RIPEM 1.1 Available
mathew
>+++ Included modifications to RSAREF to use the faster GMP
>(GNU Multiple Precision) library. This is not integrated into
>RSAREF because of possible conflicts with GMP's GNU license.
>See rsaref/test/gmp-hooks-for-rsaref.tar.Z. (Mark Henderson)
So we have a program with a very restrictive license (RSAREF), being
combined with what I assume is GPL or GLPLed software.
Yet by simply leaving the user to do the final linking together, the
GNU license has seemingly been completely bypassed. PKP are free to take
advantage of GNU software without having to give users permission to
freely redistribute RSAREF or freely re-link it.
Also, if these modifications to the RSAREF library have been rubber-stamped
through by PKP's corporate lawyers, why won't they allow the modifications
to let PGP (a freely-redistributable program, unlike RIPEM) be linked with
RSAREF?
It seems to me that PKP is being pretty obnoxious towards the cause of free
software. It is trying to disallow legal use of freely-redistributable
software (PGP), whilst taking advantage of freely-redistributable software for
its own use.
Will the FSF have anything to say about this, I wonder?
mathew
Richard Pieri
potential infringement against the GPL and the GNU Copyleft. If there is an
infringement, RMS will be sure to deal with it appropriately.
--Rat <rat...@ccs.neu.edu> Northeastern's Stainless Steel Rat
PGP Public Key Block available upon request Ask about rat-pgp.el
||||| | | | | | | | | | | | | | | | | | | | | | | |||||
One likes to believe in the freedom of baseball. --Geddy Lee
Mark C. Henderson
>I've forwarded the announcement to Richard M. Stallman, pointing out the
>potential infringement against the GPL and the GNU Copyleft. If there is an
>infringement, RMS will be sure to deal with it appropriately.
As we don't include any code from gmp or any code under the GPL in
the RIPEM distribution, there is no infringement.
I don't expect the FSF to like this. But on the other hand
NO GNU CODE IS INCLUDED IN THE RIPEM DISTRIBUTION
Mark
--
Mark Henderson
ma...@wimsey.bc.ca (personal account)
RIPEM key available by key server/finger/E-mail
MD5OfPublicKey: F1F5F0C3984CBEAF3889ADAFA2437433
Mark C. Henderson
>Also, if these modifications to the RSAREF library have been rubber-stamped
>through by PKP's corporate lawyers, why won't they allow the modifications
>to let PGP (a freely-redistributable program, unlike RIPEM) be linked with
>RSAREF?
RSAREF users have a licence to make performance enhancements. The
RIPEM developers have already replaced much of the large integer
code in RSAREF independently of the gmp hooks.
The legal problem with PGP+RSAREF is that PGP needs to access
RSAREF at a lower level than the published interface. RIPEM
doesn't.
to...@gnu.ai.mit.edu
Organization: Mantis Consultants Ltd, Cambridge. UK.
From: mat...@mantis.co.uk (mathew)
Sender: gnu-misc-dis...@cis.ohio-state.edu
In alt.security.ripem,sci.crypt m...@scss3.cl.msu.edu (Mark Riordan) writes:
>+++ Included modifications to RSAREF to use the faster GMP
>(GNU Multiple Precision) library. This is not integrated into
>RSAREF because of possible conflicts with GMP's GNU license.
>See rsaref/test/gmp-hooks-for-rsaref.tar.Z. (Mark Henderson)
So we have a program with a very restrictive license (RSAREF), being
combined with what I assume is GPL or GLPLed software.
Yet by simply leaving the user to do the final linking together, the
GNU license has seemingly been completely bypassed. PKP are free to take
advantage of GNU software without having to give users permission to
freely redistribute RSAREF or freely re-link it.
Leaving the link to the user is illegal and does not limit PKP's
liability for violating the GPL in anyway.
