gnu@hoptoad.uucp (John Gilmore) (04/01/88)
jdc@naucse.UUCP (John Campbell) wrote: > > (1) My understanding of BISON is that if you use it, then > > the thing you created with it must be freely distributed. > > Agreed? This is true, since bison's output includes a literal copy of a parser (bison.simple or bison.hairy) that contains the GNU copyright. You could write your own parser, pulling out the copyrighted GNU one, and distribute the rest of the bison output. (Bison generates a bunch of tables and some code, and appends the actual parser to the generated stuff. The actual parser is what's copyright. The tables and such are not.) > > (2) What about gcc? My understanding is that if you port > > gcc to a new machine, it must be freely distributed. > > Agreed? No, you can port gcc to a machine and never give it to anybody. But if you do give it away or sell it, you must distribute source (or make it accessible at no higher cost than binaries, for 3 years) and must not restrict redistribution of the sources or binaries. > > (3) Now, what about programs compiled with gcc? (I.e. my > > own code, compiled to binary using gcc. Must the > > binaries be freely distributed? What about the original > > source? Since the output of gcc contains no copyrights (unlike the output of bison, which contains a substantial piece of copyrighted code), you can do whatever you want with it. The GNU project would be happy if you *would* freely distribute your code, but there is no compulsion to do so. > (Would anyone from Free Software > Foundation, Inc. care to respond?) I'm not from the Free Software Foundation but have worked with them for a while, and have seen this question answered before. -- {pyramid,pacbell,amdahl,sun,ihnp4}!hoptoad!gnu gnu@toad.com "Don't fuck with the name space!" -- Hugh Daniel
rbbb@acornrc.UUCP (David Chase) (04/06/88)
>This is true, since bison's output includes a literal copy of a parser >(bison.simple or bison.hairy) that contains the GNU copyright. Last I checked, bison.hairy contained no copyright. I don't know if this omission was intentional or not, but it was that way on a tape shipped to me in late 1987 (the date on the file is 16 Feb 1987). David Chase Olivetti Research Center, Menlo Park