dynamic linking doesn't count.... Read.... http://digital-law-online.inf o/lpdi1.0/treatise22. html and Read the GPL... 'work based on the Program" means either the Program or any derivative work under copyright law'
My argument is that when I dynamically link an application I CAN only EXPECT equivalent functionality as the library, not the library itself. Which is why you cannot way it is derived from anything more than the interface, and the interface isn't copyrightable.
'The act of linking against someone else's code is the same as incorporating their code into your program'
No it's not, you are an Idoit if you link dynamically and expect to have exactly the same library on the users machine.
" What happens if Without their code, your program won't work" Wrong again... because of the point above (there can be many versions of the library written by different people under different terms), and because, you can always use a application without the library. I'm going to use an obvious case (to make it easier on you) I am a source only commercial.
I distribute my source that (must in you books) needs to be linked against a GPL library (or a library with similar interfaces).
The user doesn't have the library, so instead he picks out the functions from my source and uses them.
You can do exactly the same with a binary distribution it's just technically more difficult for 'some' people.
Under European law I have the right to reverse engineer an application for the porpoises of interoperability. If that means I didn't have the GPL library and I want to use the functions of that application then so be it. The Law says that I can use the application in part or in whole.
"Slashdot still works perfectly whether the link to Websters works or not. " No it doesn't it 'crashes' when you try and click on my link (if websteres is removed) in part/. stops working.
"In the case of linked software libraries, if the library is missing, your program won't work. That's what makes your program a derivative work of the library."
no more than/..., also see points 1 and 2. If I don't have MESA, ATI's opengl will do. etc.......
Dynamicly linking only involves headers, headers are a representation of fasts and the fasts are not bound by copyright. I am allowed by law to copy those facts, GPL or no GPL. When I distribute my dynamically linked application I only distribute my work and facts.
I suggest you read http://digital-law-online.info/lpdi1.0/treat ise22. html
oh, and the GPL..
'work based on the Program" means either the Program or any derivative work under copyright law'
see 'work under copyright law', no copyright violation no GPL.
No they don't. They model the basic version, then spend millions of dollars on super computers to do wind tunnel simulation.... that's 'windtunnel' not every nut and bolt.
and that's just cars.
When they get the data they factor in a good margin of error (10th percentile or something) and build to that.
A the simulations have got better they've been able to reduce the margin of error. If you look at old plains being build they were at least 100% bulkier than todays plains.
' Altered source versions must be plainly marked as such, and must not be misrepresented as being the original software.' says that I can cut and paste from libpng so long as I put a comment around the code.
SFAIK, cleanroom (which is what I described) has been taken to court and is ok. http://www.macnewsworld.com/story/33839.html
Thanks to that bit of googling I've found the exact defnintion and process that the courts use to determine if a work is a copyright infringement, it's called 'abstraction, filtration, comparison'
" Courts use the "abstraction, filtration, comparison" test to determine if the defendant's accused software infringes the copyright on the plaintiff's software. Basically, this test compares the original work with the accused work by removing any parts that are strictly functional (permitted to be copied) and examines the sameness of what is left."
my "but my diffs to Webster's could apply to lots of books" argument, was an artistic one. I produce the diffs (with the intent of them being against websters) and publish them in the knowledge that they can do nothing about it (since they don't even know who they are).
Yes, I don't have access to Windows or Directx 9, but look http://www.oliverthered.f2s.com/projects/wine/ I can independently create a DirectX 9 library with nothing more than the API for the library and some example programmes written for Directx 9.
This is why I assert, GPL can not cover dynamic linking. (plagiarism is something completely different)
cleanroom libraies have a long tradition of getting around copyright laws, I can cleanroom any GPL library you want me to, so GPL can not cover dynamic linking.
'However, hypothetically, you might agree with most of Webster's definitions but not agree with a few words from lots of the entries (or might think the some entries needed updating due to changes in the language). You might, therefore, want to put a pile of instructions or diffs here for which words to change (e.g.: "change word 5 of line 18 of page 789 from `intellectual property' to `exclusive intellectual exploitation rights'").'
But I'm not posting one line diffs to GPL code and trying to sell it as my own.
