
---------------------------------------------------------------------------
From the GUPRO-Project
http://gcc.gnu.org/ml/gcc/2000-02/msg00511.html
From Bellabs ATT:
http://gcc.gnu.org/ml/gcc/2001-10/msg01240.html
From Felix Dechamps
http://gcc.gnu.org/ml/gcc/2002-03/msg00155.html
---------------------------------------------------------------------------
>
> > May people would willling only use this only by GPLed code, and for
> > GPled projects. There is a huge market for work on gpled code.
> > Let me try and explain again :
> >
> >
> > 1. lots of gpled code to maintain.
> > 2. hardly any tools to do code transformations.
> > 3. the tools available dont have good parsers or dont support a raplid
> > development, ide and scripting environment.
> > 4. the gcc provides all the information the ide would need, but you
> > cannot use your freedom to do so without getting baned from the list
> > and getting absolutly no support.
> > 5. the IDEs roll thier own bad parser, and free software looks schizo.
> > 6. Microsoft create lots of hidden interfaces and protected file
> > formats, limit the access to the program database to only thier
> > programs. But the have a great IDE and people are productive because
> > they have embedded the compiler and dont have any problem with using
> > it.
> >
> > So you tell me if noone would use this software, the IDE can be under
> > GPLed, but the interface to the gcc can only be used by GPLed
> > software.
>
> I wouldn't use it, because I want to have the right to put any copyright I
> desire on my software, I want my right to freedom. But I probably wouldn't
> ever use an IDE, because vim is all the IDE I need.
For me it is emacs :)
What about function name completion and a good type browser? How can
you implement that only on vi?
Extracting the public interface of a class programmatically?
Finding all the usages (directly and indirectly) of a type?
The extraction of one function and all the functions is calls from
code, automatically?
The generation of a cleaner representation of code?
All of these things are possible with such an interface. They take
code and an input and spit code as an output. The problem is that you
still need to interface to the compiler to get the annotated source
tree.
>
> > My proposed R-GPL would be PRO GPL, it would allow for the creation of
> > many interesting programs and web services, it would promote free
> > software and foster the creation of many more programs. The only
> > restriction is the restrict that is in place by convention in the
> > gcc, that you dont give the data to thier parties.
>
> Non-free software cannot promote free software.
I am proposing free software. You tell me if there is any way to
provide this functionality without jepordizing the gcc?
>
> >
> > So you ask yourself who is restricting freedom here?
> > It is ok to create aggreements that restrict peoples freedom in the
> > name of protecting the gpl, but is not ok to just put down into a
> > license, however restrictive excatly what is being aggreed. All of you
> > seem to be shocked about what I am proposing, it is just what is
> > happening right now, but not codified.
>
> Like I said before -- it's not ok for a pro lifer to kill abortion doctors.
Read the mails I sent to you and look who is doing the shooting. It is
not me.
Mike
mdupont
"Christopher Browne" <cbbr...@acm.org> wrote in message
news:ae7k31$4q32t$1...@ID-125932.news.dfncis.de...
> Centuries ago, Nostradamus foresaw when "mdupont" <mdupo...@yahoo.com>
would write:
> >> "Building walls" is what the GNU project is aimed at stopping, you
> >> are proposing doing exectly that.
> >
> > OK, so you are saying that I should not care how the data is used
> > and if it is harming the GNU project.
>
> A "safety net" that people refuse to use is not a useful safety net.
>
> And you're proposing building a "safety net" that many will refuse to
> use.
The point is that there is no currently supported was to do this in the GCC
or GNU area.
Should we just ignore the issue and open up gnu software to become shared
XML-RPC libs for non-free software?
Should we not try an provide advanced concepts out of fear?
or should we limit the usage of this data to a small group of people who
want to contribute back?
That is the idea behind the GPL that makes it grow, the sticky license that
keeps people from running away with the ball.
Mike
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> >>>>On Wednesday 12 June 2002 06:09 pm, Michael DuPont boldly stated:
>
> >> How is my freedom not being protected ?
> >
> > If you allow third parties to create interfaces to the gcc, then they
> > take your freedom to look at the code.
>
> It is anyone's right to release non-free proprietary code,
>I don't nor do I
> expect anyone to have the divine right to access anyone's code.
That is the basic freedom of the GPL.
>Sure I wish
> it was all free, but it's not and I can't force it on anyone.
You can by limiting the access to functions to staticly linked libs.
>
> >> True freedom respects the beliefs of others, even if they are against
> >> freedom.
>
> > Ok, so you think I can just put this interface via XML under GPL and
> > not care if it used by third parties to create non-free gcc backends?
>
> If you want to promote free software, you have to write free-software.
Address the point. I am working on free software, my project is under the
GPL,
Bet there is no protection from it being derived from by non-free software.
That is why that FSF and GCC dont like software like it.
Mike
Anthony Ventimiglia
> I have had extensive contact with him on this issue. This proposal
> represents my last stand on this issue.
>
> Why is it that OpenCyc can license data under the LGPL as a
> declarative language and all derived facts? How can knowledge bases
> and prolog fact tables be licensed? DAML, RDF files? What is it that
> we cannot put the intermediate data into such a programming language
> and license all derived works? Or is the OpenCyc license invalid?
I'm gonna have to look that over better before I answer that, but I'll
address it in the new thread you started.
> >> I have no fear of abuse.
>
> The gcc team does.
> Read these (many) mails
Yeah I read them, it seems the biffest fear is attatchin proprietary
backends or frontends to gcc, isn't that pretty much what you are proposing?
And no matter what your intentions, how can you be safely allowed to make a
proprietary piece of code, while every one else is denied that right.
>> I wouldn't use it, because I want to have the right to put any copyright
>> I
>> desire on my software, I want my right to freedom. But I probably
>> wouldn't ever use an IDE, because vim is all the IDE I need.
> For me it is emacs :)
>
> What about function name completion and a good type browser? How can
> you implement that only on vi?
What does vim's command/word completion have to do with anything?
> Extracting the public interface of a class programmatically?
> Finding all the usages (directly and indirectly) of a type?
> The extraction of one function and all the functions is calls from
> code, automatically?
> The generation of a cleaner representation of code?
>
> All of these things are possible with such an interface. They take
> code and an input and spit code as an output. The problem is that you
> still need to interface to the compiler to get the annotated source
> tree.
Or you have to write it yourself, which I agree is againsty the spirit of
the GPL, if you were planning on writing GPL'd softwere.
>> > My proposed R-GPL would be PRO GPL, it would allow for the creation of
>> > many interesting programs and web services, it would promote free
>> > software and foster the creation of many more programs. The only
>> > restriction is the restrict that is in place by convention in the
>> > gcc, that you dont give the data to thier parties.
>>
>> Non-free software cannot promote free software.
> I am proposing free software. You tell me if there is any way to
> provide this functionality without jepordizing the gcc?
There's no way you can resrtict its use in this way and keep it free. The
only way you'd really be able to restrict its use is to not release the
code or give it a non-free license.
>> Like I said before -- it's not ok for a pro lifer to kill abortion
>> doctors.
> Read the mails I sent to you and look who is doing the shooting. It is
> not me.
I don't agree, I think you're a little clouded by your opinions, and so you
feel victimized. I didn't say you're doing the shooting, but what you're
proposing would be.
Anthony Ventimiglia
>> It is anyone's right to release non-free proprietary code,
> If they dont link to GPL software.
But they can use GPL software to build it.
> You can by limiting the access to functions to staticly linked libs.
No I can't because I won't force free software down anyone's throat, and If
I placed that restriction on something, it wouldn't be free.
>> If you want to promote free software, you have to write free-software.
>
> Address the point. I am working on free software, my project is under the
> GPL,
> Bet there is no protection from it being derived from by non-free
> software. That is why that FSF and GCC dont like software like it.
I will address the point, you are proposing a license that would restrict
use to only creatte free software, that is Not-free. Furthermore, ther is
no way you could ever enforce such a restriction without making the code
proprietary, unless you want to start a FREE-SOFTWARE WITCHUNT.
mdupont
> >
> > The gcc team does.
> > Read these (many) mails
>
> Yeah I read them, it seems the biffest fear is attatchin proprietary
> backends or frontends to gcc, isn't that pretty much what you are
proposing?
The new software is not as free as the gpl, but it would be open for anyone
to add to.
I think I see your point.
In order to limit the problem I would make a new form of non-free software,
one that its not free because it limits how it is used.
That is in effect what the gcc team is doing via politics right now.
> And no matter what your intentions, how can you be safely allowed to make
a
> proprietary piece of code, while every one else is denied that right.
I am saying that I would be making a software that is almost free, free
enought to give you access to the
source code, to create derived works under the same license, free to use for
truly free software.
Free to redistribute, free to modify.
But not free to use with other non-free software and not free to exchange
data with non-free software.
So not truly free, but protecting free software.
But you dont have this freedom now in the gcc anyway, so it is in effect non
free by the way they limit the interfaces,
not by license.
>
>
> >> I wouldn't use it, because I want to have the right to put any
copyright
> >> I
> >> desire on my software, I want my right to freedom. But I probably
> >> wouldn't ever use an IDE, because vim is all the IDE I need.
> > For me it is emacs :)
> >
> > What about function name completion and a good type browser? How can
> > you implement that only on vi?
>
> What does vim's command/word completion have to do with anything?
I am saying that in order to get the full list of names to complete, you
need to extract them from the symbol table of the compiler. This word list
is difficult to extract, for example to tell you what identifiers are
defined at a given point in code.
This needs full parsing of the function bodies and code, and can only be
done by the compiler.
Such a feature is what people pay for, and would be willing to give up that
bit of freedom to use. The editor would have to be free, but only the
interface between the compiler and the editor would be RGPL. That means that
all free editors could license the data.
>
> > Extracting the public interface of a class programmatically?
> > Finding all the usages (directly and indirectly) of a type?
> > The extraction of one function and all the functions is calls from
> > code, automatically?
> > The generation of a cleaner representation of code?
> >
> > All of these things are possible with such an interface. They take
> > code and an input and spit code as an output. The problem is that you
> > still need to interface to the compiler to get the annotated source
> > tree.
>
> Or you have to write it yourself, which I agree is againsty the spirit of
> the GPL, if you were planning on writing GPL'd softwere.
Building such and integrated IDE is not possible within the current politics
of the FSF.
>
> >> > My proposed R-GPL would be PRO GPL, it would allow for the creation
of
> >> > many interesting programs and web services, it would promote free
> >> > software and foster the creation of many more programs. The only
> >> > restriction is the restrict that is in place by convention in the
> >> > gcc, that you dont give the data to thier parties.
> >>
> >> Non-free software cannot promote free software.
> > I am proposing free software. You tell me if there is any way to
> > provide this functionality without jepordizing the gcc?
>
> There's no way you can resrtict its use in this way and keep it free.
I see your point.
> The only way you'd really be able to restrict its use is to not release
the
> code or give it a non-free license.
Fine, non free as in, "you cannot use it against other free software. We
take your freedom to drop the bomb."
> >> Like I said before -- it's not ok for a pro lifer to kill abortion
> >> doctors.
> > Read the mails I sent to you and look who is doing the shooting. It is
> > not me.
>
> I don't agree, I think you're a little clouded by your opinions, and so
you
> feel victimized. I didn't say you're doing the shooting, but what you're
> proposing would be.
Well it is reality, and it is what is happening right now. Deal with the
situation.
mike
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> >>>>On Thursday 13 June 2002 03:48 am, mdupont boldly stated:
>
> >> It is anyone's right to release non-free proprietary code,
> > If they dont link to GPL software.