-len
Member, League for Programming Freedom. Ask: l...@uunet.uu.net
Richard Pieri
Mark> As we don't include any code from gmp or any code under the GPL in
Mark> the RIPEM distribution, there is no infringement.
Mark> I don't expect the FSF to like this. But on the other hand
Mark> NO GNU CODE IS INCLUDED IN THE RIPEM DISTRIBUTION
But RIPEM does include specific hooks for GPL code. This may be an
infringement, it may not.
--Rat <rat...@ccs.neu.edu> Northeastern's Stainless Steel Rat
PGP Public Key Block available upon request Ask about rat-pgp.el
||||| | | | | | | | | | | | | | | | | | | | | | | |||||
And now we meet again, for the first time, for the last time. --Dark Helmet
Mark C. Henderson
>Leaving the link to the user is illegal and does not limit PKP's
>liability for violating the GPL in anyway.
GNU code. We aren't. So I'm going to repeat myself here:
THE RIPEM DISTRIBUTION DOES NOT INCLUDE ONE LINE OF GNU CODE.
Now as for PKP's liability. They have none.
I personally wrote these hooks. I'm not employed by PKP or RSADSI.
I am giving these hooks away for free.
Surely you're not asserting that the following code is covered by the
GPL. (Apologies in advance if it has any errors, I just typed this in
off the top of my head).
/* THIS CODE IS HEREBY PLACED IN THE PUBLIC DOMAIN */
#include <gmp.h>
main()
{
MP_INT a,b,c;
mpz_init(&a); mpz_init(&b); mpz_init(&c);
mpz_set_ui(a,1);
mpz_set_ui(b,1);
mpz_add(c,a,b);
}
Of course, if I compile gmp, compile this code, and link it, the
resulting binary is covered by the GPL.
If I compile the above fragment link it with a free library which has
functions called mpz_init, mpz_set_ui, and mpz_add then the resulting
binary is not covered by the GPL.
Pretty simple.
I repeat. PKP and RSADSI have nothing to do with this (they probably don't
even know about it). If you want to go after someone, you'll have to go
after me personally. I developed this with my own time and own equipment.
Now, I'm going to say it again, for all those who have thick skulls
(you know who you are >:-> ).
THE RIPEM DISTRIBUTION DOES NOT INCLUDE ONE LINE OF GNU CODE.
What it does include is some code that calls various functions with names
like mpz_mul, mpz_sub, mpz_add &c.
Enough said,
(I sure I hope so)
Mark Henderson
Marc VanHeyningen
Thus said ma...@vanbc.wimsey.com (Mark C. Henderson):
>In article <RATINOX.93...@denali.ccs.neu.edu> rat...@denali.ccs.neu.edu (Richard Pieri) writes:
>>I've forwarded the announcement to Richard M. Stallman, pointing out the
>>potential infringement against the GPL and the GNU Copyleft. If there is an
>>infringement, RMS will be sure to deal with it appropriately.
Make sure you mention that the distribution also includes makefiles
for compiling the software with GNU's C compiler, using the software
from within GNU Emacs, and who knows what other horrible stuff.
>As we don't include any code from gmp or any code under the GPL in
>the RIPEM distribution, there is no infringement.
>
>I don't expect the FSF to like this. But on the other hand
>NO GNU CODE IS INCLUDED IN THE RIPEM DISTRIBUTION
Curiously, the GMP distribution appears to be covered by the vanilla
GPL and not by the Library GPL; is this correct? If it's covered by
the former, then distributing precompiled binaries with GMP linked in
(which nobody is doing, to my knowledge) would seem a violation.
--
Marc VanHeyningen mvan...@cs.indiana.edu MIME, RIPEM & HTTP spoken here
Christopher Davis
M> == mat...@mantis.co.uk (mathew)
LT> == Len Tower <to...@gnu.ai.mit.edu>
MRR> +++ Included modifications to RSAREF to use the faster GMP (GNU
MRR> Multiple Precision) library. This is not integrated into RSAREF
MRR> because of possible conflicts with GMP's GNU license. See
MRR> rsaref/test/gmp-hooks-for-rsaref.tar.Z. (Mark Henderson)
M> So we have a program with a very restrictive license (RSAREF), being
M> combined with what I assume is GPL or GLPLed software.