Also you diff included `intellectual property', what about "change word 5, line 18 page 789 'exclusive intellectual exploitation rights'" hmm I can see an art project, now how many books could "change word 5, line 18 page 789 'exclusive intellectual exploitation rights'" be applied to and still be valid, I would expect at least a few thousand different titles.
Now I can publish anonymous changes against copyrighted works (in the form of a diff) "change word 5, line 18 page 789 'exclusive intellectual exploitation rights'"
and produce a book, that contains nothing that is copyrightable..... but you have to find the books to prove it. The problem is, it could match a thousand books out of all the books that have ever been published.
I could do that for you now with any library you choose. well, unless a particular algorithm is a trade secret, which is almost never the case (and certainly not for GPL'ed code.)
So, given 1 me, and 1 GPL library, that GPL library can cover dynamic linking 0% of the time, so no need for LGPL, which was my original argument, and why I have posted so many replys to buck the trend of 'wrongly' asserting that the GPL covers dynamic linking.
GPL can never pratically cover Dynalic linking because when you 'dynamically link' an application you don't link, you just compile, which can take place with only the headers present, the user does the linking.
I may possibly consent on that minor usage of dynamic linking (I've almost never see outside of a hello world), but again the GPL cannot cover the wider usage of dynamic linking, Generally the 'value added' and function of the application calling the libraries is so great that the user cannot tell what function the underling libraries perform. This is nothing like a diff.
If I implement and entire graphics driver as a linux module (ignoring it's binary module clause) how could it ever be considered a diff when perhaps it references linux less than 0.01% of the time.
My point is that I could give the headers I've created to anyone that are Pblic domain but provide the same interface as some GPL libraries.
That person could link against the headers and a description of the function of each call without even knowning about the GLP libraries (and certainly without agreeing to the GPL) because when you dynamically link the compiler only looks at the headers. (i.e. the author of the applicaion didn't do any linking at all, that's done dynamically by the user)
This would constiture a work created without the presence or knowlage of any GPL material, so must be considered to be independant (cleanroom).
Because I can do this for any GPL library I choose, the GPL cannot cover dynamic linking, it's impossible.
"if I took a copyrighted novel and made a diff for it along the lines of "change line 35 on page 63 to read..."
I'd be interested to read that legislation, I've never seen it (it doesn't appear in the Berne convention), or the FSF documentation. I'm very sceptical as to the scope of your argument, it doesn't take long before "change line 35 on page 63 to read..." becomes a work in it's own right.
I still don't believe this affects dynamic linking in any way, it's nothing like a diff.
"the ONLY time dynamic linking forms a derived work is when your work is based on the work you are dynamic linking to as opposed to being a truly separate work."
Which, if I create my own representation of GPLed headers can be never, meaning that the GPL cannot cover dynamic linking. (without a search warrant for headers that may no longer exist). This is why I can give you example after example and be right every time.
1."No because it is not based on Linux (assuming you don't include Linux code or code based on Linux). It isn't even a similar thing. One is a kernel; the other is a manual. (Also, note that Linux isn't a good example to use here as it has an added clause allowing linking that wouldn't normally be allowed by the GNU GPL.)"
And what is the difference? none, code, a book, a speech all derived works from your point of view.
If I take linux away from 'Linux for dummies' it becomes useless it cannot be used independently.
So where does it end? (at the dynamic linking level?)
"Microsoft wouldn't agree with you on interfaces", well they lost that one in court didn't they.
"but the FSF (and most same people) do. For that reason, if the dynamic linking uses a standard interface it is not considered to be based on the original work wrt copyright law. If your software doesn't require those libraries (e.g.: because it can be used with others created independently), it is definitely not based on the libraries."
What is a 'standard' interface?
I can write my own version of the libraries with the same interfaces, how is this different? I personally can use any programme you give me 'without' the dynamic libraries being present, just because it's technically difficult for you to use it independently doesn't make it a fact.
"For more information on how the FSF believe this works in Prentice"
For the above two reasons it doesn't.
you have failed to define a 'derived' work in a way that says I cannot dynamically link a commercial product against GPL. you have wishfully thought.
The FSF is very fluffy, because you cannot define a derived work in that way, it's impossible.