>
> But they can use GPL software to build it.
I am talking about "meta" software here, software that modifies other
software.
Software that deals with other programs as data.
Not normal software.
>
> > You can by limiting the access to functions to staticly linked libs.
>
> No I can't because I won't force free software down anyone's throat, and
If
> I placed that restriction on something, it wouldn't be free.
Well then GPLed code is not free, because you cannot link to it.
>
> >> If you want to promote free software, you have to write free-software.
> >
> > Address the point. I am working on free software, my project is under
the
> > GPL,
> > Bet there is no protection from it being derived from by non-free
> > software. That is why that FSF and GCC dont like software like it.
>
> I will address the point, you are proposing a license that would restrict
> use to only creatte free software, that is Not-free. Furthermore, ther is
> no way you could ever enforce such a restriction without making the code
> proprietary, unless you want to start a FREE-SOFTWARE WITCHUNT.
My proposed license would prevent the publishing and sale of non-free
software that uses RGPL code.
You could use it privatly and inside a company, just not publish the
results.
This would prevent a company from creating a for-sale compiler backend, but
they could experiment with it and use it at work. This is an important
point.
Right now the swagkit is a tool that is not free but uses the GCCs tree
nodes via a file.
When the introspector goes live, you will have a global database of source
code replicated across the globe. All majoect open source projects will
exists in this meta-data repository. With GUIS, query tools, visualization
tools and editors,
all free software. The only thing that is restricted is the backbone, the
infrastructure. This infrastructure will cost money to maintain and keep up.
There is no need to give it away, access can be restricted on a user basis,
if you contribute free software, you can use it otherwise tough.
It would be very dangerous to allow any non-free software to access it
unrestricted.
mike
Frank Tobin
> In order to limit the problem I would make a new form of non-free
> software, one that its not free because it limits how it is used. That
> is in effect what the gcc team is doing via politics right now.
Note that it seems that the GCC team is limiting itself via technical
means to achieve political goals (which leaves a bad taste in my mouth).
With the proposed RGPL, one is not limiting anything via technical, means,
but by political means (a more restrictive license), to achieve the same
goals.
The goals are the same, the means are different.
Then there is the issue of whether the RGPL itself is 'free'. Well, I
think much is getting hung up on the semantics of the word 'free' in this
discussion, and whether or not certain software belongs to the group
'free' and whether it promotes other 'free' software .Sure, there is the
FSF concept of 'free', the BSD concept of 'free', and the public domain
concept of 'free'. Each of these, except for public domain, places
limitations, rougly on a gradient. If we put these licenses on a line
roughly according to their restrictiveness, including the RGPL:
Restrictive -> Unrestrictive
RGPL -> GPL -> LGPL -> BSD -> Public domain
Now, in my mind, there is no 'real' reason that the RGPL couldn't also be
called 'free'. As long as it was tweaked to say that it could be used to
work on other RGPL software (so as to include this entire range of
licenses). Of course, this provides a sort of circular definition (the
RGPL allows you to work on RGPL software), but this is okay if you think
about it.
The only thing that worries me about the RGPL is that it doesn't 'play
friendly' with the 'other' 'free' licenses (those less restrictive than
the GPL). To some extent the GPL does play friendly with them, because
the independent works keep their own license, but the 'combined' work is
under the GPL. However, I'm not convinced this difference has any
practical merit.
For what its worth, this discussion highlights why I refuse to talk about
software 'freedom'; the semantics are just poor, and there are several
possible consistent definitions you can apply to 'free'/libre.
--
Frank Tobin http://www.neverending.org/~ftobin/
Anthony Ventimiglia
>> And no matter what your intentions, how can you be safely allowed to make
> a
>> proprietary piece of code, while every one else is denied that right.
> I am saying that I would be making a software that is almost free, free
> enought to give you access to the
> source code, to create derived works under the same license, free to use
> for truly free software.
> Free to redistribute, free to modify.
Ah but there is no almost free, free is like being pregnant or dead. You
can't be a little pregnent or near death. You either are or you aren't.
> But not free to use with other non-free software and not free to exchange
> data with non-free software.
> So not truly free, but protecting free software.
That's just a pipe dream, there's no possible way you can release the code
and restrict the modification. In order for this license to be effective,
you would need to restrict any modifications to also keep the software
restricted to only produce GPL'd code, how can you possibly enforce that ?
>> > What about function name completion and a good type browser? How can
>> > you implement that only on vi?
>>
>> What does vim's command/word completion have to do with anything?
> I am saying that in order to get the full list of names to complete, you
> need to extract them from the symbol table of the compiler. This word list
> is difficult to extract, for example to tell you what identifiers are
> defined at a given point in code.
> This needs full parsing of the function bodies and code, and can only be
> done by the compiler.
> Such a feature is what people pay for, and would be willing to give up
> that bit of freedom to use. The editor would have to be free, but only the
> interface between the compiler and the editor would be RGPL. That means
> that all free editors could license the data.
Well we must be talking about two different things, because vim uses open
files and header files to get the list of commands to complete. It doesn't
get anything from the compiler other than output from a make command.
>> Or you have to write it yourself, which I agree is againsty the spirit of
>> the GPL, if you were planning on writing GPL'd softwere.
>
> Building such and integrated IDE is not possible within the current
> politics of the FSF.
Is that because of your desire to restrict it's use?
And you have to realize the politics of the FSF are their politics, there
have been some, such as Larry Wall who feel the GPL is too restrictive, and
this could be an example of why, but nevertheless, they are the FSF's
politics, and I don't expect them to change.
>> I don't agree, I think you're a little clouded by your opinions, and so
> you
>> feel victimized. I didn't say you're doing the shooting, but what you're
>> proposing would be.
> Well it is reality, and it is what is happening right now. Deal with the
> situation.
Fair enough.
Anthony Ventimiglia
>> >> It is anyone's right to release non-free proprietary code,
>> > If they dont link to GPL software.
>>
>> But they can use GPL software to build it.
> But they cannot link to it.
Ok we're going back and forth here.
> I am talking about "meta" software here, software that modifies other
> software.
> Software that deals with other programs as data.
> Not normal software.
I understand, and it is a whole different issue, however I just don't see
how you can restrict it's use.
>> I will address the point, you are proposing a license that would restrict
>> use to only creatte free software, that is Not-free. Furthermore, ther is
>> no way you could ever enforce such a restriction without making the code
>> proprietary, unless you want to start a FREE-SOFTWARE WITCHUNT.
> My proposed license would prevent the publishing and sale of non-free
> software that uses RGPL code.
> You could use it privatly and inside a company, just not publish the
> results.
> This would prevent a company from creating a for-sale compiler backend,
> but they could experiment with it and use it at work. This is an important
> point.
Ok I understand, and the proposal of the RGPL is like the antithesis of th
LGPL. But here's a real reason why the FSF will reject it and certainly
would not want to hold the copyright:
The LGPL essentially reduces the amount of policing that the FSF will have
to do regarding FSF owned copyright infringement. Now if the FSF adopted a
RGPL, and then decided to actually take copywritten software under the
RGPL, their work to police Copyright infringement would increase many times
over. Think about it, with a license like that, the copyright holder could
spend all their time tracking non-free software trying to determine if it
was made with RGPL software. I'm pretty sure the FSF doesn't have the
financial resources to start a witchhunt like that, and I doubt any
independent developer would ever have the money or time to police their
copyright.
Then there would be the whole public image of the "Free" Software
Foundation tracking down and filing lawsuits against all these companies
who have violated their RGPL copyrights. Sure would take away the
impression of keeping things Free.
> Right now the swagkit is a tool that is not free but uses the GCCs tree
> nodes via a file.
> When the introspector goes live, you will have a global database of source
> code replicated across the globe. All majoect open source projects will
> exists in this meta-data repository. With GUIS, query tools, visualization
> tools and editors,
> all free software. The only thing that is restricted is the backbone, the
> infrastructure. This infrastructure will cost money to maintain and keep
> up.
>
> There is no need to give it away, access can be restricted on a user
> basis, if you contribute free software, you can use it otherwise tough.
Well that's not too good, what about people who write free software, as
well as non-free software, do you revoke their rights? How about the people
who want to start writing free software? They can't have access until they
contribute, hardly free is it ?
I think the Global database of sourcecode we have right now is just fine.
It sounds like your proposal is pretty centralized, which isn't really a
good way to preserve freedom. That's one of the reasons the Savannah GNU
project was started, because of the centralized nature of Sourceforge.
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> >>>>On Thursday 13 June 2002 10:58 am, mdupont boldly stated:
>
> >> And no matter what your intentions, how can you be safely allowed to
make
> > a
> >> proprietary piece of code, while every one else is denied that right.
> > I am saying that I would be making a software that is almost free, free
> > enought to give you access to the
> > source code, to create derived works under the same license, free to use
> > for truly free software.
> > Free to redistribute, free to modify.
>
> Ah but there is no almost free, free is like being pregnant or dead. You
> can't be a little pregnent or near death. You either are or you aren't.
>
> > But not free to use with other non-free software and not free to
exchange
> > data with non-free software.
> > So not truly free, but protecting free software.
>
> That's just a pipe dream, there's no possible way you can release the code
> and restrict the modification.
Restrict the usage. Not the modification.
> In order for this license to be effective,
> you would need to restrict any modifications to also keep the software
> restricted to only produce GPL'd code, how can you possibly enforce that ?
How does the microsoft shared source work? What about QT/Free and QT-for
money?
What about MSQL?
>
> >> > What about function name completion and a good type browser? How can
> >> > you implement that only on vi?
> >>
> >> What does vim's command/word completion have to do with anything?
> > I am saying that in order to get the full list of names to complete, you
> > need to extract them from the symbol table of the compiler. This word
list
> > is difficult to extract, for example to tell you what identifiers are
> > defined at a given point in code.
> > This needs full parsing of the function bodies and code, and can only be
> > done by the compiler.
> > Such a feature is what people pay for, and would be willing to give up
> > that bit of freedom to use. The editor would have to be free, but only
the
> > interface between the compiler and the editor would be RGPL. That means
> > that all free editors could license the data.
>
> Well we must be talking about two different things, because vim uses open
> files and header files to get the list of commands to complete. It doesn't
> get anything from the compiler other than output from a make command.
But it cannot provide these features without implementing a c parser
and symanantic analysis, so you do not have it.
I bet you wont ever have it under the current licensing even if you could
becuase anyone
who wanted to use the gcc to do so would get attacked.
>
> >> Or you have to write it yourself, which I agree is againsty the spirit
of
> >> the GPL, if you were planning on writing GPL'd softwere.
> >
> > Building such and integrated IDE is not possible within the current
> > politics of the FSF.
>
> Is that because of your desire to restrict it's use?
This has nothing to do with me. This is an old topic.
> And you have to realize the politics of the FSF are their politics, there
> have been some, such as Larry Wall who feel the GPL is too restrictive,
and
> this could be an example of why, but nevertheless, they are the FSF's
> politics, and I don't expect them to change.
Well the Perl community is much less paranoid. I really like perl.
I had my code under the PAL to begin with, and moved to GPL later.
Now I am wondering if I should not distribute the code at all and only
provide a webservice to registers GNU users.
The GPL2.0 cannot force you to distribute your changes and code of a web
service.
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
People will police it, it is not that much work. The amount of abuse right
now is limited.
This will close the hole and make it pretty clear to anyone who thinks they
are clever that we are on to them.
>
> Then there would be the whole public image of the "Free" Software
> Foundation tracking down and filing lawsuits against all these companies
> who have violated their RGPL copyrights. Sure would take away the
> impression of keeping things Free.