No, merely being distributed with optional patches that allow it to be
linked with GPL'ed code. No GPL'ed code is included in the distribution.
Or does the GPL now assert some form of rights to the programmatic interface?
M> Yet by simply leaving the user to do the final linking together, the
M> GNU license has seemingly been completely bypassed. PKP are free to take
M> advantage of GNU software without having to give users permission to
M> freely redistribute RSAREF or freely re-link it.
Since it's distributed as source, it's hard to say that they're not
allowing users to freely re-link it. As for freely distributing it, the
RSAREF license included in the package, while nowhere near as useful as the
GPL in terms of
LT> Leaving the link to the user is illegal and does not limit PKP's
LT> liability for violating the GPL in anyway.
Which, it would seem, they have not done, as they have not distributed
GPLed source code, nor binaries compiled from GPLed source.
It's not as if they're distributing a modified version of gmp that's only
useful with their proprietary library or anything, either.
--
* Christopher Davis * <c...@eff.org> * <c...@kei.com> * [CKD1] * MIME * RIPEM *
"Those who cannot remember history are doomed to repost it every month,
with diffs marked with change bars." --Ed Vielmetti <e...@msen.com>
david carlton
> Leaving the link to the user is illegal and does not limit PKP's
> liability for violating the GPL in anyway.
I haven't looked at the source code in question, but if the above is
intended as a general statement, I don't understand why it should be
true. It seems to me that, in general, if I write a program that
happens to work with a GPL'd library, I don't have to put my program
under the GPL, because I'm not using the library's code, merely
conforming to its interface. Is that not so, or is there something
special about this case? Or is it the case that it's okay for me to
write a program that happens to work with a GPL'd library but once it
is linked with that library, the whole aggregate becomes a GPL'd work,
so in this case you could maybe distribute the program but it would be
illegal for the user to link it? (Of course, things change if the
library is GLPL'd, but if I'm remembering right, gmp isn't.)
david carlton
car...@husc.harvard.edu
.. does your DRESSING ROOM have enough ASPARAGUS?
Marc VanHeyningen
>No, merely being distributed with optional patches that allow it to be
>linked with GPL'ed code. No GPL'ed code is included in the distribution.
>
>Or does the GPL now assert some form of rights to the programmatic interface?
Considering the feelings RMS has about interface copyrights, it seems
ironic, not to mention disturbing, that the FSF would assert what
effectively amounts to one.
Ozan S. Yigit
...
What it does include is some code that calls various functions with names
like mpz_mul, mpz_sub, mpz_add &c.
Can the RIPEM library use something else, say the DEC (paris labs)
bignum library instead of gmp?
oz
Tim Smith
Instead of writing one, I run GNU Emacs on a pty and use it as the text
editing engine.
I do not include GNU Emacs in my distribution. My word processor simply
requires that there be a program named "emacs" in the user's path, and
that it implements the GNU Emacs commands.
Would FSF claim that my word processor falls under GPL?
--
"Pope moved that we strike from the State's brief and appendix a selection from
the Year Book of 1484 written in Medieval Latin and references thereto. The
State provided no translation and conceded a total lack of knowledge of what it
meant. The motion is granted" 396 A.2d 1054 --Tim Smith
Per Abrahamsen
It is interesting to compare to scenarios:
1) Write a program that uses the public interface of a GPL'ed
library. Distribute the program (without library) under a different
copyright.
2) Modify a GPL'ed program by writing a library for the new
functionality, and change the program to use this library. Distribute
the changed program under the GPL and distribute the library under a
different copyright.