"Mere aggregation of two programs means putting them side by side on the same CD-ROM or hard disk. We use this term in the case where they are separate programs, not parts of a single program. In this case, if one of the programs is covered by the GPL, it has no effect on the other program."
fine....
"What constitutes combining two parts into one program? This is a legal question, which ultimately judges will decide. "......Because we can't because it's impossible.
"We believe that a proper criterion depends both on the mechanism of communication (exec, pipes, rpc, function calls within a shared address space, etc.) and the semantics of the communication (what kinds of information are interchanged)."
'exec, pipes, rpc, function calls within a shared address space, etc' are all just semantics put in to confuse the reader into thinking that they represent something different.
"If the modules are included in the same executable file, they are definitely combined in one program."
Yes, we agree on that one (so far as distribution).
"If modules are designed to run linked together in a shared address space, that almost surely means combining them into one program.", for 'almost surely' read doesn't , you cannot design software in that way, you design it with the INTENT of a SIMILAR interface being present at runtime. (e.g. what happens when version 1.4 of a library is released, doe the interface become 'standard' ?, what about mpg321 and mpg123, or GTK and QtGTK?
"By contrast, pipes, sockets and command-line arguments"....... there we go, trying to confuse people again.
With RPC my application could be running on a server on the other side of the universe from the library, going over all manner of pipes, sockes and command-line arguments. is HTTP also covered by the GPL?
/** * Message levels **/ #define KERN_EMERG "<0>"// emergency (unusable) #define KERN_ALERT "<1>"// about to die #define KERN_CRIT "<2>"// critical #define KERN_ERR "<3>"// not quite so critical #define KERN_WARNING "<4>"// ignore this of you want #define KERN_NOTICE "<5>"// something you should know #define KERN_INFO "<6>"// something you may want to know #define KERN_DEBUG "<7>"// somethig I may want to know /** * MIN max constants.. **/
It's about a meg or so, if you fancy testing, or playing around and implements the core of DirectX9, beyond cedega in some areas. I'm merging with wine head at the moment so wine should be on a similar level to Cedega in a month or so when shaders are fixed.
Hopefully when Cedega get some competition they will actually improve there product or release more source so that others can.
I've got a few months research into UI's to put up when I get the time.
Including things like the answer to, just how long does it take to click that button, or Is this really a better user interface?
I would like to see soft 'modes' so that if I'm at work and I get a fault call I can switch over to 'fault call mode' and all my book marks and menus are configured optimally to 'fault call' mode, then when I'm done I can switch back to internet mode or debugging mode.
Re:SHA-1 is alien to me.
on
SHA-1 Broken
·
· Score: 1
Jesus used alien tech to bend spoons with his dead grandmother whilst the random number generator looked into the future or asteroids hitting the planet and killing everyone.
Yeh, right crack open another bottle of anti-psycotics and stop trying to rule the world Mr Bush.
could you honestly trust someone who 'believed' that Jesus was here to save us, God is real, and if your a good boy you go to Heaven when you die, when all the facts point to it being all in his head?
Slashdot Mirror
Why, is this a discussion about books?
So Apache for windows is more secure than Apache for linux?
encryption has been cracked before by sensing the changes in current in an IC.
You will always have control (or at least someone will and that's usually enough)
SFAIK, cleanroom (which is what I described) has been taken to court and is ok.
. cgih tm
http://www.macnewsworld.com/story/33839.html
and
http://www.chillingeffects.org/reverse/faq
and for the UK/europe
http://www.jenkins-ip.com/serv/serv_6.
Also look up "abstraction, filtration, comparison" it tells you exactly what the courts do.
Why do you think Apple changed their DRM instead of taking Real to court.
dynamic linking doesn't count....f o/lpdi1.0/treatise22. html
Read....
http://digital-law-online.in
and Read the GPL...
'work based on the Program" means either the Program or any derivative work under copyright law'
My argument is that when I dynamically link an application I CAN only EXPECT equivalent functionality as the library, not the library itself.
Which is why you cannot way it is derived from anything more than the interface, and the interface isn't copyrightable.
'The act of linking against someone else's code is the same as incorporating their code into your program'
/. stops working.