It is very simple, this would eliminate piggybacking of non-free software on
RGPLed software. Very simple.
Nonone else would really be affected.
>
> > Right now the swagkit is a tool that is not free but uses the GCCs tree
> > nodes via a file.
> > When the introspector goes live, you will have a global database of
source
> > code replicated across the globe. All majoect open source projects will
> > exists in this meta-data repository. With GUIS, query tools,
visualization
> > tools and editors,
> > all free software. The only thing that is restricted is the backbone,
the
> > infrastructure. This infrastructure will cost money to maintain and keep
> > up.
> >
> > There is no need to give it away, access can be restricted on a user
> > basis, if you contribute free software, you can use it otherwise tough.
>
> Well that's not too good, what about people who write free software, as
> well as non-free software, do you revoke their rights?
They cannot use this for non-free software publically.
> How about the people
> who want to start writing free software?
No, they cannot publish works that use it.
>They can't have access until they
> contribute, hardly free is it ?
They can use it all they want in private.
>
> I think the Global database of sourcecode we have right now is just fine.
It is not parsed, it is not cross referenced, you cannot process it
automatically, cannot query it.
Who uses this? What does that use? What is this? What do I need to call that
function?
What are examples of this function being used?
> It sounds like your proposal is pretty centralized, which isn't really a
> good way to preserve freedom.
It is not, it could be distributed easily. Each project would host its own
softwares meta-data.
>That's one of the reasons the Savannah GNU
> project was started, because of the centralized nature of Sourceforge.
Well I have been in contact with them already.
mdupont
The next issue is that the interfaces that have to be GPL because they are
connected to the GCC
could be limited to webservices under the GPLv2, then we can limit the
access to the modified compiler to registered users.
This would keep people from redistributing the compiler with the patches.
"Frank Tobin" <fto...@neverending.org> wrote in message
news:Pine.LNX.4.44.02061...@palanthas.neverending.org...
> On Thu, 13 Jun 2002, mdupont wrote:
>
> > In order to limit the problem I would make a new form of non-free
> > software, one that its not free because it limits how it is used. That
> > is in effect what the gcc team is doing via politics right now.
>
> Note that it seems that the GCC team is limiting itself via technical
> means to achieve political goals (which leaves a bad taste in my mouth).
> With the proposed RGPL, one is not limiting anything via technical, means,
> but by political means (a more restrictive license), to achieve the same
> goals.
>
> The goals are the same, the means are different.
Yes, you summed it all up.
>
> Then there is the issue of whether the RGPL itself is 'free'. Well, I
> think much is getting hung up on the semantics of the word 'free' in this
> discussion, and whether or not certain software belongs to the group
> 'free' and whether it promotes other 'free' software .Sure, there is the
> FSF concept of 'free', the BSD concept of 'free', and the public domain
> concept of 'free'. Each of these, except for public domain, places
> limitations, rougly on a gradient. If we put these licenses on a line
> roughly according to their restrictiveness, including the RGPL:
>
> Restrictive -> Unrestrictive
> RGPL -> GPL -> LGPL -> BSD -> Public domain
>
> Now, in my mind, there is no 'real' reason that the RGPL couldn't also be
> called 'free'. As long as it was tweaked to say that it could be used to
> work on other RGPL software (so as to include this entire range of
> licenses). Of course, this provides a sort of circular definition (the
> RGPL allows you to work on RGPL software), but this is okay if you think
> about it.
Thanks, you get my point. It is another level of software, a different
dimension if you will.
>
> The only thing that worries me about the RGPL is that it doesn't 'play
> friendly' with the 'other' 'free' licenses (those less restrictive than
> the GPL). To some extent the GPL does play friendly with them, because
> the independent works keep their own license, but the 'combined' work is
> under the GPL. However, I'm not convinced this difference has any
> practical merit.
The scope of the RGPLed software is limited.
You wont be able to create artistic licensed code with it, but you can use
artistic license.
> For what its worth, this discussion highlights why I refuse to talk about
> software 'freedom'; the semantics are just poor, and there are several
> possible consistent definitions you can apply to 'free'/libre.
The free issue is getting boring. The question is if this is legal and
doable?
mike
Anthony Ventimiglia
It's not so much the work of policing it, but the expense of enforcing the
copyrights legelly. I'm sure sooner or later the terms would be violated by
a multimillion dollar software corporation that would keep the whole issue
tied up in courts for years exhausting all the expenses of a party pursuing
a copyright infringement.
The amount right now is limited, but what about in the future.
>> Well that's not too good, what about people who write free software, as
>> well as non-free software, do you revoke their rights?
> They cannot use this for non-free software publically.
>
>> How about the people
>> who want to start writing free software?
> No, they cannot publish works that use it.
>
>>They can't have access until they
>> contribute, hardly free is it ?
> They can use it all they want in private.
So until they contribute, they wouldn't be allowed to publicy use the
source from the database. So can their initial contribution be derived from
this code, or does it have to be totally original? If that's the case then
who will decide if their contribution is worthy of being accepted?
If you are forcing someone to contribute, then won't you start getting a
duplication of already made efforts. Look at the majority of projects on
sourceforge, personally I see at least two-thirds (and that's a
conservative estimate) of the projects are reinventing the wheel.
I see the idea you have here and I agree it would be a wonderful thing if
we could quickly have references to find certain libraries and utilites to
facilitate programming. I'd definitley say one of the hardest things for me
has been to find the right libraries to do the job, partially because there
are often many libraries which serve the same purpose. Often I find that
the majority of these libraries may be poorly written or incomplete.
There's a new project on sourceforge (http://libs.sourceforge.net/) with a
similar idea, only they are just trying to build a database of information
on libraries, not actually containing the code as you propose. However the
real goal of this should be to refine existing libraries and codes, and
discouraging the creation of more libraries which serve the same purpose of
existing libraries.
Mike,
I think we can agree to disagree on some of this stuff, I'm sure by now you
have a good idea of what my opinion is, which was your original goal of
this thread.
mdupont
If you want to really protect the GPLed code and not expose the interfaces
that you used to extract the data,
then you need to make private changes to the GPLed code and not publish
them.
Only the programs results via a web service.
I bet that no-one will like it,but here she comes :
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
news:NznO8.32541$MW2.1...@news4.srv.hcvlny.cv.net...
> >>>>On Friday 14 June 2002 06:05 am, mdupont boldly stated:
> >> Well that's not too good, what about people who write free software, as
> >> well as non-free software, do you revoke their rights?
> > They cannot use this for non-free software publically.
> >
> >> How about the people
> >> who want to start writing free software?
> > No, they cannot publish works that use it.
> >
> >>They can't have access until they
> >> contribute, hardly free is it ?
> > They can use it all they want in private.
>
> So until they contribute, they wouldn't be allowed to publicy use the
> source from the database. So can their initial contribution be derived
from
> this code, or does it have to be totally original?
The code itself is GPL, the way you connect to the server is RGPL.
An GPled application might be an HTML gateway that allows for
simple browsing. Then the source code would be GPLed in html form.
But the structure of the HTML that contains all the cross references would
be RGPLEd declarative language source code.
You could not publish a non-free compiler that reads html from a RGPLEd
application and uses it for eliminating parsing.
>If that's the case then
> who will decide if their contribution is worthy of being accepted?
As I said, GPLed projects.
This gets into the nasty bits of protecting the changes to the GPL code from
the public to implement the RGPLed interfaces.
Who would be able to run a server,and get the source code?
Assuming that the modifications to the GPLed compilers would cause SO MUCH
stress to the authors, then they can only be
distributed privatly under a NDA and using a encrypted data exchange system.
We would limit access to the server source code to a few groups with a
specific license aggreement.
The server license would limit the distribution of the source code
willy-nilly. The server would contain all the modified GPL compilers under
the v2 gpl and the changes would not have to be distributed.
The GCC group would not have to worry about us publishing the changes
because no one can access the modified compiler except via an RGPLed web
service.
The aggreement would be not to distribute it to third parties without this
aggreement,
and to verify that the users are in the list of registered user.
Kind of like the debian keyring.
This might sound quite crazy, but it makes good sense.
We might have to create a company or club that requires all the users that
run a server to be member of the group. This would allow for the
non-publication of the source code.
It is fully legal under the GPL to distribute the changes inside a company
or closed user group.
I bet I will get lots of hate mail on this one, but it would at least make
it difficult for non-free derived works from the GCC. That might be a higher
goal then promoting only free software.
There are allusions to this happening ANYWAY on the gcc mailing list. Not
all things are published, people are free to exchange patches as they
please.
Mike
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
How can MYSQL limit useage and still be GPL?
> >> Well that's not too good, what about people who write free software, as
> >> well as non-free software, do you revoke their rights?
> > They cannot use this for non-free software publically.
> >
> >> How about the people
> >> who want to start writing free software?
> > No, they cannot publish works that use it.
> >
> >>They can't have access until they
> >> contribute, hardly free is it ?
> > They can use it all they want in private.
>
> So until they contribute, they wouldn't be allowed to publicy use the
> source from the database.
The contribution of source code to the database? The source would have to be
GPL to be contributed.
Of tools to the system? The tools would have to be RGPL if they are to link
to the RGPL.
They can use the source in the database, just not via non-free tools, but
via the RGPL tools.
> If you are forcing someone to contribute, then won't you start getting a
> duplication of already made efforts. Look at the majority of projects on
> sourceforge, personally I see at least two-thirds (and that's a
> conservative estimate) of the projects are reinventing the wheel.
>
> I see the idea you have here and I agree it would be a wonderful thing if
> we could quickly have references to find certain libraries and utilites to
> facilitate programming.
And refactor programs by applying code transformation tools.
> I'd definitley say one of the hardest things for me
> has been to find the right libraries to do the job, partially because
there
> are often many libraries which serve the same purpose. Often I find that
> the majority of these libraries may be poorly written or incomplete.
>
> There's a new project on sourceforge (http://libs.sourceforge.net/) with a
> similar idea, only they are just trying to build a database of information
> on libraries, not actually containing the code as you propose. However the
> real goal of this should be to refine existing libraries and codes, and
That is right, by building a repository we can then tranform them.
> discouraging the creation of more libraries which serve the same purpose
of
> existing libraries.
That is good, but we need to contain all the function bodies to get the
examples of usage, the
way they are used, the dependencies.
>
> Mike,
>
> I think we can agree to disagree on some of this stuff, I'm sure by now
you
> have a good idea of what my opinion is, which was your original goal of
> this thread.
Yes, I have been arguing a very provokative topic as well.
All is fair and your comments were taken in perspective.
Anthony Ventimiglia
>>
>> It's not so much the work of policing it, but the expense of enforcing
>> the copyrights legelly. I'm sure sooner or later the terms would be
>> violated
> by
>> a multimillion dollar software corporation that would keep the whole
>> issue tied up in courts for years exhausting all the expenses of a party
> pursuing
>> a copyright infringement.
>>
>> The amount right now is limited, but what about in the future.
> That is why I am bringing this up here and now. Any ideas?
> How can MYSQL limit useage and still be GPL?
I don't know, but I don't feel like it's my responsibility to make sure
people don't break the MySQL Licencse by using is commercially without
buying the commercial license. It's really up to MySQL to police their own
stuff.
The Qt license carries a similar dual license, covered by the GPL for
non-commercial Linux use and a commercial license for commercial use. I
guess in this case it is also up to Trolltech to police their software.
The trooltech approach to this problem may be very similar to yours, They
finally agreed to use the GPL from pressure by many people, but obviously
wanted to make sure they would be able to profit from it's commercial use.
