Hi,
Is there a version of zebu that works well with CMUCL. I've played
around with the latest version from cmu repository
but its pretty broken for CMUCL, especially pathnames and typechecking.
Thanks,
Tunc
>>>>> "Tunc" == Tunc Simsek <······@robotics.eecs.berkeley.edu> writes:
Tunc> Is there a version of zebu that works well with CMUCL. I've played
Tunc> around with the latest version from cmu repository
I have one that works. I think the version on ftp.digitool.com also
works.
Tunc> but its pretty broken for CMUCL, especially pathnames and typechecking.
I'm not sure mine gets around these problems. I don't recall having
problems with this.
Ray
Hi Ray,
May I have it.
Raymond Toy wrote:
>
> >>>>> "Tunc" == Tunc Simsek <······@robotics.eecs.berkeley.edu> writes:
>
> Tunc> Is there a version of zebu that works well with CMUCL. I've played
> Tunc> around with the latest version from cmu repository
>
> I have one that works. I think the version on ftp.digitool.com also
> works.
>
> Tunc> but its pretty broken for CMUCL, especially pathnames and typechecking.
>
> I'm not sure mine gets around these problems. I don't recall having
> problems with this.
>
> Ray
See
ftp://ftp.digitool.com/pub/mcl/contrib/
Andrew
http://www.andrewcooke.free-online.co.uk/index.html
In article <·················@robotics.eecs.berkeley.edu>,
······@robotics.eecs.berkeley.edu wrote:
> Hi Ray,
>
> May I have it.
>
> Raymond Toy wrote:
> >
> > >>>>> "Tunc" == Tunc Simsek <······@robotics.eecs.berkeley.edu>
writes:
> >
> > Tunc> Is there a version of zebu that works well with CMUCL. I've
played
> > Tunc> around with the latest version from cmu repository
> >
> > I have one that works. I think the version on ftp.digitool.com also
> > works.
> >
> > Tunc> but its pretty broken for CMUCL, especially pathnames and
typechecking.
> >
> > I'm not sure mine gets around these problems. I don't recall having
> > problems with this.
> >
> > Ray
>
Sent via Deja.com http://www.deja.com/
Before you buy.
In this directory I found v3.5.5. which again does not work on
CMU/Solaris. Any more options?
Also, a parser written in Zebu, how do you distribute it. It
seems that Zebu needs to be around even after you generate the .TAB
file (unlike yacc/lex where once you generate C you don't really
need YACC/LEX anymore).
Thanks,
Tunc
On Fri, 24 Mar 2000, Andrew Cooke wrote:
>
> See
>
> ftp://ftp.digitool.com/pub/mcl/contrib/
>
> Andrew
> http://www.andrewcooke.free-online.co.uk/index.html
>
> In article <·················@robotics.eecs.berkeley.edu>,
> ······@robotics.eecs.berkeley.edu wrote:
> > Hi Ray,
> >
> > May I have it.
> >
> > Raymond Toy wrote:
> > >
> > > >>>>> "Tunc" == Tunc Simsek <······@robotics.eecs.berkeley.edu>
> writes:
> > >
> > > Tunc> Is there a version of zebu that works well with CMUCL. I've
> played
> > > Tunc> around with the latest version from cmu repository
> > >
> > > I have one that works. I think the version on ftp.digitool.com also
> > > works.
> > >
> > > Tunc> but its pretty broken for CMUCL, especially pathnames and
> typechecking.
> > >
> > > I'm not sure mine gets around these problems. I don't recall having
> > > problems with this.
> > >
> > > Ray
> >
>
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
>
>
3.5.5 seems to be the latest. I don't have CMU - it works with CLISP.
I also have 3.5.3 and 3.4.8 (can't be sure 3.4.8 is unedited as it is in
a directory, not a gzipped tar file). I can email you either of those
if you want to try them, but I doubt it would help...
As for distribution - I guess you could distribute Zebu too (it's
licence conditions are pretty relaxed). Again, I've got no experience
of distributing it - I just used it to play with (there's a nice package
that uses Zebu and rewrites parse trees).
I couldn't find the original author about 6 months ago.
Good luck,
Andrew
In article
<·········································@tudor.EECS.Berkeley.EDU>,
Tunc Simsek <······@tudor.EECS.Berkeley.EDU> wrote:
> In this directory I found v3.5.5. which again does not work on
> CMU/Solaris. Any more options?