/..., also see points 1 and 2.
t ise22. html
No it's not, you are an Idoit if you link dynamically and expect to have exactly the same library on the users machine.
"
What happens if Without their code, your program won't work"
Wrong again... because of the point above (there can be many versions of the library written by different people under different terms), and because, you can always use a application without the library.
I'm going to use an obvious case (to make it easier on you) I am a source only commercial.
I distribute my source that (must in you books) needs to be linked against a GPL library (or a library with similar interfaces).
The user doesn't have the library, so instead he picks out the functions from my source and uses them.
You can do exactly the same with a binary distribution it's just technically more difficult for 'some' people.
Under European law I have the right to reverse engineer an application for the porpoises of interoperability. If that means I didn't have the GPL library and I want to use the functions of that application then so be it.
The Law says that I can use the application in part or in whole.
"Slashdot still works perfectly whether the link to Websters works or not. "
No it doesn't it 'crashes' when you try and click on my link (if websteres is removed) in part
"In the case of linked software libraries, if the library is missing, your program won't work. That's what makes your program a derivative work of the library."
no more than
If I don't have MESA, ATI's opengl will do. etc.......
Dynamicly linking only involves headers, headers are a representation of fasts and the fasts are not bound by copyright. I am allowed by law to copy those facts, GPL or no GPL. When I distribute my dynamically linked application I only distribute my work and facts.
I suggest you read
http://digital-law-online.info/lpdi1.0/trea
oh, and the GPL..
'work based on the Program" means either the Program or any derivative work under copyright law'
see 'work under copyright law', no copyright violation no GPL.
No they don't.
They model the basic version, then spend millions of dollars on super computers to do wind tunnel simulation.... that's 'windtunnel' not every nut and bolt.
and that's just cars.
When they get the data they factor in a good margin of error (10th percentile or something) and build to that.
A the simulations have got better they've been able to reduce the margin of error. If you look at old plains being build they were at least 100% bulkier than todays plains.
that's what the parent said.
' Altered source versions must be plainly marked as such, and must not be misrepresented as being the original software.' says that I can cut and paste from libpng so long as I put a comment around the code.
if you think that there is clear need for LGPL, then please google abstraction, filtration, comparison.
LGPL is only good for dynamiclinking.
that's a 'anomaly' or applying a technical process in order to reflect the underling GPL library and not the generality of dynamic linking.
I still don't believe that it counts because the GPL library can be anything..
Abstraction, Filtration, Comparison would probably have to go on the side of Satan with the exception of extrema prejudice.
SFAIK, cleanroom (which is what I described) has been taken to court and is ok.
http://www.macnewsworld.com/story/33839.html
Thanks to that bit of googling I've found the exact defnintion and process that the courts use to determine if a work is a copyright infringement, it's called 'abstraction, filtration, comparison'
"
Courts use the "abstraction, filtration, comparison" test to determine if the defendant's accused software infringes the copyright on the plaintiff's software. Basically, this test compares the original work with the accused work by removing any parts that are strictly functional (permitted to be copied) and examines the sameness of what is left."
my "but my diffs to Webster's could apply to lots of books" argument, was an artistic one. I produce the diffs (with the intent of them being against websters) and publish them in the knowledge that they can do nothing about it (since they don't even know who they are).
Yes, I don't have access to Windows or Directx 9, but look http://www.oliverthered.f2s.com/projects/wine/
I can independently create a DirectX 9 library with nothing more than the API for the library and some example programmes written for Directx 9.
This is why I assert, GPL can not cover dynamic linking. (plagiarism is something completely different)
cleanroom libraies have a long tradition of getting around copyright laws, I can cleanroom any GPL library you want me to, so GPL can not cover dynamic linking.
'However, hypothetically, you might agree with most of Webster's definitions but not agree with a few words from lots of the entries (or might think the some entries needed updating due to changes in the language). You might, therefore, want to put a pile of instructions or diffs here for which words to change (e.g.: "change word 5 of line 18 of page 789 from `intellectual property' to `exclusive intellectual exploitation rights'").'
..... but you have to find the books to prove it. The problem is, it could match a thousand books out of all the books that have ever been published.
But I'm not posting one line diffs to GPL code and trying to sell it as my own.