As a consequence, you can't get the Qt libraries for Windows, not even
Cygwin without buying a commercial license. I guess their belief was that
the only Free software development using Qt would be on a *nix platform.
I guess the answer is that technically they are only GPL when used for
non-commercial use. I would say the advantage of having a dual license over
a license that restricts non-free use is there's a quick remedy for
violation of the GPL. In other words if MySQL is found to be used
commercially, they can just settle it by buying a license. Where if you
tell them they can't use it at all for non-free use, then you'll have a
tougher time settling a copyright dispute.
>> I think we can agree to disagree on some of this stuff, I'm sure by now
> you
>> have a good idea of what my opinion is, which was your original goal of
>> this thread.
> Yes, I have been arguing a very provokative topic as well.
> All is fair and your comments were taken in perspective.
It's certainly been fun.
Jeffrey Siegal
> How can MYSQL limit useage and still be GPL?
MySQL usage is not limited.
Anthony Ventimiglia
I'm sorry, you got my head spinning now, I can't think about this anymore.
Anthony Ventimiglia
> mdupont wrote:
>> How can MYSQL limit useage and still be GPL?
>
> MySQL usage is not limited.
I'm not sure what your definition of limited is, but I think you should
read the licensing Policy before you say it's not limited:
Michael DuPont
> I don't know, but I don't feel like it's my responsibility to make sure
> people don't break the MySQL Licencse by using is commercially without
> buying the commercial license. It's really up to MySQL to police their own
> stuff.
There is an initiative call DOTGNU to produce a secure web service
with authentication and lots of network options. If they cannot
provide the features to protect software from abuse, then I dont know.
> The trooltech approach to this problem may be very similar to yours, They
Just opposite, I want the software to be free, but the pressure from
the GNU community is making me be more restrictive.
> commercially, they can just settle it by buying a license. Where if you
> tell them they can't use it at all for non-free use, then you'll have a
> tougher time settling a copyright dispute.
It is simple I think, we need to give the FSF a legal method on which
they can start a case. Right now they have nothing.
If we come up with a solution that can protect the software just from
explotation, and If people start to try and create non-free derived
works, then we can start looking into it.
The policing is not an issue, because when a violation occurs, it
could take a while to find it, but if a user is just using the data,
we can make sure that they dont distribute the Gpled software
publically.
There is a issue that someone might create a non-free module and just
distribute it according to the file formats.
The users who could download the software from a third party and then
violate the GPl on thier own, well that could get interesting.
Given XML files, I dont think that there is a way to stop that.
That would be an issue, only by taking away the users freedoms
completly would we be able to stop that.
Then we are back to square one..... I will have to think about this.
mike
Michael DuPont
> mdupont wrote:
> > How can MYSQL limit useage and still be GPL?
>
> MySQL usage is not limited.
>You have a commercial application that ONLY works with MySQL and
ships the >application with the MySQL server. This is because we view
this as linking >even if it is done over the network.
Linking and function invocation is usage. Limiting that is limiting
usage.
mike
Sam Holden
Offering a commercial license to those who don't wish to comply with the GPL
is not a 'limitation'. It increases the number of things you are allowed to do.
--
Sam Holden
Sam Holden
But that isn't what the are doing.
They are offering an additional choice. If you need to 'link' then the GPL
requires that your code be GPL'd as well, if you don't wish to do that then
you can get a commercial license from them. If you don't want to get a
commercial license then you can just GPL your code and everything is fine.
That is not a limitation of usage. That is an expansion of your options.
If you disagree with their view regarding what is linking then your
interpretation of what the GPL allows and what they think it allows are
different. That doesn't stop you acting according to your view, but you should
expect to end up in court over it (but if you (and your lawyer) think your view
is valid then you would expect to win anyway).
--
Sam Holden
Jeffrey Siegal
> If you disagree with their view regarding what is linking then your
> interpretation of what the GPL allows and what they think it allows are
> different. That doesn't stop you acting according to your view, but you should
> expect to end up in court over it (but if you (and your lawyer) think your view
> is valid then you would expect to win anyway).
I think you'd have a hard if not impossible time ever imposing your view
on them in court since you have been clearly informed ahead of time what
their view is. Your choices are to accept the license terms as they've
explained them, or reject them and have no license at all.
The real issue, then, is whether their interpretation of the GPL might
be incompatible with someone else's interpretation of the GPL,
preventing MySQL from being linked with other GPLed code.
Anthony Ventimiglia
> There is an initiative call DOTGNU to produce a secure web service
> with authentication and lots of network options. If they cannot
> provide the features to protect software from abuse, then I dont know.
Yeah I know, it's the GNU answer to .NET.
> It is simple I think, we need to give the FSF a legal method on which
> they can start a case. Right now they have nothing.
Sure they do, but the FSF only protects software in which they hold the
copyright.
> If we come up with a solution that can protect the software just from
> explotation, and If people start to try and create non-free derived
> works, then we can start looking into it.
>
> The policing is not an issue, because when a violation occurs, it
> could take a while to find it, but if a user is just using the data,
> we can make sure that they dont distribute the Gpled software
> publically.
>
> There is a issue that someone might create a non-free module and just
> distribute it according to the file formats.
>
> The users who could download the software from a third party and then
> violate the GPl on thier own, well that could get interesting.
>
> Given XML files, I dont think that there is a way to stop that.
>
> That would be an issue, only by taking away the users freedoms
> completly would we be able to stop that.
>
> Then we are back to square one..... I will have to think about this.
You also have to remember that, AFAIK the GPL has never been challenged
legally, so I don't know if it's such a good idea to possibly write
something that will be expected to go to court one day If the GPL is one
day defeated in a court of law, this whole thing changes.
> mike
Anthony Ventimiglia
I don't know if I wouldn't call it a limitation, rather a comprimise. I
also think that it's an easier option rather than trying to bar all
commercial use of the software.
Michael DuPont
one that contains data about other software.
Specifically I am talking about licensing a changed GCC embedded via
and XML interface and how to protect that XML interface from use by
others.
I think that the only way to protect these changes to the gcc is not
to publish them at all, only via a webservice and only for registered
users for GPLed source code. That does not need any special license.
The XML dumps published would then being only from GPL code and
containing either the entire source code in an xml annotated form or
just the annonotated parse tree with links to the original.
The PARSE tree as a program unto itself could then be put under the
RGPL as declarative data file, or just left under the GPL.
The RGPL is limit the types of software that it will process to
only free ("GPL") software. The data contents would then be
tranformations on GPLed code.
Now when we add all types of parse tree data and symantic analysis to
the XML/HTML data, we can say that we are adding information.
Lets say that we just reference the original source code without
containing one bit of the original file. The XML document just
contains links to the original source code. The parse trees and
symantic type information is gained out of lots of work from the
original file. This output is then stored in a declarative language
(DAML or RDF or XML) for execution.
Is that parse tree still the copyright of the author of the original
text file, or is an creative work from the compiler author?
Lets say I do not distribute the code that produces this tree, but
only the output with a reference to the original code. Can I copyright
the Parse tree under the RGPL? So that any transformations of it are
then still RGPLed?
> > It is simple I think, we need to give the FSF a legal method on which
> > they can start a case. Right now they have nothing.
>
> Sure they do, but the FSF only protects software in which they hold the
> copyright.
I mean that right now the FSF has no legal means against the usage of
the gcc that way. Via my RPLed interface they would have a starting
point.
Ok, lets take the RGPL one more time.
If the license can be implemented, and usage of RGPL software can be
limited to GPLed code. Someone goes against it, then the author will
have at least a starting point for complaining.
The XML interface between GPL and RGPL can be used by anyone without
license, therefore any GPLd data transfer to an RGPL system has to be
over a secure channel, and under the GPLv2. That way the GPLed code is
not distributed, only the interface via RGPL is published via web
services.
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
So then the GPL is the one that limits the usage without forcing the GPL on
the clients.
It is perfectly legit to then use the RGPL to give them less rights.
mike
mdupont
to make it understandable.
Basically I am going the next step into less and less freedom for the users.
The final result is a very restrictive system that basically represents the
current situation but has a legal foundation.
This is the only way to protect the GCC from abuse and still produce this
project.
It is legal and doable with only the addition of the RGPL for handling of
self-replicating declarative language files.
You have the following components :
1. GPLed source code with patches to dump the data to non-licensed XML.
2. Patches are not made available to the public, only the results of
that.
3. XML extractors that read that data into an RGPLed service, the
program is only ran by me so there is no question of public display.
4. The data output from the service is copyrighted by me, and put under
the RGPL license.
5. Restrictive license aggreements for submitters of input data to
aggree that all output is licensed that way.
6. Any execution of the RGPLed software results in source code that is
also RGPL.
7. All data from the RGPLed software is outputed as RGPLed source code
in a declarative form.
8. All data input into the RGPLed software is RGPLed source code in a
declarative form.
9. Any usage of the RGPLed software is only allowed by GPled software,
that is execution and linking.
10. The source code in RGPLe declarative language remains RPLed
even if it is carried by and processed by the GPLEd software.
It remains RGPL because the software cannot change the license of
the data.
That basically summs it up.
"mdupont" <mdupo...@yahoo.com> wrote in message
news:aed4et$rq5$1...@quimby.gnus.org...
> This is were it gets ugly!
> If you want to really protect the GPLed code and not expose the interfaces
> that you used to extract the data,
> then you need to make private changes to the GPLed code and not publish
> them.
In order to copyright the output of a program, it has to be run by the
copyright holder.
If I want to copyright the parse trees of the gcc for a peice of software, I
need to run that myself.
The output would be then edited and declared a creative work from me. That
can then be put under the RGPL.
To prevent the usage of that software by others, the changes to the gcc
should not be published.
The GPL v2 allows for web services to not publish thier sources.
> Only the programs results via a web service.
The program could also be ran as a web service,
the gcc in the background with the agreement that all output is RGPLEd.
> > >> Well that's not too good, what about people who write free software,
as
> > >> well as non-free software, do you revoke their rights?
> > > They cannot use this for non-free software publically.
> > >
> > >> How about the people
> > >> who want to start writing free software?
> > > No, they cannot publish works that use it.
> > >
> > >>They can't have access until they
> > >> contribute, hardly free is it ?
> > > They can use it all they want in private.
> >
> > So until they contribute, they wouldn't be allowed to publicy use the
> > source from the database. So can their initial contribution be derived
> from
> > this code, or does it have to be totally original?
> The code itself is GPL, the way you connect to the server is RGPL.
The data in the database would be GPLed.
The database itself RGPLEd.
You cannot connect to the database without using GPled software. All query
results would be RGPL.
The contained data is then GPL.
> An GPled application might be an HTML gateway that allows for
> simple browsing. Then the source code would be GPLed in html form.
> But the structure of the HTML that contains all the cross references would
> be RGPLEd declarative language source code.
>
> You could not publish a non-free compiler that reads html from a RGPLEd
> application and uses it for eliminating parsing.
>
> >If that's the case then
> > who will decide if their contribution is worthy of being accepted?
> As I said, GPLed projects.
I would decide, with an aggreement from the submitter that they accept the
conditions.
Jay Maynard
>Basically I am going the next step into less and less freedom for the users.
>The final result is a very restrictive system that basically represents the
>current situation but has a legal foundation.
>This is the only way to protect the GCC from abuse and still produce this
>project.
You know, it's been funny, watching all of the GPV apologists who've been
claiming for years that the only way to preserve freedom is to take it away,
argue against someone here who's taking that argument one step further...
and I fully expect to see more of the same.
