Stallman stalls again
James Michael DuPont
>Late '99 I wrote a backend for gcc targeting the java virtual
>machine.
[SNIP]
>I recently asked RMS if he figured it would be worth my while to go
>and ask for the assignment again as I figured that after reaping
>nothing from the code for 18 months they may be more forthcoming.
>This is the response I got. RMS essentially tells me to bury the code
>in the backyard because it might be "dangerous".
I am currently involved in a similar situation.
My introspector project extracts the ASTS out of the compiler into XML,
I am getting hostile mails from the gcc/fsf group.
It seems that there are not any real rules on this,
only FUD and opinions from the side of the people trying to stop all
"Dangerous" patches to the gcc.
I will be meeting with rms about this soon and need to know his
arguments from the past. He has sent me almost identical mails as well,
I think that I will have to fight over this.
Please tell me on what is going on with this issue,
I would like some advice on what my options are.
Is is not funny how the GCC people did not try and bury this guys
attempt at the same :
http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
的'm writing a backend for GCC to compile to Java bytecode, and I'm
having some problems with function calls."
http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
this mail got him help from the compiler team.
Best Regards,
Mike
=====
James Michael DuPont
http://introspector.sourceforge.net/
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
To Unsubscribe: send mail to majo...@FreeBSD.org
with "unsubscribe freebsd-advocacy" in the body of the message
Terry Lambert
> >Late '99 I wrote a backend for gcc targeting the java virtual
> >machine.
> [SNIP]
> >I recently asked RMS if he figured it would be worth my while to go
> >and ask for the assignment again as I figured that after reaping
> >nothing from the code for 18 months they may be more forthcoming.
> >This is the response I got. RMS essentially tells me to bury the code
> >in the backyard because it might be "dangerous".
>
> I am currently involved in a similar situation.
> My introspector project extracts the ASTS out of the compiler into XML,
I think you meant ASTs (Abstract Syntax Tree, plural)?
> I am getting hostile mails from the gcc/fsf group.
Unless you are quoting a post that you made previously (and you
aren't) I don't understand the relevence. I also don't understand
why you would get email complaining about the code.
However, any code you want to add to GCC is going to have to go
through the FSF gatekeepers, so you are probably out of luck. I
personally would publicize the issue.
But I personally don't see how this is a FreeBSD-advocacy issue;
probably you wanted FreeBSD-chat? The FreeBSD-advocacy list is
about advocating FreeBSD, not about getting FreeBSD people to
advocate things on your behalf. 8-).
FWIW, I like the idea of being able to parse a program into a
data dictionary, and regenerate the code. THis is how most of
the COBOL Y2K conversion software was written, and if it were
generalized, it should allow source translation of code, which
is always a neat thing.
> It seems that there are not any real rules on this,
> only FUD and opinions from the side of the people trying to stop all
> "Dangerous" patches to the gcc.
>
> I will be meeting with rms about this soon and need to know his
> arguments from the past. He has sent me almost identical mails as well,
> I think that I will have to fight over this.
I've done one or two things that have gotten me email from RMS
in the past. I think that the answer is that you need to read
the GNU Manifesto to understand where he is coming from. He's
not really about free software (whether "free" means "free" or
it means "liberated", and you are trying to avoid using the word
so you use "free" instead), he's about the intellectual commons
and strong opposition to intellectual property law. The GPL is
just a tool he uses (and, as I've pointed out before, not really
a very good one, technically, but we aren't talking technically,
we're talking politically).
> Please tell me on what is going on with this issue,
> I would like some advice on what my options are.
Read the Manifesto. If the condemning letters are from RMS
himself, he likely understands exactly what you intend to do,
and has extrapolated the long term consequences in a binary
decision based on whether it promotes or does not promote the
GNU Manifesto. Your only argument that could convince him
has to do with the long term consequences, as they apply to
forwarding or at least not hindering the goals of the FSF.
If you are going to go in cold, and try to think on your feet,
because you can't figure out what his objections are without
him telling you, then you should just ask up front what they
are, and try to address them.
> Is is not funny how the GCC people did not try and bury this guys
> attempt at the same :
> http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
> 的'm writing a backend for GCC to compile to Java bytecode, and I'm
> having some problems with function calls."