Also you diff included `intellectual property', what about
"change word 5, line 18 page 789 'exclusive intellectual exploitation rights'"
hmm I can see an art project, now how many books could "change word 5, line 18 page 789 'exclusive intellectual exploitation rights'" be applied to and still be valid, I would expect at least a few thousand different titles.
Now I can publish anonymous changes against copyrighted works (in the form of a diff)
"change word 5, line 18 page 789 'exclusive intellectual exploitation rights'"
and produce a book, that contains nothing that is copyrightable
all hypothetically.
I could do that for you now with any library you choose.
well, unless a particular algorithm is a trade secret, which is almost never the case (and certainly not for GPL'ed code.)
So, given 1 me, and 1 GPL library, that GPL library can cover dynamic linking 0% of the time, so no need for LGPL, which was my original argument, and why I have posted so many replys to buck the trend of 'wrongly' asserting that the GPL covers dynamic linking.
sorry, wrong meeting.
GPL can never pratically cover Dynalic linking because when you 'dynamically link' an application you don't link, you just compile, which can take place with only the headers present, the user does the linking.
I may possibly consent on that minor usage of dynamic linking (I've almost never see outside of a hello world), but again the GPL cannot cover the wider usage of dynamic linking, Generally the 'value added' and function of the application calling the libraries is so great that the user cannot tell what function the underling libraries perform. This is nothing like a diff.
If I implement and entire graphics driver as a linux module (ignoring it's binary module clause) how could it ever be considered a diff when perhaps it references linux less than 0.01% of the time.
My point is that I could give the headers I've created to anyone that are Pblic domain but provide the same interface as some GPL libraries.
That person could link against the headers and a description of the function of each call without even knowning about the GLP libraries (and certainly without agreeing to the GPL) because when you dynamically link the compiler only looks at the headers. (i.e. the author of the applicaion didn't do any linking at all, that's done dynamically by the user)
This would constiture a work created without the presence or knowlage of any GPL material, so must be considered to be independant (cleanroom).
Because I can do this for any GPL library I choose, the GPL cannot cover dynamic linking, it's impossible.
"if I took a copyrighted novel and made a diff for it along the lines of "change line 35 on page 63 to read..."
I'd be interested to read that legislation, I've never seen it (it doesn't appear in
the Berne convention), or the FSF documentation. I'm very sceptical as to the scope of your argument, it doesn't take long before "change line 35 on page 63 to read..." becomes a work in it's own right.
I still don't believe this affects dynamic linking in any way, it's nothing like a diff.
"the ONLY time dynamic linking forms a derived work is when your work is based on the work you are dynamic linking to as opposed to being a truly separate work."
Which, if I create my own representation of GPLed headers can be never, meaning that the GPL cannot cover dynamic linking. (without a search warrant for headers that may no longer exist). This is why I can give you example after example and be right every time.
1."No because it is not based on Linux (assuming you don't include Linux code or code based on Linux). It isn't even a similar thing. One is a kernel; the other is a manual. (Also, note that Linux isn't a good example to use here as it has an added clause allowing linking that wouldn't normally be allowed by the GNU GPL.)"
......Because we can't because it's impossible.
And what is the difference? none, code, a book, a speech all derived works from your point of view.
If I take linux away from 'Linux for dummies' it becomes useless it cannot be used independently.
So where does it end? (at the dynamic linking level?)
"Microsoft wouldn't agree with you on interfaces", well they lost that one in court didn't they.
"but the FSF (and most same people) do. For that reason, if the dynamic linking uses a standard interface it is not considered to be based on the original work wrt copyright law. If your software doesn't require those libraries (e.g.: because it can be used with others created independently), it is definitely not based on the libraries."
What is a 'standard' interface?
I can write my own version of the libraries with the same interfaces, how is this different? I personally can use any programme you give me 'without' the dynamic libraries being present, just because it's technically difficult for you to use it independently doesn't make it a fact.
"For more information on how the FSF believe this works in Prentice"
For the above two reasons it doesn't.
you have failed to define a 'derived' work in a way that says I cannot dynamically link a commercial product against GPL. you have wishfully thought.
The FSF is very fluffy, because you cannot define a derived work in that way, it's impossible.