Phillip Lord
Jay> You know, it's been funny, watching all of the GPV apologists
Jay> who've been claiming for years that the only way to preserve
Jay> freedom is to take it away
I think the word "apologist" is wrong here. Most of those who you are
describing are perfectly happy with the freedoms that GPL gives, and
also perfectly happy with those that it does not. No real apology
needed.
Phil
mdupont
"Jay Maynard" <jmay...@thebrain.conmicro.cx> wrote in message
news:slrnagrjto....@thebrain.conmicro.cx...
It is the classic Owellian Animal Farm Revolution,
the "have nots" come to power and then become just as restrictive if not
more so than before for futher development.
The software development tools that we have under the GPL are in the stone
age, and
there is no politically correct way to fix it, or legal way to protect the
new development under the GPL.
In the end we spend more time playing lawer than hacking.
It is time that we come to terms with the situation,
the GPL is not up to date, it needs critical review,
not religious worship.
I am pretty pissed off with the hypocrisy of the GNU crowd,
it is alright to violate the GPL if it fits them, but not if they dont like
your project.
I asked for permission to use the GCC to make things better,
questions that many do not ask, and get negative answers.
Others do not ask, they just do so, and noone says anything.
Very inconsistent.
Mike
Anthony Ventimiglia
>> > Offering a commercial license to those who don't wish to comply with
>> > the GPL is not a 'limitation'. It increases the number of things you
>> > are allowed to do.
>>
>> I don't know if I wouldn't call it a limitation, rather a comprimise. I
>> also think that it's an easier option rather than trying to bar all
>> commercial use of the software.
>
> So then the GPL is the one that limits the usage without forcing the GPL
> on the clients.
> It is perfectly legit to then use the RGPL to give them less rights.
Ok, if you want it that way, but with the RGPL, there's no option for
comprimise as with the LGPL. You see in a way the LGPL reduces the amount
of disputes by offering an answer to those who wish to use it in non-free
software.
Anthony Ventimiglia
> It is the classic Owellian Animal Farm Revolution,
> the "have nots" come to power and then become just as restrictive if not
> more so than before for futher development.
Well, in this case I'm not too sure who the pigs are? You obviously thing
it's the GNU, while I think your proposal represents the restriction.
> The software development tools that we have under the GPL are in the stone
> age, and
> there is no politically correct way to fix it, or legal way to protect the
> new development under the GPL.
>
> In the end we spend more time playing lawer than hacking.
>
> It is time that we come to terms with the situation,
> the GPL is not up to date, it needs critical review,
> not religious worship.
I worship nothing on papar, (or electronic media for that matter). But I
respect what it says, and I don't expect it to be rewritten to satisfy my
needs.
> I am pretty pissed off with the hypocrisy of the GNU crowd,
> it is alright to violate the GPL if it fits them, but not if they dont
> like your project.
I think that should be taken up with the people you're pissed off at.
> I asked for permission to use the GCC to make things better,
> questions that many do not ask, and get negative answers.
>
> Others do not ask, they just do so, and noone says anything.
Well then less talk more work True freedom == Anarchy, the GPL is just
there, because we have governments we'll never be rid of.
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
are offering no comprise either.
This RGPL will plug that whole and give a legal basis for creating a great
set of tools for software developers.
mike
mdupont
"Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> >>>>On Monday 17 June 2002 08:23 am, mdupont boldly stated:
>
> > It is the classic Owellian Animal Farm Revolution,
> > the "have nots" come to power and then become just as restrictive if not
> > more so than before for futher development.
>
> Well, in this case I'm not too sure who the pigs are? You obviously thing
> it's the GNU, while I think your proposal represents the restriction.
My restriction is a codification of the restriction that is imposed DEFACTO
by the GCC and FSF.
It is just a abstraction that represents the TRUE nature of the GCC license,
not what is on paper, but what is reality,
it is not free to modify and not free to extend in any direction.
By the way that the APIs have been not documented, exposed and opened to
scripting languages, C++ and repositories, it is very very difficult to use
the AST data freely.
More difficult than it needs to be, that is the true sad story.
I think that much money is spent on gcc via big companies,
the small guy like me is easy to squeeze out with stupid accusations and BS
arguements.
You wont believe how many people like students have asked for similar help
to do what I want to do, they just get BS.
No real help, just confusing statements and disinformation.
That is job security in the GCC world, obscurity.
> > The software development tools that we have under the GPL are in the
stone
> > age, and
> > there is no politically correct way to fix it, or legal way to protect
the
> > new development under the GPL.
> >
> > In the end we spend more time playing lawer than hacking.
> >
> > It is time that we come to terms with the situation,
> > the GPL is not up to date, it needs critical review,
> > not religious worship.
>
> I worship nothing on papar, (or electronic media for that matter). But I
> respect what it says, and I don't expect it to be rewritten to satisfy my
> needs.
My needs are not alone. It is not about me at all.
This is a problem that has to be fixed. pretty simple.
>
> > I am pretty pissed off with the hypocrisy of the GNU crowd,
> > it is alright to violate the GPL if it fits them, but not if they dont
> > like your project.
>
> I think that should be taken up with the people you're pissed off at.
It is, it has been. It will be resolved I am sure. This is my attempt at it.
>
> > I asked for permission to use the GCC to make things better,
> > questions that many do not ask, and get negative answers.
> >
> > Others do not ask, they just do so, and noone says anything.
>
> Well then less talk more work True freedom == Anarchy, the GPL is just
> there, because we have governments we'll never be rid of.
Well we have anarchy on this meta-data exchange with the gcc right now.
Everyone is doing thier own thing, noone is making any clear statements or
giving direction.
I hope that some order will come out out this effort. You are helping by
your comments and Ideas.
Mike
Robert Uhl <ruhl@4dv.net>
>
> You know, it's been funny, watching all of the GPV apologists who've
> been claiming for years that the only way to preserve freedom is to
> take it away, argue against someone here who's taking that argument
> one step further... and I fully expect to see more of the same.
The only way to preserve freedom (e.g. to live) _is_ to limit it
(e.g. to kill). The GPL is much the same. This, OTOH, goes somewhat
too far IMHO--like Prohibition.
--
Robert Uhl <ru...@4dv.net>
Ever wonder if taxation without representation might have been cheaper?
Jay Maynard
wrote:
>jmay...@thebrain.conmicro.cx (Jay Maynard) writes:
>> You know, it's been funny, watching all of the GPV apologists who've
>> been claiming for years that the only way to preserve freedom is to
>> take it away, argue against someone here who's taking that argument
>> one step further... and I fully expect to see more of the same.
>The only way to preserve freedom (e.g. to live) _is_ to limit it
>(e.g. to kill). The GPL is much the same. This, OTOH, goes somewhat
>too far IMHO--like Prohibition.
This discussion shows that this is exactly the same as the guy who asked a
woman if she'd go to bed with him for a million bucks, and got a positive
response, but offended her greatly when he offered her merely a hundred: at
that point, they're merely haggling over price.
Robert Uhl <ruhl@4dv.net>
>
> >The only way to preserve freedom (e.g. to live) _is_ to limit it
> >(e.g. to kill). The GPL is much the same. This, OTOH, goes somewhat
> >too far IMHO--like Prohibition.
>
> This discussion shows that this is exactly the same as the guy who asked a
> woman if she'd go to bed with him for a million bucks, and got a positive
> response, but offended her greatly when he offered her merely a hundred: at
> that point, they're merely haggling over price.
That's what any society is founded upon. I'm quite happy that murder
is generally illegal; it keeps the odds of me being murdered down a
fair bit. I'm not so happy that I cannot build a still.
I'm quite happy with the GPL: it keeps my code free. I'm not so happy
with proprietary licensing--which enslaves the user--or public
domain--which allows my code to be enslaved.
--
Robert Uhl <ru...@4dv.net>
Politicians and nappies should both be changed at regular intervals, and
for exactly the same reason.
Jay Maynard
wrote:
>I'm quite happy with the GPL: it keeps my code free. I'm not so happy
>with proprietary licensing--which enslaves the user--or public
>domain--which allows my code to be enslaved.
No, public domain does not allow your code to be enslaved. Your code is and
will always remain your code, to do with as you will; if you release it as
PD, that version is and will always remain PD, no matter what others do with
it.
It is this misconception (and the related one about the BSD/X/MIT licenses)
that, IMAO, gives the GPV more legitimacy than it would otherwise enjoy, and
I believe that it's in some part deliberately fostered by GPV advocates.
Phillip Lord
Jay> On Mon, 17 Jun 2002 15:49:31 GMT, Robert Uhl <ru...@4dv.net>
Jay> <ru...@4dv.net>
Jay> wrote:
>> I'm quite happy with the GPL: it keeps my code free. I'm not so
>> happy with proprietary licensing--which enslaves the user--or
>> public domain--which allows my code to be enslaved.
Jay> No, public domain does not allow your code to be enslaved. Your
Jay> code is and will always remain your code, to do with as you
Jay> will; if you release it as PD, that version is and will always
Jay> remain PD, no matter what others do with it.
So you have said before, many times on this news group.
Its only true if you accept that code is set in time. Most people see
a code base as a growing thing. Hence if I realise an application
under a free license, and then some one releases a new version under a
non free license, my original code is being used none free software.
You may feel that this is acceptable. You may feel that this does not
mean that the code is "enslaved". But others may feel
differently. More over, I know that you know others feel
differently. And I know that you know exactly why they feel
differently. So why do you pretend that you do not. It is not that
others have a misconception, its that they disagree with yours.
Phil
Stefaan A Eeckels
jmay...@thebrain.conmicro.cx (Jay Maynard) wrote:
> This discussion shows that this is exactly the same as the guy who
> asked a woman if she'd go to bed with him for a million bucks, and got
> a positive response, but offended her greatly when he offered her
> merely a hundred: at that point, they're merely haggling over price.
"The guy" is George Bernard Shaw.
--
Stefaan
--
Even in business, the greatest reputations are sometimes the result of
the glamor of city superstition. I could point to railway chairmen reputed
indispensable, whom the shareholders and the travelling public might with
great profit and comfort to themselves send to St. Helena with a pension
of £0,000 a year. -- George Bernard Shaw
Robert Uhl <ruhl@4dv.net>
>
> No, public domain does not allow your code to be enslaved. Your
> code is and will always remain your code, to do with as you will; if
> you release it as PD, that version is and will always remain PD, no
> matter what others do with it.
My code and work can be used by someone else without compensation. I
suppose it is me that is enslaved thus, rather than my code.
I suppose one could argue that I prefer compensation in kind to
compensation in cash. There's also the fact that I believe that free
software is better for all concerned: users, because it generates
better software; developers, because they can solve problems more
easily.
--
Robert Uhl <ru...@4dv.net>
Drawing on my extensive covert operations training I curled up into a
foetal postion and whimpered.
Frank Tobin
> The only way to preserve freedom (e.g. to live) _is_ to limit it (e.g.
> to kill). The GPL is much the same. This, OTOH, goes somewhat too far
> IMHO--like Prohibition.
A key thing to understand is that the proposed RGPL does *nothing* more
than what the GCC people are *already* doing; each is limiting how other
products can interact with their own. As pointed out earlier, it
appears the GCC devs are intentionaly making it difficult to interoperate
with their product because that would be a 'way around' the GPL. I think
this is very obvious given the messages linked to in
news: <8118d547.02061...@posting.google.com>
and
news: <8118d547.02061...@posting.google.com>
The proposed RGPL is really more of a modern-day version of the GPL.