>
> http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
> this mail got him help from the compiler team.
Duplicate URLs... Intentional?
> http://introspector.sourceforge.net/
It looks like what you are doing is intentionally to interoperate
with Microsoft's .NET; so RMS's being reticent is understandable,
given his platform and view from there.
PS: If you want to get FreeBSD folks on your bandwagon, fine, bur
like I said above, this probably belongs in FreeBSD-chat, and if
you really want the FreeBSD folks to listen to you, you may want
to think about making the front page of your project site accessible
to a browser other than Internet Explorer or a "view source" in
Netscape...
-- Terry
James Michael DuPont
> > I am getting hostile mails from the gcc/fsf group.
>
> Unless you are quoting a post that you made previously (and you
> aren't) I don't understand the relevence. I also don't understand
> why you would get email complaining about the code.
The problem is that the GCC team does not want to support and XML
interface, their argument is that a GDBM dump of the tree data is
protectable via some strange arguments.
See
http://groups.google.de/groups?selm=20020624100433.79219.qmail%40web13301.mail.yahoo.com
or
http://groups.google.de/groups?selm=20020301114404.74579.qmail%40web13301.mail.yahoo.com
> However, any code you want to add to GCC is going to have to go
> through the FSF gatekeepers, so you are probably out of luck. I
> personally would publicize the issue.
Yes, I have setup a sourceforge project and have a patch to the gcc.
> But I personally don't see how this is a FreeBSD-advocacy issue;
> probably you wanted FreeBSD-chat? The FreeBSD-advocacy list is
> about advocating FreeBSD, not about getting FreeBSD people to
> advocate things on your behalf. 8-).
I repsponded to this thread:
http://docs.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2001/freebsd-advocacy/20010225.freebsd-advocacy
> FWIW, I like the idea of being able to parse a program into a
> data dictionary, and regenerate the code. THis is how most of
> the COBOL Y2K conversion software was written, and if it were
> generalized, it should allow source translation of code, which
> is always a neat thing.
Yes, and there are many interesting projects that do this.
> > It seems that there are not any real rules on this,
> > only FUD and opinions from the side of the people trying to stop
> all
> > "Dangerous" patches to the gcc.
> >
> > I will be meeting with rms about this soon and need to know his
> > arguments from the past. He has sent me almost identical mails as
> well,
> > I think that I will have to fight over this.
>
> I've done one or two things that have gotten me email from RMS
> in the past. I think that the answer is that you need to read
> the GNU Manifesto to understand where he is coming from.
That is a good idea.
> > Please tell me on what is going on with this issue,
> > I would like some advice on what my options are.
>
> Read the Manifesto. If the condemning letters are from RMS
> himself, he likely understands exactly what you intend to do,
> and has extrapolated the long term consequences in a binary
> decision based on whether it promotes or does not promote the
> GNU Manifesto.
> Your only argument that could convince him
> has to do with the long term consequences, as they apply to
> forwarding or at least not hindering the goals of the FSF.
My project will help the users of free software and help the gcc.
> > Is is not funny how the GCC people did not try and bury this guys
> > attempt at the same :
> > http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
> > 的'm writing a backend for GCC to compile to Java bytecode, and I'm
> > having some problems with function calls."
> >
> > http://gcc.gnu.org/ml/gcc/2001-08/msg00696.html
> > this mail got him help from the compiler team.
>
> Duplicate URLs... Intentional?
Sorry, the original was :
http://gcc.gnu.org/ml/gcc/2001-08/msg00407.html
Also See the thread from Trent :
http://gcc.gnu.org/ml/gcc/2001-02/msg00895.html
> > http://introspector.sourceforge.net/
>
> It looks like what you are doing is intentionally to interoperate
> with Microsoft's .NET; so RMS's being reticent is understandable,
> given his platform and view from there.
I am not to interoperate with .NET, that is the DOTGNU project.
I would like to create a free set of tools, and have no problem putting
them under the GPL, the gcc people have a problem with the extraction
of the compiler internal data.
> PS: If you want to get FreeBSD folks on your bandwagon, fine, bur
> like I said above, this probably belongs in FreeBSD-chat, and if
> you really want the FreeBSD folks to listen to you, you may want
> to think about making the front page of your project site accessible
> to a browser other than Internet Explorer or a "view source" in
> Netscape...
