Re: Copyleft and embedded Forth
- From: Aleksej Saushev <asau@xxxxxxxx>
- Date: Sat, 09 Feb 2008 01:34:16 +0300
byron@upstairs.(none) (Byron Jeff) writes:
In article <87zluctxgm.fsf@xxxxxxxx>, Aleksej Saushev <asau@xxxxxxxx> wrote:
byron@upstairs.(none) (Byron Jeff) writes:
In article <8763x2vtc7.fsf@xxxxxxxx>, Aleksej Saushev <asau@xxxxxxxx> wrote:
byron@upstairs.(none) (Byron Jeff) writes:
In article <8763x3xnri.fsf@xxxxxxxx>, Aleksej Saushev <asau@xxxxxxxx> wrote:
"Jean-François Michaud" <cometaj@xxxxxxxxxxx> writes:
[Getting too long. Snipping...]
That's about it. All the complexity comes in when trying to mix those
rules with closed source software.
The complexity comes when trying to mix those rules with any software,
except GPLd. This means the software isn't free,
Back to the same question I asked earlier: Free for whom?
For anyone except copyright holder and dumb user. Isn't it clear?
No it isn't. Any right asserted to one party will interfere with the
rights of another.
No, copyright holder still owns copyright and may do whatever he wants,
though he's slightly more restricted by the law after he publishes
the work.
Every right you give to a developer takes away a right from the user.
That's apparently wrong. Developer may do with p.d. work anything
he likes, and user may do with p.d. work whatever he wishes,
except claiming copyright or authorship.
But there's no interaction between the two. And that's the problem.
Start with a P.D. base A. To you it's free because both the developer
and the user can get that base. But what happens when the developer
modifies that base to produce A+B? Since P.D. software has no asserted
copyright, then the user's access to A+B (which cannot exist with A) is
limited to whatever the developer decides to encumber that combination.
No. If B is negligible, it doesn't constitute art work, if it is
significant rework, and A is negligible, then it is quite another work,
and the license of A doesn't matter, if A and B are interspersed,
then it is still doubtful, that "A+B" can be viewed as derived work.
Stallman's view is biased, and shouldn't be taken into account seriously.
In the GPL sense of free, both the base and modifications have the same
level of access. That's what makes the software free.
Don't use this wording, if you want to be heard.
Software is an object and can't be free or non-free therefore.
If you stick to unnatural meanings, you are to go to psychiatrist,
recognising objects as if they're living creatures isn't sane.
It seems to me, I've understood, what is confusing here.
If you say "free software", from now on I read it as "gratis",
because software isn't living creature. Use proper words, please.
So, availability of source and free license to use, modify and redistribute
in modified and unmodified forms is fair to users and creators, who would
like to take contributions back. Because: a) both sides have profit,
developers gain popularity and higher social status, users gain free
software and protection from abandonement; b) both sides take equal risks,
the independent fork become more available, users take risk of loosing
support, if they don't contribute in code or material goods (donations),
in addition the most active users may become developers.
But to do that you
have the limit the rights of the modifiers, so that users/developers
that receive the modifications downstream (A+B) have the same access as
the developer who received base A.
If you restrict developers, then they take higher risks, hence worse support
and decline unless monopoly (like in GCC).
The GPL is designed to protect end user rights at the expense of
developers and redistributors.
It isn't. End user is forced to pay lawyer consultant to understand
how he can modify work.
The end user can modify anything they want. The end user needs to
distribute those modifications when they distrubute the modified
product. Exactly how hard is that?
I have explained in previous messages, did you read them?
Note, that end user tends to use _any_
software tool (even acquired illegally) without licence segregation.
Use isn't an issue with any free software license. It's all about
distribution.
Everything is about distribution, since anything acquired lawfully
is allowed to be modified.
So, end user is forced to either consult lawyer before he shares
modified work, or take a risk of being sued by FSF for violating
some unclear or even illegal point in their licence.
So we agree that it's distribution. Share the modifications.
Change the licence, and I'll share.
End of story.
In each and every instance that the FSF gets involved, the source for
modifications have not been distributed. No lawyer is required if you
are not violating the license.
Lawyer is required always, if the licence isn't clear enough.
GPL is that very case. Programmers are not lawyers most frequently,
and any "veto" distracts (can distract) from doing useful work.
As a consequence, GPLd software is of worse quality usually.
to tell more,
the issues are much worse than with classic BSD: even when you
have more free license, you have to deliver not the simple list
of contributors but the whole unmodified source of components to
the user, who doesn't care about what have you used, and doesn't
want to see the guts.