Back when the GPL was written, shared libraries were the de-facto way to
combine programs on Unix. Nowadays, we have things like XML-RPC, Corba,
RMI, etc. These are also 'linking'; it's just that the GPL doesn't really
cover them.
IMHO, the RGPL *opens* doors, not closes them. It allows people to attach
the RGPL to products that they are currently limiting technically, for
fear that someone will 'work around' the GPL. Applying the RGPL will
lessen their fears and encourage them to make the innards of their
programs more accessible via a variety of means.
--
Frank Tobin http://www.neverending.org/~ftobin/
Robert Uhl <ruhl@4dv.net>
>
> IMHO, the RGPL *opens* doors, not closes them. It allows people to
> attach the RGPL to products that they are currently limiting
> technically, for fear that someone will 'work around' the GPL.
> Applying the RGPL will lessen their fears and encourage them to make
> the innards of their programs more accessible via a variety of
> means.
IMHO they need to get over themselves. I enver bought the linking
argument, and I don't think it will ever hold up. The GPL rests on
copyright law, as is right and proper--I don't believe that copyright
should or does cover linking.
In other words, I consider GPLed software to be equivalent to LGPLed
software.
Write software--empower your users. Don't impose technical means to
political ends.
--
Robert Uhl <ru...@4dv.net>
We don't just borrow words; on occasion, English has pursued other
languages down alleyways to beat them unconscious and rifle their
pockets for new vocabulary. --James D. Nicoll
Jay Maynard
wrote:
>In other words, I consider GPLed software to be equivalent to LGPLed
>software.
If this were the case, I'd have no quarrel with the FSF, for it is precisely
this difference that I feel makes the GPV unconscionable: it denotes the
point where the line is crossed between telling me what I may do with their
software and telling me what I may do with my software.
Somehow, though, I suspect the FSF would fall over in a collective
convulsive fit were a court to rule that way.
Frank Tobin
> IMHO they need to get over themselves. I enver bought the linking
> argument, and I don't think it will ever hold up. The GPL rests on
> copyright law, as is right and proper--I don't believe that copyright
> should or does cover linking.
Well, since it can cover end-user 'use', then I think it most surely can
cover 'automated-use', which linking would be a tight form of.
Anthony Ventimiglia
stated:
>
> That's what any society is founded upon. I'm quite happy that murder
> is generally illegal; it keeps the odds of me being murdered down a
> fair bit. I'm not so happy that I cannot build a still.
You can build a still, just can't sell your liquor.
Stefaan A Eeckels
> Frank Tobin <fto...@neverending.org> writes:
> >
> > IMHO, the RGPL *opens* doors, not closes them. It allows people to
> > attach the RGPL to products that they are currently limiting
> > technically, for fear that someone will 'work around' the GPL.
> > Applying the RGPL will lessen their fears and encourage them to make
> > the innards of their programs more accessible via a variety of
> > means.
>
> IMHO they need to get over themselves. I enver bought the linking
> argument, and I don't think it will ever hold up. The GPL rests on
> copyright law, as is right and proper--I don't believe that copyright
> should or does cover linking.
>
> In other words, I consider GPLed software to be equivalent to LGPLed
> software.
With that difference that the LGPL allows static linking,
and the GPL does not (because it's pretty damn obvious
that a statically linked binary contains/is a derivative
work of the library or program, and a dynamically linked
binary is not, unless one wants to open Pandora's box
(rather a nasty can of worms, IMHO).)
> Write software--empower your users. Don't impose technical means to
> political ends.
And accept that the true strength of free software is that
sharing knowledge is more productive than hiding it.
--
Stefaan
--
Even in business, the greatest reputations are sometimes the result of
the glamour of city superstition. I could point to railway chairmen
Stefaan A Eeckels
"mdupont" <mdupo...@yahoo.com> wrote:
> "Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> news:golP8.25160$qL5.1...@news4.srv.hcvlny.cv.net...
> > I guess M Dupont wrote the following (attribution snipped
> > by previous posters):
> > > It is the classic Owellian Animal Farm Revolution,
> > > the "have nots" come to power and then become just as
> > > restrictive if not more so than before for futher development.
> > Well, in this case I'm not too sure who the pigs are? You obviously
> > thing it's the GNU, while I think your proposal represents the
> > restriction.
>
> My restriction is a codification of the restriction that is imposed
> DEFACTO by the GCC and FSF.
>
> It is just a abstraction that represents the TRUE nature of the GCC
> license, not what is on paper, but what is reality,
> it is not free to modify and not free to extend in any direction.
What bitterness! I have the source code to gcc-3.02 here on
my machine, and I can extend and modify it as I please. What's
stopping you?
> By the way that the APIs have been not documented, exposed and opened to
> scripting languages, C++ and repositories, it is very very difficult to
> use the AST data freely.
Ah, but that's not part of the deal - gcc isn't _easy_ to
modify or extend, necessarily, and it's not optimally documented
(why don't you document it?)
> More difficult than it needs to be, that is the true sad story.
Alas, that's the state for many a software project. I'm told
Microsoft Word is in much the same state.
> I think that much money is spent on gcc via big companies,
> the small guy like me is easy to squeeze out with stupid accusations and
> BS arguements.
It's much easier to get help when you pay for it.
> You wont believe how many people like students have asked for similar
> help to do what I want to do, they just get BS.
The code is freely available, and legally modifiable and
redistributable. Help is _not_ freely available, and if
that makes it difficult for you to do what you'd like
to do, that's tough cookie.
> No real help, just confusing statements and disinformation.
>
> That is job security in the GCC world, obscurity.
You get the source code. You don't get anything more. That
has always been the deal.
> > > It is time that we come to terms with the situation,
> > > the GPL is not up to date, it needs critical review,
> > > not religious worship.
> >
> > I worship nothing on papar, (or electronic media for that matter). But
> > I respect what it says, and I don't expect it to be rewritten to
> > satisfy my needs.
> My needs are not alone. It is not about me at all.
> This is a problem that has to be fixed. pretty simple.
It's not about your needs. It's about a bunch of people
who make source code available. They don't cater to the
needs of others. If you feel that there is a need for a
better compiler under a different license, please write
one. And document it.
> > > I am pretty pissed off with the hypocrisy of the GNU crowd,
> > > it is alright to violate the GPL if it fits them, but not if they
> > > dont like your project.
> >
> > I think that should be taken up with the people you're pissed off at.
> It is, it has been. It will be resolved I am sure. This is my attempt at
> it.
Writing code or documentation would be a lot more productive.
In this business, the more you give, the more you get.
> > > I asked for permission to use the GCC to make things better,
> > > questions that many do not ask, and get negative answers.
> > >
> > > Others do not ask, they just do so, and noone says anything.
> >
> > Well then less talk more work True freedom == Anarchy, the GPL is just
> > there, because we have governments we'll never be rid of.
> Well we have anarchy on this meta-data exchange with the gcc right now.
> Everyone is doing thier own thing, noone is making any clear statements
> or giving direction.
> I hope that some order will come out out this effort. You are helping by
> your comments and Ideas.
Lack of direction has been one of gcc's perennial problems.
egcs did jolt all and sundry and motivated work on gcc 3. I
suggest that changing the license isn't the way to achieve
something similar.
What might work is bold leadership - and if you want to build
on the achievements of gcc it will have to be under the GPL,
because it's the commitment to the GPL that has motivated
contributors to sign their copyright over to the FSF.
Beggars can't be choosers, you know. If you neither have the
knowledge nor the dough, you have to accept what people deign
to give you.
Russ Allbery
> A key thing to understand is that the proposed RGPL does *nothing* more
> than what the GCC people are *already* doing; each is limiting how other
> products can interact with their own. As pointed out earlier, it
> appears the GCC devs are intentionaly making it difficult to
> interoperate with their product because that would be a 'way around' the
> GPL.
So fork GCC and do it yourself.
You've now strayed outside of the scope of licensing. It's not the GCC
licensing that's doing that; it's that a set of maintainers have chosen to
not to take a certain type of code. They're free people; they get to do
that. It's a personal choice. If you don't like how a maintainer is
maintaining a package, the GPL offers you the freedom to fork it and
maintain it the way you want to maintain it.
--
Russ Allbery (r...@stanford.edu) <http://www.eyrie.org/~eagle/>
Rob Ristroph
Phillip>
Jay> You know, it's been funny, watching all of the GPV apologists
Jay> who've been claiming for years that the only way to preserve
Jay> freedom is to take it away
Phillip>
Phillip> I think the word "apologist" is wrong here. Most of those who
Phillip> you are describing are perfectly happy with the freedoms that
Phillip> GPL gives, and also perfectly happy with those that it does
Phillip> not. No real apology needed.
Phillip>
Phillip> Phil
"Apologist" as used here doesn't mean someone who is admitting fault
and acknowledging it and asking forgiveness. It means someone who is
explaining the position. The greek word "apologia" is a formal
written defence of something that you belive in very strongly.
A good example is G.H. Hardy's book "A Mathematician's Apology" -- he
wasn't asking forgiveness for the mistake of being a mathematician, he
was explaining why he was a mathematician.
--Rob
Jeffrey Siegal
> So fork GCC and do it yourself.
Bad idea.
> You've now strayed outside of the scope of licensing. It's not the GCC
> licensing that's doing that; it's that a set of maintainers have chosen to
> not to take a certain type of code. They're free people; they get to do
> that. It's a personal choice. If you don't like how a maintainer is
> maintaining a package, the GPL offers you the freedom to fork it and
> maintain it the way you want to maintain it.
Legally it does, but the custom in the community is not to fork lightly,
and that's a good thing. It is preferable to try to educate and
persuade. If the current project maintainers are unwilling to accept
the argument, and if the problem becomes (or imminently threatens to
become) severe and real, as opposed to hypothetical, then perhaps
forking might be reasonable as a last resort. See ESR's excellent
writings on the subject for more discussion.
Stefaan A Eeckels
Frank Tobin <fto...@neverending.org> wrote:
> On Mon, 17 Jun 2002, Robert Uhl <ru...@4dv.net> wrote:
>
> > The only way to preserve freedom (e.g. to live) _is_ to limit it (e.g.
> > to kill). The GPL is much the same. This, OTOH, goes somewhat too
> > far IMHO--like Prohibition.
>
> A key thing to understand is that the proposed RGPL does *nothing* more
> than what the GCC people are *already* doing; each is limiting how other
>
> products can interact with their own. As pointed out earlier, it
> appears the GCC devs are intentionaly making it difficult to
> interoperate with their product because that would be a 'way around' the
> GPL. I think this is very obvious given the messages linked to in
> news: <8118d547.02061...@posting.google.com>
> and
> news: <8118d547.02061...@posting.google.com>
>
> The proposed RGPL is really more of a modern-day version of the GPL.
> Back when the GPL was written, shared libraries were the de-facto way to
> combine programs on Unix. Nowadays, we have things like XML-RPC, Corba,
> RMI, etc. These are also 'linking'; it's just that the GPL doesn't
> really cover them.
When the GPL was written, static linking was the norm. Shared
objects came a lot later. This is the reason why the FSF holds
that dynamic linking is the same as static linking. The point
is that (in copyright terms) this is the same as saying that
a reference to a text is the same as quoting the entire text,
which it emphatically is _not_.
Philosophically, RPC is no different from dynamic linking, or
from a HTTP client sending a request to a server. In all cases,
the functionality experienced by the user of the program is
provided by separate bits, executing on the same or a different
processor, in the same or a different address space. All this
sits _very_ badly with the concepts (even when they have been
expanded to cater specifically with problems introduced by
less sophisticated hardware and software) of copyright.