Ooop... I will try and fix the html,
I use opera and mozilla, it works ok... :(
I will be redoing that page, it is hopelessly out of date :(
Regards,
mike
=====
James Michael DuPont
http://introspector.sourceforge.net/
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
To Unsubscribe: send mail to majo...@FreeBSD.org
Johnson David
> I am currently involved in a similar situation.
> My introspector project extracts the ASTS out of the compiler into XML,
> I am getting hostile mails from the gcc/fsf group.
The basic problem here (I think) is that "extracting" anything out of the
compiler, other than what it normally outputs, will create a derivative work.
It is another form of translation, which is covered by copyright law. You
have the permission to do so under the GPL, if the resulting XML extraction
is also under the GPL.
GNU is a funny organization. Among their various beliefs is that utilizations
of loopholes in the GPL are not permitted. They may feel (my guess) that your
XML extraction is a "loophole". You cannot write a non-GPL program that
directly extracts the ASTS out of the compiler. But your XML extraction can
be used by a non-GPL program for the same thing. Thus, they may feel that
your project (again, my guess) is an end run, or can be used as an end run,
around the GPL, by non-GPL programs.
> It seems that there are not any real rules on this,
> only FUD and opinions from the side of the people trying to stop all
> "Dangerous" patches to the gcc.
I don't think your project violates the GPL at all. I am confident that
copyright law and the wording of the GPL allows this. But that won't stop the
FSF from suing to prove otherwise. That latter point is all important. The
FSF doesn't need to be right, they just need to be able to outlast you in a
court of law.
Your only defense, as I see it, is to convince enough people associated with
GNU that your project is not in violation of the GPL. In order to do this,
you must argue from the perspective of the GNU Philosophy, and not logic,
law, rights or anything else.
> I will be meeting with rms about this soon and need to know his
> arguments from the past. He has sent me almost identical mails as well,
> I think that I will have to fight over this.
Having been in discussions with RMS before, I know that once he has made up
his mind on an issue, he will never change it no matter what. If he gets it
into his head that your project can be used as an end run around the GPL, you
will have lost.
One of RMS's overriding concerns is the GNU Project itself. He actually
opposes the Uniform Driver Interface because it would benefit Windows users
more than GNU users. So keep that in mind.
cheers,
David
Giorgos Keramidas
> > Your only argument that could convince him has to do with the long
> > term consequences, as they apply to forwarding or at least not
> > hindering the goals of the FSF.
>
> My project will help the users of free software and help the gcc.
I think Terry meant that you should expand this part a bit.
James Michael DuPont
--- Giorgos Keramidas <kera...@ceid.upatras.gr> wrote:
> On 2002-06-27 03:46 +0000, James Michael DuPont wrote:
> > > Your only argument that could convince him has to do with the
> long
> > > term consequences, as they apply to forwarding or at least not
> > > hindering the goals of the FSF.
> >
> > My project will help the users of free software and help the gcc.
>
> I think Terry meant that you should expand this part a bit.
Yes, I have followed his advice and updated my project webpage
: http://introspector.sf.net
and you will find an ton of chaotic and related snippets under
the blogger : http://gccintrospector.blogspot.com
I have a ton of mails written describing the project, but let me tell
in a couple of lines :
All types of data is stored about software on files in various formats,
all types of programs exist to process this information. The idea
behind the introspector is to extract that data out of the running
tools via an XML Dumper patch into a common format (XML/RDF/DAML) and
then have all types of tools to collate that into a repository of meta
data. This would then be presented back to the programmer, be
accessable at run time and even the end user might benefit.
This introspection means that a program "knows" about itself and its
environment, of course it is only interesting for a user to process
this information intelligently.
My current version contains a meta data framework in perl and a XML
dumper for the gcc. The perl collects data from the gcc, and then can
do code generation in SQL, PERL, and XML.
There is a lot of bits and bobs to the introspector, and now I am
finally putting it all together.
Also the first targets for the introspector were Perl, the gcc itself
and the DOTGNU PNET C# compiler Cscc.
Mike
=====
James Michael DuPont
http://introspector.sourceforge.net/
__________________________________________________
Do You Yahoo!?