It doesn't matter what the end user wants. It's the developers
responsibility to deliver according to the license.
The problem is that it's a slipperly slope. Presuming that the end user
doesn't want source means thinking that you don't need to deliver it.
If you can't read, maybe you should quit the discussion.
I can read just fine thank you. You are saying that why bother
delivering source when the user doesn't want it. I am telling why you
need to deliver the source even when the user doesn't want it.
You're wrong, because user's will is supreme.
I have told that GPL imposes problematic limitations.
Then don't use it. You have to choose to use GPL software. If it's
restrictions causes you problems, then simply don't use it. Either find
software that is under a license that doesn't cause you problems, or
negotiate a license with the copyright holders that releases you from
the limitations that the GPL imposes.
That's why I don't contribute to GPLd projects willingly, even when I can.
I don't like commercial closed source licenses. You don't hear me saying
that those licenses should be banned because I don't have access to the
source and that I can't redistribute their product as I see fit. I
either don't user their software or try to negotiate some other
arrangement.
I never talked about commercial software, so why this passage?
You represent GPL as if it is the only free open source and the only
free non-commercial licence.
RMS threatened everyone around, that BSD (original) is problematic
for its acknowledgement requirement, which is less than 5-line note
"This software includes parts..." He replaced it with requirement
to distribute the whole source instead, the source, which carries
the repetitive "This is free software..." text. The provision of
the source via public accessible media isn't sufficient to him.
Why not? To quote from the GPL FAQ:
http://www.gnu.org/licenses/gpl-faq.html#AnonFTPAndSendSources
FAQ has no legal power, it's not the licence. So, I still have to find
a lawyer to check, whether the unwanted legal action is possible or not.
Then modifications start to get distributed without source.
It's easy to get the source, build it and check, whether
software carries unknown modifications. The only additional
step is getting the source.
This is a long thread. What are we arguing about? As the FAQ outlined
above, as long as you make the source available to the user, then your
obligation is finished.
What exactly is RMS and the FSF asking someone to do that's such a
problem?
1. Transfer copyright to FSF.
2. Unwillingness to recognise contribution, I still want to be sure,
that noone will clear the list of contributors. If I work for free,
I want to get compensation in recognition, classic BSDL provides that,
so do modern BSD, MIT, and X11 licences. GPL does not and FSF supports
that.
It's not 2-3 line "This contains parts of
software written by John Doe and contributors," it is the same
full page "THIS IS FREE SOFTWARE" advertising in each file.
That's not how the GPL works.
Oh, really?
Have you _ever_ read the source of any GNU tool?
Yes. I've written, modified, and released GPL software. So yes I have
read the source.
So, you think, that it's fine, that every file starts from full page
of useless for common user legal stuff, don't you?
"If the program is interactive, make it output a short notice like
this when it starts in an interactive mode:
"Gnomovision version 69, Copyright (C) YYYY NAME OF AUTHOR
Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type
`show w'. This is free software, and you are welcome to
redistribute it under certain conditions; type `show c' for details."
Again it's intent that's the issue. The notice is designed to ensure
that the end user is aware of their rights.
The end user becomes aware of his rights, when he downloads or
otherwise get the software. If it is unclear, "all rights reserved"
by default, otherwise, the terms are written in accompanying documentation.
If it is 1-page BSDL, user can read it, it takes short time, if
it is GPL, user can't understand it, since it involves not only
reading 10 pages licence itself, but 150 K FAQ, and you still
have to consult lawyer, since it isn't clear, whether there're
really no underwater stones, since it took so long to write.
And as usual if you have a problem with it, then don't use the license
or software that's under it. You always have that right.
That's why I tell that GPL isn't that useful to end user.
And each file should carry even longer advertising, that it's
"free" software.
How can it be free if the end user isn't aware that it's free?
End user can read accompanying documentation, there's no need to waste
his time to transmit more useless text.
That's
why every time you bring this up, I ask "Free to whom?".
Free to modify by advanced user or programmer.
The GPL
purposefully and intensely asserts every right of the end user while in
fact suppressing the rights of intermediaries to that end user.
It complexifies realisation of those rights to a very great extent,
since FSF deceives user of his rights to prepare integrated works.
The end
user who gets software must have access to the source of that exact
software (including any intermediate modifications), and be aware of
their rights. Each and every portion of the GPL supports that. To do
that impositions are made on developers so that end users get their
rights and know about it.
And that impositions are illegal in some parts.
Questions like:
"I received the source for a GPL program. Can I make changes and sell
the changed program without giving up the source for my changes?"