The simple fact is that copyright specifies quite clearly what
a derivative work is, and that "requesting the services, and
using the output of another program", cannot ever make the
requesting/using program a derivative work in copyright terms.
Depending on the language, a dynamically linked executable
can be argued to be a derivative of a library it used based
on whether it includes code from the library (e.g. in C++
header files often contain code that is copied into the
executable), but there is certainly no reason to assume that
the functional equivalence of dynamic and static linking
would make the individual objects the same under copyright
law.
> IMHO, the RGPL *opens* doors, not closes them. It allows people to
> attach the RGPL to products that they are currently limiting
> technically, for fear that someone will 'work around' the GPL. Applying
> the RGPL will lessen their fears and encourage them to make the innards
> of their programs more accessible via a variety of means.
It's a fallacy to believe that the innards of gcc are not
accessible. After all, the source code is available and
it suffices to study it to know exactly what intermediate
formats are used.
The argument is that gcc is not composed of a number of
individually usable modules that generate output that can
be used by other tools. This is largely a modern concern
(older systems were so "slow" that all this overhead was
never considered), and it does indeed make the use of these
components by non-free-software a lot easier (if not
trivial).
I would argue that this is not a problem, as free software
has always been used to facilitate the production of
non-free programs (cfr flex and bison). There's little
difference between using gcc to produce a non-free binary,
and a gcc-component to produce a parse tree. It can be
that the output is a derivative work of the tool (again,
cfr bison), but it need not be so (even the use of XML
with tag names that are generated by the compiler component
does not mean that the intermediate is a derivative work
of the compiler component in copyright terms). In all
likelihood, the presence of enough material from the
compiler component to make the output a derivative work
would also make it less than ideal as a generic format
for the use by other tools.
The real strength of the GPL is that it limits itself
to restricting distribution of derivative works and
redistribution of the original work. Any license that
tries to impose usage restrictions (which is the inevitable
result of wanting to exert control over the output of
a compiler component) would to all intents and purposes
be indistinguishable from the EULAs of this world. The
only "advantage" would be that they are gratis.
Quite frankly, such programs would have not much more
success than the "for educational/academic purposes
only" software (which is not very much, really).
Take care,
Stefaan A Eeckels
Frank Tobin <fto...@neverending.org> wrote:
> Robert Uhl <ru...@4dv.net>, on 2002-06-17, wrote:
>
> > IMHO they need to get over themselves. I enver bought the linking
> > argument, and I don't think it will ever hold up. The GPL rests on
> > copyright law, as is right and proper--I don't believe that copyright
> > should or does cover linking.
>
> Well, since it can cover end-user 'use', then I think it most surely can
> cover 'automated-use', which linking would be a tight form of.
Correction - copyright law does not cover end-user use
at all. It specifically allows the owners of a legal copy
of the software to use it has they see fit. It restricts
copying, (re-)distribution, and the preparation of derivative
works.
Sam Holden
> Russ Allbery wrote:
>> So fork GCC and do it yourself.
>
> Bad idea.
Probably, buy GCC has some precedent to indicate that forking can be
successful in moving the project forward. :)
However, in many cases you can do the equivalent of a fork without actually
forking the project. By creating patches that turn the 'official' project code
into what you think it should be. This of course entails more work on your part,
and is next to impossible if the 'official' project actively works against
you (by changing things to break your patches for no real reason other than to
break your patches).
The linux kernel is a great example of how this approach can work. It of course
reduces the user base of your code since most people use precompiled binaries
from some distributor as opposed to applying patches and compiling code
themselves. However, in the linux kernel case many distributors apply some
'non-official' patches when creating their binaries.
After all most free software people are more easily persuaded by showing them
the code, then by explaining why your idea is so wonderful.
--
Sam Holden
John Hasler
> You can build a still, just can't sell your liquor.
Last time I checked (quite a few years ago) it was illegal to distill
liquor without a license in the US no matter what you did with it.
--
John Hasler
jo...@dhh.gt.org
Dancing Horse Hill
Elmwood, Wisconsin
Anthony Ventimiglia
> Anthony Ventimiglia writes:
>> You can build a still, just can't sell your liquor.
>
> Last time I checked (quite a few years ago) it was illegal to distill
> liquor without a license in the US no matter what you did with it.
Illegal doesn't mean can't.
Sam Holden
Anthony Ventimiglia <shep...@pack.0f.wo1v35> wrote:
>
>> Anthony Ventimiglia writes:
>>> You can build a still, just can't sell your liquor.
>>
>> Last time I checked (quite a few years ago) it was illegal to distill
>> liquor without a license in the US no matter what you did with it.
>
> Illegal doesn't mean can't.
In that context it obviously does, since otherwise "just can't sell your
liquor" wouldn't be true. This entire section of the thread was clearly
talking about legality not physical possibility.
--
Sam Holden
John Hasler
> Illegal doesn't mean can't.
...And we aint paid no whiskey tax since 1792
--
John Hasler
jo...@dhh.gt.org (John Hasler)
Dancing Horse Hill
Elmwood, WI
Russ Allbery
> Legally it does, but the custom in the community is not to fork lightly,
> and that's a good thing. It is preferable to try to educate and
> persuade. If the current project maintainers are unwilling to accept
> the argument, and if the problem becomes (or imminently threatens to
> become) severe and real, as opposed to hypothetical, then perhaps
> forking might be reasonable as a last resort. See ESR's excellent
> writings on the subject for more discussion.
Sure. I agree. But regardless of the approach you take, maintainers
choosing not to accept code is not a licensing issue. That decision in no
way reduces your freedom to modify and redistribute the package, take it
over, maintain patches that do what you want, or anything else you choose.
And in the case of making it easier to plug proprietary front-ends into
GCC, the FSF has even openly acknowledged that, but stated that they will
request that people not write such code because they don't like the
possibilities that it opens up.
This is, on their part, a political stance. They're free people; they get
to hold political stances. Many people who work on GCC don't agree with
that political stance, but so far the amount of work required to maintain
a GCC without that political stance has not been sufficient to justify
taking that step.
People can quite reasonably be annoyed that the decision is being made on
political rather than technical reasons, and a lot of hackers are very
apolitical about most things. But surely I think we can all support
people's right to make decisions about what they are and are not willing
to support based on their own personal ethics.
Robert Uhl <ruhl@4dv.net>
>
> Well, since it can cover end-user 'use', then I think it most surely
> can cover 'automated-use', which linking would be a tight form of.
Can it cover end-user use? It was my impression that the GPL strives
not to be so much a license (despite the name) as a granting of
certain rights in return for certain other concessions.
--
Robert Uhl <ru...@4dv.net>
...a language is just an dialect with an army and a navy.
--Paul Tomblin, in a.s.r.
Robert Uhl <ruhl@4dv.net>
>
> > That's what any society is founded upon. I'm quite happy that
> > murder is generally illegal; it keeps the odds of me being
> > murdered down a fair bit. I'm not so happy that I cannot build a
> > still.
>
> You can build a still, just can't sell your liquor.
Not in the US (or rather, the license is too expensive for anyone to
consider for personal use).
Homebrewing is another matter--I'm quite the proud homebrewer.
--
Robert Uhl <ru...@4dv.net>
A PC without windows is like a chocolate cake without mustard.
Robert Uhl <ruhl@4dv.net>
>
> > In other words, I consider GPLed software to be equivalent to
> > LGPLed software.
>
> With that difference that the LGPL allows static linking, and the
> GPL does not (because it's pretty damn obvious that a statically
> linked binary contains/is a derivative work of the library or
> program, and a dynamically linked binary is not, unless one wants to
> open Pandora's box (rather a nasty can of worms, IMHO).)
I'm not convinced that static linking creates a derived work in any
meaningful sense. Certainly, in a purely physical sense it may, but
logically it does not. The library is still a seperate library--just
one which is linked in for the moment.
> And accept that the true strength of free software is that sharing
> knowledge is more productive than hiding it.
Yep. I GPL everything I write, precisely for that reason. Just today
I had a major issue with Solaris. The installed awk is
vendor-supplied, and is quite limited and lame. I had installed a
patch cluster, and was then installing Veritas. Unfortunately, the
vendor awk could not deal with the number of patches a certain patch
dealt with, and failed, thus causing Veritas to believe that I'd not
installed a necessary patch. I ended up installed gawk, renaming awk,
renaming gawk to awk (because the Veritas script hard-codes the awk
path), and ran the install just fine.
This is an object lesson in why free software is better. It is
better, because at one point in time someone had the same problem. He
had the source, and fixed it. Solaris 9 brags that df has gotten the
-h flag. GNU df, OTOH, has so many more flags that the comparison is
as between a grass hut and the Houses of Parliament.
--
Robert Uhl <ru...@4dv.net>
All I want is a warm bed and a kind word and unlimited power.
--Ashleigh Brilliant
Michael DuPont
> Frank Tobin <fto...@neverending.org> writes:
> >
> > IMHO, the RGPL *opens* doors, not closes them. It allows people to
> > attach the RGPL to products that they are currently limiting
> > technically, for fear that someone will 'work around' the GPL.
> > Applying the RGPL will lessen their fears and encourage them to make
> > the innards of their programs more accessible via a variety of
> > means.
>
> IMHO they need to get over themselves.
> I enver bought the linking
> argument, and I don't think it will ever hold up.
Well linking is just an indication of derived work. It is hard to
argue that you have a derived work when you are just reading a
standard file format, even if
you the gcc is the only reliable source of that information.
The issue with the RGPL is that it also needs to include some security
provision that we have not talked about. If anyone can access the
service over the web, then you have to limit the users of the service
by a special license of some form.
Then we get back to my question of "JAILing" the gpled software and
only providing the webservice and not the source code. That is
possible under the v2, if the AGPL turns into GPLv3 then there will be
no way to stop the "dangerous" interfaces from being propagated.
>The GPL rests on
> copyright law, as is right and proper--I don't believe that copyright
> should or does cover linking.
But does it cover derived works?
> In other words, I consider GPLed software to be equivalent to LGPLed
> software.
And I can embedd the gcc into a non-free program and call any function
I like, with no remorse?
> Write software--empower your users. Don't impose technical means to
> political ends.
It is way to late for that.
mike
mdupont
"Stefaan A Eeckels" <hoen...@ecc.lu> wrote in message
news:20020618000614.1...@ecc.lu...
> On Mon, 17 Jun 2002 16:26:25 +0200
> "mdupont" <mdupo...@yahoo.com> wrote:
>
> > "Anthony Ventimiglia" <shep...@pack.0f.wo1v35> wrote in message
> > news:golP8.25160$qL5.1...@news4.srv.hcvlny.cv.net...
>
> > > I guess M Dupont wrote the following (attribution snipped
> > > by previous posters):
>
> > > > It is the classic Owellian Animal Farm Revolution,
> > > > the "have nots" come to power and then become just as
> > > > restrictive if not more so than before for futher development.
>
> > > Well, in this case I'm not too sure who the pigs are? You obviously
> > > thing it's the GNU, while I think your proposal represents the
> > > restriction.
> >
> > My restriction is a codification of the restriction that is imposed
> > DEFACTO by the GCC and FSF.
> >
> > It is just a abstraction that represents the TRUE nature of the GCC
> > license, not what is on paper, but what is reality,
> > it is not free to modify and not free to extend in any direction.
>
> What bitterness! I have the source code to gcc-3.02 here on
> my machine, and I can extend and modify it as I please. What's
> stopping you?
I was asked by RMS not to make it easy to allow third partys to have access
to the fully annotated source trees of the gcc.
That would allow for third party code generation.
What is stopping me is a license that covers this exporting of the ASTs in a
protected way.