>
> Also, a parser written in Zebu, how do you distribute it. It
> seems that Zebu needs to be around even after you generate the .TAB
> file (unlike yacc/lex where once you generate C you don't really
> need YACC/LEX anymore).
>
> Thanks,
> Tunc
Sent via Deja.com http://www.deja.com/
Before you buy.
Andrew Cooke <······@andrewcooke.free-online.co.uk> writes:
>
> In article
> <·········································@tudor.EECS.Berkeley.EDU>,
> Tunc Simsek <······@tudor.EECS.Berkeley.EDU> wrote:
> > In this directory I found v3.5.5. which again does not work on
> > CMU/Solaris. Any more options?
I'm using it here with CMU/Solaris. I have written some defsystem
files that kinda work, but I am still hesitant to change files from
the Zebu distribution themselves (some package and parameter issues
need to be fixed.)
> As for distribution - I guess you could distribute Zebu too (it's
> licence conditions are pretty relaxed). Again, I've got no experience
> of distributing it - I just used it to play with (there's a nice package
> that uses Zebu and rewrites parse trees).
IMO it is anything but pretty relaxed. If I have seen it right
(COPYRIGHT file of the 3.5.5 distribution), it's a (very slightly
edited) GPL and you need the ZEBU package (albeit not ZEBU-COMPILER)
with your application, so anything that uses Zebu is GPL also. (Note
that I do not want to start any flames here -- the author makes the
distribution rules and I respect his decision.)
> I couldn't find the original author about 6 months ago.
Neither could I.
If anyone wants to use / debug my zebu[-compiler].defsystem, drop me a
mail. I was thinking of making a nice package for CLOCC, but haven't
come around to clean everything up yet.
Rudi
In article <··············@ist.tu-graz.ac.at>,
Rudolf Schlatte <········@ist.tu-graz.ac.at> wrote:
> Andrew Cooke <······@andrewcooke.free-online.co.uk> writes:
> > As for distribution - I guess you could distribute Zebu too (it's
> > licence conditions are pretty relaxed). Again, I've got no
experience
> > of distributing it - I just used it to play with (there's a nice
package
> > that uses Zebu and rewrites parse trees).
>
> IMO it is anything but pretty relaxed. If I have seen it right
> (COPYRIGHT file of the 3.5.5 distribution), it's a (very slightly
> edited) GPL and you need the ZEBU package (albeit not ZEBU-COMPILER)
> with your application, so anything that uses Zebu is GPL also. (Note
> that I do not want to start any flames here -- the author makes the
> distribution rules and I respect his decision.)
sorry - my post was misleading. since all my own code is gpl, i
simple read the licence some time in the past and thought "no problems
there"...
andrew
Sent via Deja.com http://www.deja.com/
Before you buy.
I've written some yacc parsers before without worrying about the
distribution,
I guess the reason is that yacc generates C code which distributes well
under
all kinds of licenses including GPL.
IMO, lisp code does not distribute well under GPL, especially in this
case
where as Rudolf points out that if you use my Zebu generated parser,
both my parser will
have to be GPL and your code too.
BTW, the parser I'm working on is a Matlab(ish) parser that simplifies
some mathematical
expressions. The parser will work like this:
(with-matlisp-syntax ( ... code generation options)
l = expm(a*t)*x;
z = l/norm(l);
etc ... )
Tunc
Andrew Cooke wrote:
>
> In article <··············@ist.tu-graz.ac.at>,
> Rudolf Schlatte <········@ist.tu-graz.ac.at> wrote:
> > Andrew Cooke <······@andrewcooke.free-online.co.uk> writes:
> > > As for distribution - I guess you could distribute Zebu too (it's
> > > licence conditions are pretty relaxed). Again, I've got no
> experience
> > > of distributing it - I just used it to play with (there's a nice
> package
> > > that uses Zebu and rewrites parse trees).
> >
> > IMO it is anything but pretty relaxed. If I have seen it right
> > (COPYRIGHT file of the 3.5.5 distribution), it's a (very slightly
> > edited) GPL and you need the ZEBU package (albeit not ZEBU-COMPILER)
> > with your application, so anything that uses Zebu is GPL also. (Note
> > that I do not want to start any flames here -- the author makes the
> > distribution rules and I respect his decision.)