Have you read my message? I don't ask such simple questions.
I _don't_ make any changes to GPL components, I deliver those source
to end user (with questions "WTF you waste my bandwidth for this
stuff?", rarely with addition "If I wanted it, I know about gnu.org
and google.com."),i
So simply make the offer for source. That's all you need to do.
In written form? Thanks! It just proves that GPL is licence for lawyers,
not for users nor for developers either.
Sigh. I think our conversation is running out of steam. The offer
asserts the right for users to have access to the source of the GPL
software they are using. It's the developers responsibility to make it
convenient to access that source for a period of time. Hence the offer.
Why does that make it a license for lawyers?
Because all impositions are made against both, users and developers,
the only winning party is lawyers, since both, users and developers
are forced to pay lawyers for consultation about their actual rights.
Because it's unclear which points of GPL are valid and which are not,
since they contradict the general copyright law.
so what's the difference between LGPL and GPL?
For the LGPL's code itself, virtually nothing. The difference is that
LGPL allows other licensed code to use it.
GPL isn't different, if you don't take RMS' meaning for words,
rather natural one.
Maybe that's why you think it's a lawyers license. So tell me why it
isn't different?
Explained in previous messages, and even repeated below.
All the nuances come in when trying to determine if something is a
derivative. Anything that isn't standalone, requires either
recompilation or relinking to change, is considered derivative.
Most of the derivative issue is based around rule #2.
dlopen/dlsym/dlclose isn't relinking, using external linker is my own,
internal problem, GNU has nothing to do with that.
You have code in the same process space as GPL code. So that combined
entity is a derivative work of the GPL code. So it all has to be GPL.
No, it isn't. GPLd code is clearly separated from any other code,
the code remains functional, when the library is replaced by any
other one with similar interface. Non-GPLd part doesn't need to
have any deep knowledge about GPLd part and vice versa.
Memory space doesn't matter, it is the same memory on this system,
either logically (as concept) and physically (as device).
It does matter. Obviously the dynamic library offers functionality to
the whole work, otherwise it has no reason to exist. It exists in the
same process memory space as the GPL work. Presumably the DLL cannot
function as a standalone component without the GPL work.
It's a derivative. It needs to be GPLed. Or alternatively rewrite all
of the GPLed code under a different license.
No it isn't derivative, since the original code isn't modified in any way
and it is free to use and to redistribute. The definition of "derivative"
in the licence contradicts to the natural meaning of the word,
and to the legal meaning as well. Read the law.
As for other points, the dynamic library provides functionality to the whole
system, and any program, even those with no assumption on internal structure
of the library or interface of the latter, can use it any way.
E. g. for entropy source or statistical measures.
And it will be loaded into the same process memory space, if that construct,
of separate process memory space, exists at all.
So, to make your point valid, you should force anyone to license their
operating systems, linking and loading tools, any file tools that could
possibly operate on shared libraries or applications under GPL, thus
violating rights of other copyright holders. Hence your point isn't
valid, the dynamic library stays separate work. and linking to it
through standard, details-ignorant mechanism, doesn't produce derivative
work.
Same for static linking by standard code-ignorant mechanisms.
So, it is obvious, that RMS deceives licensees of their rights,
the law protect to them.
This is what I'm talking about when I say intent. You've given an
example of deliberate code segregation specifically for the purpose of
not having to deliver source for part of the code.
That's why you have to have licenses with all the legalese.
We have the law, that protects our user, _consumer_ rights.
If you don't know the law, you ought to go and learn it,
as for now you're just another liar.
I'm done. This is tiring.
So, you can't stand legal arguments. When you loose, "it is tiring."
--
CKOPO BECHA...
CKOPO CE3OH...
.
- Follow-Ups:
- Re: Copyleft and embedded Forth
- From: none
- Re: Copyleft and embedded Forth
- From: Stephen J. Bevan
- Re: Copyleft and embedded Forth
- References:
- Re: Copyleft and embedded Forth
- From: Aleksej Saushev
- Re: Copyleft and embedded Forth
- From: none
- Re: Copyleft and embedded Forth
- From: Aleksej Saushev
- Re: Copyleft and embedded Forth
- From: none
- Re: Copyleft and embedded Forth
- Prev by Date: Re: common place apps in forth available, for example irc, or multi chat, browser, etc?
- Next by Date: Re: Copyleft and embedded Forth
- Previous by thread: Re: Copyleft and embedded Forth
- Next by thread: Re: Copyleft and embedded Forth
- Index(es):
Relevant Pages
|