"Mere aggregation of two programs means putting them side by side on the same CD-ROM or hard disk. We use this term in the case where they are separate programs, not parts of a single program. In this case, if one of the programs is covered by the GPL, it has no effect on the other program."
fine....
"What constitutes combining two parts into one program? This is a legal question, which ultimately judges will decide. "
"We believe that a proper criterion depends both on the mechanism of communication (exec, pipes, rpc, function calls within a shared address space, etc.) and the semantics of the communication (what kinds of information are interchanged)."
'exec, pipes, rpc, function calls within a shared address space, etc' are all just semantics put in to confuse the reader into thinking that they represent something different.
"If the modules are included in the same executable file, they are definitely combined in one program."
Yes, we agree on that one (so far as distribution).
"If modules are designed to run linked together in a shared address space, that almost surely means combining them into one program.",
for 'almost surely' read doesn't , you cannot design software in that way, you design it with the INTENT of a SIMILAR interface being present at runtime. (e.g. what happens when version 1.4 of a library is released, doe the interface become 'standard' ?, what about mpg321 and mpg123, or GTK and QtGTK?
"By contrast, pipes, sockets and command-line arguments"....... there we go, trying to confuse people again.
With RPC my application could be running on a server on the other side of the universe from the library, going over all manner of pipes, sockes and command-line arguments. is HTTP also covered by the GPL?
"Read the GNU GPL FAQ."
/usr/include/not_linux/kernel.h
// emergency (unusable) // about to die // critical // not quite so critical // ignore this of you want // something you should know // something you may want to know // somethig I may want to know
I have.
Compile against this representation of FACTS.. and you can produce a dynamically linked kernel module that isn't GPL.
because as the 'the dynamic linking is done by the end user'
Also, you still can't answer who own the copyright on 'Linux for dummies'
cat
#ifndef _LINUX_KERNEL_H
#define _LINUX_KERNEL_H
/**
* This file is not GPL, it is public domain
* Copyright oliverthered
**/
#ifdef __KERNEL__
#include <stdarg.h>
#include <linux/linkage.h>
#include <linux/types.h>
#include <linux/stddef.h>
#include <linux/compiler.h>
#include <asm/byteorder.h>
/**
* Message levels
**/
#define KERN_EMERG "<0>"
#define KERN_ALERT "<1>"
#define KERN_CRIT "<2>"
#define KERN_ERR "<3>"
#define KERN_WARNING "<4>"
#define KERN_NOTICE "<5>"
#define KERN_INFO "<6>"
#define KERN_DEBUG "<7>"
/**
* MIN max constants..
**/
#define INT_MAX ((int)(~0U>>1))
#define INT_MIN ( -1 -INT_MAX )
/* unsigned */
#define UINT_MAX (~0U)
#define LONG_MAX ((long)(~0UL>>1))
#define LONG_MIN (-1 -LONG_MAX )
/* unsinged */
#define ULONG_MAX (~0UL)
...
you get my intersting, well if telling someone to use google can be interesing, so can a joke.
Cedega don't release there source so I thought I'd write my own patch against winehead/a>.
It's about a meg or so, if you fancy testing, or playing around and implements the core of DirectX9, beyond cedega in some areas. I'm merging with wine head at the moment so wine should be on a similar level to Cedega in a month or so when shaders are fixed.
Hopefully when Cedega get some competition they will actually improve there product or release more source so that others can.
so was I.
I've got a few months research into UI's to put up when I get the time.
Including things like the answer to, just how long does it take to click that button, or Is this really a better user interface?
I would like to see soft 'modes' so that if I'm at work and I get a fault call I can switch over to 'fault call mode' and all my book marks and menus are configured optimally to 'fault call' mode, then when I'm done I can switch back to internet mode or debugging mode.
Jesus used alien tech to bend spoons with his dead grandmother whilst the random number generator looked into the future or asteroids hitting the planet and killing everyone.
Yeh, right crack open another bottle of anti-psycotics and stop trying to rule the world
Mr Bush.
could you honestly trust someone who 'believed' that Jesus was here to save us, God is real, and if your a good boy you go to Heaven when you die, when all the facts point to it being all in his head?