What about the c-backend that outputs the RTLs in C?
What about
>
> > By the way that the APIs have been not documented, exposed and opened to
> > scripting languages, C++ and repositories, it is very very difficult to
> > use the AST data freely.
>
> Ah, but that's not part of the deal - gcc isn't _easy_ to
> modify or extend, necessarily, and it's not optimally documented
> (why don't you document it?)
I am working on documenting it.
I have bootstrapped the entire compiler into XML and have build a big
database of it.
I have been doucumenting how the tree nodes are built and how they interact.
> > You wont believe how many people like students have asked for similar
> > help to do what I want to do, they just get BS.
>
> The code is freely available, and legally modifiable and
> redistributable. Help is _not_ freely available, and if
> that makes it difficult for you to do what you'd like
> to do, that's tough cookie.
We are not talking about help,
we are talking about dicouragement, and conflicting messages.
We are also talking about nasty accusations.
If you read this thread you will find a message from me with links to the
discussion that brought up this issue.
>
> > No real help, just confusing statements and disinformation.
> >
> > That is job security in the GCC world, obscurity.
>
> You get the source code. You don't get anything more. That
> has always been the deal.
So there is not any issue with be publishing the GCC as a perl lib with an
easy to use interface? Or a C++ interface to the gcc?
You bet there will be lots of hatemails.
>
> > > > It is time that we come to terms with the situation,
> > > > the GPL is not up to date, it needs critical review,
> > > > not religious worship.
> > >
> > > I worship nothing on papar, (or electronic media for that matter). But
> > > I respect what it says, and I don't expect it to be rewritten to
> > > satisfy my needs.
>
> > My needs are not alone. It is not about me at all.
> > This is a problem that has to be fixed. pretty simple.
>
> It's not about your needs. It's about a bunch of people
> who make source code available. They don't cater to the
> needs of others. If you feel that there is a need for a
> better compiler under a different license, please write
> one. And document it.
I feel that the GCC needs to have its interfaces opend up, we need to have
an open code repository that can be accessed with scripting tools. I am
working on that, and there are legal and moral issues involved.
>
> > > > I am pretty pissed off with the hypocrisy of the GNU crowd,
> > > > it is alright to violate the GPL if it fits them, but not if they
> > > > dont like your project.
> > >
> > > I think that should be taken up with the people you're pissed off at.
>
> > It is, it has been. It will be resolved I am sure. This is my attempt at
> > it.
>
> Writing code or documentation would be a lot more productive.
> In this business, the more you give, the more you get.
I have been working on the gcc interface for the past 2-3 years and have
lots of code to share.
>
> > > > I asked for permission to use the GCC to make things better,
> > > > questions that many do not ask, and get negative answers.
> > > >
> > > > Others do not ask, they just do so, and noone says anything.
> > >
> > > Well then less talk more work True freedom == Anarchy, the GPL is just
> > > there, because we have governments we'll never be rid of.
>
> > Well we have anarchy on this meta-data exchange with the gcc right now.
> > Everyone is doing thier own thing, noone is making any clear statements
> > or giving direction.
> > I hope that some order will come out out this effort. You are helping by
> > your comments and Ideas.
>
> What might work is bold leadership - and if you want to build
> on the achievements of gcc it will have to be under the GPL,
> because it's the commitment to the GPL that has motivated
> contributors to sign their copyright over to the FSF.
I have put my code under GPL.
I have been working on an XML/SQL/Perl interface into the gcc,
it is not under the GNU coding standards, it is a new development.
I dont know if the FSF is interested in getting the copyright,
last I can tell they want to delete my project not support it.
>
> Beggars can't be choosers, you know. If you neither have the
> knowledge nor the dough, you have to accept what people deign
> to give you.
Well we are much futher along than that, I am not begging at all anymore.
We are talking now about a new license that will cover the intropspector,
close the holes in the GPL and provide a legal foundation for protecting of
the GCC.
That is my contribution.
mike
mdupont
> On Mon, 17 Jun 2002 21:43:35 GMT
> ru...@4dv.net (Robert Uhl <ru...@4dv.net>) wrote:
>
> > Frank Tobin <fto...@neverending.org> writes:
> > >
> > > IMHO, the RGPL *opens* doors, not closes them. It allows people to
> > > attach the RGPL to products that they are currently limiting
> > > technically, for fear that someone will 'work around' the GPL.
> > > Applying the RGPL will lessen their fears and encourage them to make
> > > the innards of their programs more accessible via a variety of
> > > means.
> >
> > IMHO they need to get over themselves. I enver bought the linking
> > argument, and I don't think it will ever hold up. The GPL rests on
> > copyright law, as is right and proper--I don't believe that copyright
> > should or does cover linking.
> >
> > In other words, I consider GPLed software to be equivalent to LGPLed
> > software.
>
> With that difference that the LGPL allows static linking,
> and the GPL does not (because it's pretty damn obvious
> that a statically linked binary contains/is a derivative
> work of the library or program, and a dynamically linked
> binary is not, unless one wants to open Pandora's box
> (rather a nasty can of worms, IMHO).)
to it.
To limit thier freedom to abuse the gcc.
The RGPL will allow for the protection of that dynamic linking.
>
> > Write software--empower your users. Don't impose technical means to
> > political ends.
>
> And accept that the true strength of free software is that
> sharing knowledge is more productive than hiding it.
Well I am trying to share, but there is a limit to the amount of sharing
that will be good.
The sharing of the gcc as a dll is possible, but is it good?
mike
mdupont
"Jeffrey Siegal" <j...@quiotix.com> wrote in message
news:3D0E77D9...@quiotix.com...
> Russ Allbery wrote:
> > So fork GCC and do it yourself.
>
> Bad idea.
as I can.
The issue is that we need the idea behind the RGPL to move forward on this.
I am not going to try and just force this on everyone, but try and reach an
consensus before moving forward.
There are alot of people who are trying to do similar things, most to hook
to non-free software. Before my project can really become prime-time we need
to resolve this license issue.
mike
mdupont
> On Mon, 17 Jun 2002 16:41:05 -0400
> Frank Tobin <fto...@neverending.org> wrote:
>
> > On Mon, 17 Jun 2002, Robert Uhl <ru...@4dv.net> wrote:
> >
> > > The only way to preserve freedom (e.g. to live) _is_ to limit it (e.g.
> > > to kill). The GPL is much the same. This, OTOH, goes somewhat too
> > > far IMHO--like Prohibition.
> >
> > A key thing to understand is that the proposed RGPL does *nothing* more
> > than what the GCC people are *already* doing; each is limiting how other
> >
> > products can interact with their own. As pointed out earlier, it
> > appears the GCC devs are intentionaly making it difficult to
> > interoperate with their product because that would be a 'way around' the
> > GPL. I think this is very obvious given the messages linked to in
> > news: <8118d547.02061...@posting.google.com>
> > and
> > news: <8118d547.02061...@posting.google.com>
> >
> > The proposed RGPL is really more of a modern-day version of the GPL.
> > Back when the GPL was written, shared libraries were the de-facto way to
> > combine programs on Unix. Nowadays, we have things like XML-RPC, Corba,
> > RMI, etc. These are also 'linking'; it's just that the GPL doesn't
> > really cover them.
>
> When the GPL was written, static linking was the norm. Shared
> objects came a lot later. This is the reason why the FSF holds
> that dynamic linking is the same as static linking.
>The point
> is that (in copyright terms) this is the same as saying that
> a reference to a text is the same as quoting the entire text,
> which it emphatically is _not_.
So a parse tree that just contains reference to the original code is
copyrightable independant of the original author of the code?
> Philosophically, RPC is no different from dynamic linking, or
> from a HTTP client sending a request to a server. In all cases,
> the functionality experienced by the user of the program is
Function invocation.
> provided by separate bits, executing on the same or a different
> processor, in the same or a different address space. All this
> sits _very_ badly with the concepts (even when they have been
> expanded to cater specifically with problems introduced by
> less sophisticated hardware and software) of copyright.
The problem is that you cannot stop anyone from including non-free software
and running away with it.
the RGPL tries to create a virtual private network of data exchange between
GPled software exclusivly.
> The simple fact is that copyright specifies quite clearly what
> a derivative work is, and that "requesting the services, and
> using the output of another program", cannot ever make the
> requesting/using program a derivative work in copyright terms.
So you need to use an EULA.
> Depending on the language, a dynamically linked executable
> can be argued to be a derivative of a library it used based
> on whether it includes code from the library (e.g. in C++
> header files often contain code that is copied into the
> executable), but there is certainly no reason to assume that
> the functional equivalence of dynamic and static linking
> would make the individual objects the same under copyright
> law.
Unless the license explicitly states that, and the user aggrees to it with a
digital signature?
>
> > IMHO, the RGPL *opens* doors, not closes them. It allows people to
> > attach the RGPL to products that they are currently limiting
> > technically, for fear that someone will 'work around' the GPL. Applying
> > the RGPL will lessen their fears and encourage them to make the innards
> > of their programs more accessible via a variety of means.
>
> It's a fallacy to believe that the innards of gcc are not
> accessible.
The political and licence problems are the issue.
>After all, the source code is available and
> it suffices to study it to know exactly what intermediate
> formats are used.
That is what I am doing, I would like to publish this in form of an XML
format, SQL database, Perl modules and other dynamic interfaces.
> The argument is that gcc is not composed of a number of
> individually usable modules that generate output that can
> be used by other tools.
But it is. These modules like the mangler/demangler code are jealously
guarded.
>This is largely a modern concern
> (older systems were so "slow" that all this overhead was
> never considered),
Good point.
> and it does indeed make the use of these
> components by non-free-software a lot easier (if not
> trivial).
Fine.
>There's little
> difference between using gcc to produce a non-free binary,
> and a gcc-component to produce a parse tree.
I disagree.
The source code can be recreated from the parse tree.
The semantically annotated parse tree can be used for code generation and
non-free compiler backends.
The parse tree could be used as an input into the compiler for making it a
Super Assembler.
>It can be
> that the output is a derivative work of the tool (again,
> cfr bison), but it need not be so (even the use of XML
> with tag names that are generated by the compiler component
> does not mean that the intermediate is a derivative work
> of the compiler component in copyright terms).
If you think so, then look at the swagkit from the university of waterloo.
There is no way to make that software free, even if it needs the gcc to run.
>In all
> likelihood, the presence of enough material from the
> compiler component to make the output a derivative work
> would also make it less than ideal as a generic format
> for the use by other tools.
There are other supposed generic formats, and you can alway translate the
XML into them.
> The real strength of the GPL is that it limits itself
> to restricting distribution of derivative works and
> redistribution of the original work.
So the XML dump of a GPLed program is still GPLed, but can be processed by
non-gpl software?
>Any license that
> tries to impose usage restrictions (which is the inevitable
> result of wanting to exert control over the output of
> a compiler component) would to all intents and purposes
> be indistinguishable from the EULAs of this world. The
> only "advantage" would be that they are gratis.
What is wrong with that?
For the modified GCC, you would then be able to keep it from being abused
and embedded.
For web services you need to have some form of protection from that abuse.
The EULA makes sense in the long run, even for free software.
> Quite frankly, such programs would have not much more
> success than the "for educational/academic purposes
> only" software (which is not very much, really).
I disagree.
This is the *only* way to be able to build a full program database and
visualitation GUI for the gcc, without
opening up for abuse by third parties.
The end users will want to be able to script and dynamically link, you
cannot take that away, otherwise your development cycles get too long.
Visual tools need to have a way of testing safely, without long compile and
link cycles.
mike