Yahoo! - Official partner of 2002 FIFA World Cup
http://fifaworldcup.yahoo.com
To Unsubscribe: send mail to majo...@FreeBSD.org
Terry Lambert
> The problem is that the GCC team does not want to support and XML
> interface, their argument is that a GDBM dump of the tree data is
> protectable via some strange arguments.
Actually, it was your argument in your second posting that the
output of a program would be protectable under the GPL, even if
the input were not.
For that to be true, we would have to recognize the program as
an author, which is legally impossible.
Minimally, your code provides an intermediate language, which
RMS has already come out against; but it is potentially much,
much worse, from his perspective.
I think that, according to your first posting, the work you are
doing will also do one of two things, neither of which can be
tolerated by the FSF:
1) The program can abstract an idea from the code, which
therefore removes the license protections by removing
from the code that which constitutes authorship, the
expression of the idea, therefore leaving only the idea,
which cannot be protected (e.g. via the GPL).
2) The program can demonstrate abstract equivalence between
software, which would demonstrate that one expression
and another are fundamentally equivalent through uniform
transformation, and therefore provide reasonable doubt
as to even a clean-room reengineering of a piece of
software done in order to change its license (e.g. "tar"
and "GNU tar" source code equivalence).
In the first case, you provide a sieve through which one can strain
ideas embodies in expressions, thereby removing the expression, and
leaving the idea. Such a sieve would allow the removal of license,
such as the GPL, as well.
In the second case, you provide the ability to transform one
expression into another expression. Such a transformation would
demonstrate infringement of one work's expression on another's
expression, despite any techniques utilized to "clean-room" the
code. A side effect of this might be the claim that the ASTs
themselves were the expression, precluding any equivalence
engineering whatsoever.
From a legal perspective, absent modification of intellectual
property law, either outcome is "dangerous".
> > But I personally don't see how this is a FreeBSD-advocacy issue;
> > probably you wanted FreeBSD-chat? The FreeBSD-advocacy list is
> > about advocating FreeBSD, not about getting FreeBSD people to
> > advocate things on your behalf. 8-).
>
> I repsponded to this thread:
> http://docs.freebsd.org/cgi/getmsg.cgi?fetch=0+0+archive/2001/freebsd-advocacy/20010225.freebsd-advocacy
This wasn't really an advocacy issue, either.
If you want to argue it again, though, it was an issue of being
able to use GNU tools to target a proprietary virtual machine for
which there may not be a GNU equivalent. I understand the argument
there, but, if anything, it's a side issue compared to your code,
which strikes at the very heart of the social problem that Richard
believes exists, and doesn't resolve it. It cuts very close to the
bone, through the flesh of the boundary on which the GPL depends for
it use of the intellectual property system to subvert the intellectual
proerty system. Your code is subversive of their subversion. 8-).
> > Your only argument that could convince him
> > has to do with the long term consequences, as they apply to
> > forwarding or at least not hindering the goals of the FSF.
>
> My project will help the users of free software and help the gcc.
I disagree, but that's irrelevent. The question you must answer
for the FSF's approval will be whether or not it advances or it
detracts from the goals of the FSF. Free software, the GPL, gcc,
etc., are all just instrumentalities, to be used in achieving the
goals of the GNU Manifesto. If those goals could be achieved
more effciently or effectively with another approach then these
things would not have been created.
> Also See the thread from Trent :
> http://gcc.gnu.org/ml/gcc/2001-02/msg00895.html
RMS's reply here touches the fundamental distinction between "use"
and "utilize" that he attempts to obfuscate elsewhere.
Frankly, the posting of a private communication like this is bad
"nettiquite", and I rather expect that RMS wished that it had
not been published, since it demonstrates a distinction between
private and publically held positions with regard to liberty vs.
freedom. Personally, I try to be completely consistent in both
my public and private email, since I operate under the theory
that anything not in ones direct control may be published.
I won't compound the issue personally by reposting it here, but
I will leave your reference to the original, above.
RMS is correct in his analysis of the consequences of such a
project being a potential undermining of the goals of the GPL.
I think that your only chance is to make it a tradeoff that he
considers worthwhile. I think that would be very hard to do.