Apparently #1 is not legal (RIPEM) but #2 is legal (Win-Emacs).
This somehow seems wrong to me.
Mark C. Henderson
>Considering the feelings RMS has about interface copyrights, it seems
>ironic, not to mention disturbing, that the FSF would assert what
>effectively amounts to one.
>--
>Marc VanHeyningen mvan...@cs.indiana.edu MIME, RIPEM & HTTP spoken here
A shame too, since I have some code that I was going to contribute
to gmp.
Now, I'm not inclined to.
Tim Smith
>Leaving the link to the user is illegal and does not limit PKP's
>liability for violating the GPL in anyway.
Elaborate on this remarkable claim, please. If PKP includes no
copyrighted GPL'ed code in their code, how can they be liable? Wouldn't
the only one liable, if anyone, be the user who does the link?
Yggdrasil Computing
> GNU license has seemingly been completely bypassed. PKP are free to take
> advantage of GNU software without having to give users permission to
> freely redistribute RSAREF or freely re-link it.
>
>Leaving the link to the user is illegal and does not limit PKP's
>liability for violating the GPL in anyway.
I've heard this claim made occasionally by FSF people and I'd
like to learn more about the legal basis for it, especially in light
of Computer Associates v. Altai. I've read the FSF posting on how
an attorney (who is being paid by the FSF to think this?) thinks that
a court would consider separate linking to be a subterfuge, but I've
never seen it backed up with quotations from similar cases or legal
references.
In the absence of a clearer legal justification, it appears that
FSF is endorsing a kind of interface copyright and is relying primiarily
on fear, uncertainty and doubt. I realize that there is a certain amount
of FU&D in any legal argument since the law is sufficiently imprecise
to require judges to interpret it, but in cases where the claim being
made seems unjustified to most people, it is reasonable to at least
demand a more thorough explanation.
--
Adam J. Richter Yggdrasil Computing, Incorporated
409 Evelyn Ave., Apt. 312, Albany CA 94706 PO Box 8418, Berkeley CA 94707-8418
(510) 528-3209 (510) 526-7531, fax: (510) 528-8508
ad...@netcom.com yggd...@netcom.com
Another member of the League for Programming Freedom (l...@uunet.uu.net).
James Davies
>Suppose I write a fancy word processor. It needs a text editing engine.
>Instead of writing one, I run GNU Emacs on a pty and use it as the text
>editing engine.
>
>I do not include GNU Emacs in my distribution. My word processor simply
>requires that there be a program named "emacs" in the user's path, and
>that it implements the GNU Emacs commands.
>
>Would FSF claim that my word processor falls under GPL?
Do all Objective-C programs fall under the GPL?
Michael Golan
This is silly. Neither of these fall under the GPL.
Why? because rms clearly stated so, and indeed, any reasonable person
reading the GPL would come to the same conclusion. Such posts do not
help in clarifying the problems with the GPL, contrasting it with PD,
or solving the issue of RIPEM/GMP.
The real question is, what is a "program" an "interface" an "os"
a "linking", etc. Most people fail to see why talking to emacs over
a pipe (as above) is different than talking to emacs via shared
memory, shared address space (threads), os hook, dynamic library calls,
or static library calls. These people (including me) think they are
all allowed (by copyright law.) The FSF/GPL claims a big difference.
This claim can only be decided by the courts, and I suspect it
(a) haven't been answered yet and (b) is likely to be wrongly
answered by a judge who don't understand computer architecture :-)
For now, the difference is clear. It is whatever rms say it is...
So talking to emacs over a pipe is ok, and linking with GMP isn't.
Everything in the middle - we are still waiting for explicit
reply from rms (I'm told soon!)
-- Michael
--
Michael Golan | Duel, PD add-on to gdb, allows "x[..100] >? 0" to
m...@cs.princeton.edu | show the positive elements of x in the debugger, etc.
| annon ftp ftp.cs.princeton.edu:/duel or send me mail!