🚀 go-pugleaf

%% Regarding Re: GPL, LGPL;
%% burley@gnu.ai.mit.edu (Craig Burley) writes:

    cb> In article <ARTK.95Mar3105505@cc-color1.Congruent.COM>
    cb> artk@Congruent.COM (Arthur Kreitman) writes:

    cb>        I've heard this claim that linking with a bison output
    cb>    make your code completely subject to the gpl.  I'm wondering
    cb>    what the legal foundation for that claim is.  To me, its
    cb>    always seemed to be an unsupportable claim.

Well, look at it this way: before you use GNU code you must agree to the
terms of the licensing.

The terms say you can't give any application containing the GNU code to
anyone else unless you also provide the complete source for that
application, so they can rebuild it and/or modify it, *and* that you
cannot restrict what they do with it.  Etc.

These are the terms with which the code is provided; if you don't want
to abide by them you can't use the code: nobody says you have to do so.

Although it's true these claims have never been tested in court I
wouldn't bet they wouldn't be upheld.  Looked at in this way they don't
seem too far out.

They seem to me to be at least as reasonable as those silly licenses you
get with most commercial software now that says you've just "licensed"
the software, and you're not allowed to resell your copy to anyone else,
etc.  What kind of a crock is that?  I *bought* it, I *own* it.  If I
don't want/need it anymore, I can darn well sell it to someone else if I
want to (as long as I don't keep a copy myself of course).  I don't buy
a "license" for a book, or a videotape, or a CD.

    cb> It's not so much the use of bison output, it's the inclusion of
    cb> the copyrighted parser skeleton.  You can write your own
    cb> skeleton to get around the problem, but apparently lots of
    cb> people use some other free yacc replacement that doesn't have
    cb> all of bison's copyright restrictions.

Most people use Berkeley yacc (byacc), found at most large ftp sites.

There's an even better, IMHO, alternative out now called PCCTS which is
a true improvement in parser/lexer technology, not just a reworking of
yacc.  It was developed (and still is) by Terence Parr at Purdue
University.  Quite cool, check it out.

Here's a telling statement: almost ready is a complete C++ grammar
written in PCCTS (it'll be out shortly--the people at NeXT did the
original PCCTS grammar for their NeXTStep stuff and contributed it to
Terence and he's polishing it) with no odd hacks or weird
tricks... how's *that* for a powerful parser generator!  Like to see
anyone do that in yacc or bison :)

                                                                paul
 _____________________________________________________________________________
/                                                                             \
 Paul D. Smith                   | You're a mad bastard, Peter Gabriel!
 <psmith@wellfleet.com>          | 
 Bay Networks, Inc.              |                    -- Bono (U2)
 Network Management Development  |                       1992 MTV Music Awards
\_____________________________________________________________________________/