-- Terry
Johnson David
> Frankly, the posting of a private communication like this is bad
> "nettiquite", and I rather expect that RMS wished that it had
> not been published, since it demonstrates a distinction between
> private and publically held positions with regard to liberty vs.
> freedom.
This private communication is the heart of the whole matter. It may be against
netiquette, but it seems to me to be quite necessary given the situation. The
public writings of RMS are quite clear that information should never be
suppressed. Yet here is a letter from RMS urging information suppression.
David
Terry Lambert
> The question you must answer
> for the FSF's approval will be whether or not it advances or it
> detracts from the goals of the FSF. Free software, the GPL, gcc,
> etc., are all just instrumentalities, to be used in achieving the
> goals of the GNU Manifesto. If those goals could be achieved
> more effciently or effectively with another approach then these
> things would not have been created.
I want to emphasize this as much as possible. If you take the
time to internalize this idea, it should give you some profound
insight into the limitations of the vision under which RMS
operates.
If you can model this, you can model his behaviour, and in doing
that, you should be able to predict his reactions to potential
arguments, without biasing against your final case when you go
to present it to him.
If your proof lies outside his model, he will reject it.
-- Terry
Ted Mittelstaedt
>From: James Michael DuPont [mailto:mdupo...@yahoo.com]
>Sent: Thursday, June 27, 2002 2:32 AM
>To: d...@ofug.org; s33...@student.uq.edu.au; te...@toybox.placo.com;
>freebsd-...@FreeBSD.ORG; djoh...@acuson.com; stu...@confusion.net
>Subject: RE: Stallman stalls again
>
>
>I am currently involved in a similar situation.
>My introspector project extracts the ASTS out of the compiler into XML,
>I am getting hostile mails from the gcc/fsf group.
>
>only FUD and opinions from the side of the people trying to stop all
>"Dangerous" patches to the gcc.
>
Right. As Terry says, your doing exactly what they are doing, only you
aren't consistent with their view of the world. Thus, they will oppose
you.
>I will be meeting with rms about this soon and need to know his
>arguments from the past. He has sent me almost identical mails as well,
>I think that I will have to fight over this.
>
Why? What is it that you want from RMS?
For starters, RMS doesen't even represent the views of the GNU project,
in fact that project itself is as fractionalized as all the Linux
people with their distribution of the month mentality. The GNU has
no goals that everyone in it agrees over. They argue with each other
as much as they argue with everyone else.
RMS today is nothing more than a figurehead who's main usefulness to
anybody is that he has an opinion on your stuff.
If RMS loves something, then half the GNU will hate it and the other half
will love it.
If RMS hates something then half the GNU will love it and the other half
will hate it.
I think your operating under the delusion that getting RMS's buyoff on
something will miraculously get all the GNU to embrace it. You need to
realize that the fact that RMS hates your stuff is valuable - now you
can go wade into the fray waving your "I'm a victim of RMS" banner and
gain a following. He has done you a tremendous favor - grab it and
run!
>
>Please tell me on what is going on with this issue,
>I would like some advice on what my options are.
>
You have only one real option, and that is this:
Write useful software. If your GCC backend does something that someone
needs done, then your headed in the right direction. If it does something
that a lot of people love, then so much the better.
The GNU history is littered with projects that the core group fought
against but the users basically told the core to shove it up their ass,
then eventually the GNU capitulated and rewrote their manifesto to
include the project.
Keep this in mind: you cannot convince RMS of anything. He makes a
snap decision as to whether he likes what your doing or not, then
never changes it. By meeting him all your doing is feeding his ego,
you have no more chance of convincing him your stuff is worthwhile than
a flower has of convincing the sun to move it's rays into a better
position.
What you do to RMS is you tell him "This is what I'm doing and it's
going to be great and lots of people will like it and it's not going
to go away. So you may as well get used to it because nothing you can
say will stop me."
Your missing the point of Free software if you are seeking permission
from anyone. First your seeking permission from RMS now your seeking
permission from us. Wake up - you don't need permission from anyone.
Just do it and if it's good, enough people will start using it, THEN
you can run around doing your politicing to grease it up for easy
insertion into the GNU rectum.
Ted Mittelstaedt te...@toybox.placo.com
Author of: The FreeBSD Corporate Networker's Guide
Book website: http://www.freebsd-corp-net-guide.com