>
> sorry - my post was misleading. since all my own code is gpl, i
> simple read the licence some time in the past and thought "no problems
> there"...
>
> andrew
>
> Sent via Deja.com http://www.deja.com/
> Before you buy.
In article <·················@robotics.eecs.berkeley.edu>,
······@robotics.eecs.berkeley.edu wrote:
> IMO, lisp code does not distribute well under GPL, especially in this
> case
> where as Rudolf points out that if you use my Zebu generated parser,
> both my parser will
> have to be GPL and your code too.
Depending on personal politics this is not necessarily a Bad Thing, of
course.
Andrew
Sent via Deja.com http://www.deja.com/
Before you buy.
Tunc Simsek <······@robotics.eecs.berkeley.edu> writes:
> I've written some yacc parsers before without worrying about the
> distribution,
> I guess the reason is that yacc generates C code which distributes well
> under
> all kinds of licenses including GPL.
>
> IMO, lisp code does not distribute well under GPL, especially in this
> case
> where as Rudolf points out that if you use my Zebu generated parser,
> both my parser will
> have to be GPL and your code too.
In principle there isn't any difference between C and CL here. A
Bison-generated parser will include the C parser skeleton that comes
with Bison, so your parser (and quite possibly the whole program you
integrate that parser with) will come under the same licence that the
Parser skeleton came under. There was a time when the parser skeleton
of Bison also came under the GPL, and so Bison-generated parsers
couldn't be used in non-GPL programs. This was changed only a couple
of years ago with version 1.24. To quote from the Bison Texinfo doc:
<quote>
As of Bison version 1.24, we have changed the distribution terms for
`yyparse' to permit using Bison's output in non-free programs.
Formerly, Bison parsers could be used only in programs that were free
software.
The other GNU programming tools, such as the GNU C compiler, have
never had such a requirement. They could always be used for non-free
software. The reason Bison was different was not due to a special
policy decision; it resulted from applying the usual General Public
License to all of the Bison source code.
The output of the Bison utility--the Bison parser file--contains a
verbatim copy of a sizable piece of Bison, which is the code for the
`yyparse' function. (The actions from your grammar are inserted into
this function at one point, but the rest of the function is not
changed.) When we applied the GPL terms to the code for `yyparse', the
effect was to restrict the use of Bison output to free software.
We didn't change the terms because of sympathy for people who want to
make software proprietary. *Software should be free.* But we
concluded that limiting Bison's use to free software was doing little to
encourage people to make other software free. So we decided to make the
practical conditions for using Bison match the practical conditions for
using the other GNU tools.
</quote>
So in principle you have the same problems with C and CL. In practice,
it might turn out that since C will force a rather complete separation
between generator and generated code/runtime support code, licensing
both under different conditions will be easier to do.
Regs, Pierre.
--
Pierre Mai <····@acm.org> PGP and GPG keys at your nearest Keyserver
"One smaller motivation which, in part, stems from altruism is Microsoft-
bashing." [Microsoft memo, see http://www.opensource.org/halloween1.html]
····@acm.org (Pierre R. Mai) writes:
[...]
> There was a time when the parser skeleton
> of Bison also came under the GPL, and so Bison-generated parsers
> couldn't be used in non-GPL programs. This was changed only a couple
> of years ago with version 1.24. To quote from the Bison Texinfo doc:
Thanks for the reference - I *knew* I had seen this, but couldn't
remember where or when.
> So in principle you have the same problems with C and CL. In practice,
> it might turn out that since C will force a rather complete separation
> between generator and generated code/runtime support code, licensing
> both under different conditions will be easier to do.
In the case of Zebu, the runtime system and compiler can be separated
quite cleanly as well. All the install methods of the original
distribution gave the possibility of loading only the runtime.
Andrew Cooke <······@andrewcooke.free-online.co.uk> writes:
> See
>
> ftp://ftp.digitool.com/pub/mcl/contrib/
Didn't someone mention that they were trying to maintain a Zebu
version? If the license of Zebu doesn't prohibit it, it might be a
good idea to add it to CLOCC (http://clocc.sourceforge.net/)?
--
------------------------------------------------------------------
Stig Erik Sandoe ····@ii.uib.no http://www.ii.uib.no/~stig/