Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP Path: utzoo!watmath!clyde!burl!ulysses!bellcore!petrus!scherzo!allegra!mit-eddie!think!harvard!seismo!rochester!bullwinkle!batcomputer!garry From: garry@batcomputer.TN.CORNELL.EDU (Garry Wiegand) Newsgroups: net.unix,net.lang,net.legal Subject: lex and yacc in the public domain (responses) Message-ID: <481@batcomputer.TN.CORNELL.EDU> Date: Thu, 3-Apr-86 18:42:25 EST Article-I.D.: batcompu.481 Posted: Thu Apr 3 18:42:25 1986 Date-Received: Tue, 8-Apr-86 06:15:25 EST Reply-To: garry%cadif-oak@cu-arpa.cornell.edu.arpa Organization: Cornell Engineering && Flying Moose Graphics Lines: 125 Xref: watmath net.unix:7559 net.lang:2313 net.legal:3223 A couple weeks ago I asked whether lex and yacc, as distributed by Berkeley, were protected (they had no explicit copyright notices). The interesting discussion I received went as follows (my summary is at the bottom): *********************************************************************** Copyright law requires a copyright notice. It seems these files don't have one, so they can't be protected. Software may also be protected as "proprietary information" or as a "trade secret". Whether or not this software is protected this way I think depends on how and where you got the sources. Certainly if you signed a non-disclosure agreement before you got them then you cannot redistribute them. If, however, you found them on your local university computer system, and they were not protected, and nobody ever told you not to look/copy/redistribute them, then it seems to me that they are fair game for redistribution. The law protecting trade secrets and proprietary information requires the owner to take strong measures to protect this information. Once it slips out into public knowledge, it is no longer protected. To my mind, lots of UNIX code has slipped out, since it is readable by everyone on lots of university and business computer systems. Probably the universities signed non-disclosure agreements when they got the source, but AT&T has been a bit negligent in enforcement. -- Jim Galbiati, GenRad Inc, Production Test Division Mail Stop 6, 300 Baker Ave, Concord, Mass. 01742 {decvax,linus,wjh12,mit-eddie,cbosgd,masscomp}!genrad!panda!jwg (617) 369-4400 x2459 ************************************************************************** Absolutely not. Both are licenced under your Unix source paper from ATT, and both are trade secrets, even though not copy right. ***dan {allegra,astrovax,princeton}!fisher!djl ************************************************************************** You'd better read copyright law. You don't have to say "copyright" anywhere in order to get protection. But if you don't, you can't sue for more than the profit that was made. Yacc and lex are still covered by copyright, even if they don't say so. Check with a lawyer. Rick Richardson, PC Research, Inc. (201) 922-1134 ..!ihnp4!houxm!castor!{rer,pcrat!rer} <--Replies to here, not to homxb!!! ************************************************************************** lex and yacc are protected AT&T sources under the AT&T trade secret agreement. They cannot be copied willy-nilly unless you want to risk a major suit filed against you which would be impossible for you to win. Tony Hansen ihnp4!pegasus!hansen ************************************************************************** We've been a UNIX shop over 10 years, so my information may be out of date, but I know that both of the following were at one time true: The source for yacc and lex that is distributed with 4BSD, Sys V, V7 (pick your flavor), is covered by the AT&T software agreement. It is my understanding that there are public domain versions of both, but I don't remember when or where I got that understanding (there may be DECUS versions). At one time AT&T claimed that source code _generated_ by yacc and lex was also covered (and perhaps the binaries as well - I don't remember). I forget the exact line of reasoning, but while it was "legalistic" it was not completely unreasonable. Before you redistribute the sources for yacc and lex that you originally got from AT&T (or from an AT&T licensee, which includes all UNIX-like systems I've ever heard of) you would be wise to take a copy of your source license agreement to _very_ competent legal counsel. Personally, I would be equally careful of distributing source code that contained yacc and/or lex output, and probably just as careful about binaries created from yacc/lex output. It was to avoid exactly this problem, and to keep the UC lawyers happy, that I spent about two months some years ago hand coding a rather nasty database data-definition language parser. (Of course the other benefit of hand coding it was that it ran a lot faster than yacc/lex generated code would have and I didn't have to worry about any of the bugs in either of those two.) John Pierce, Chemistry, UC San Diego jwp%chem@ucsd.edu sdcsvax!sdchem!jwp ************************************************************************** SO: 1) it can't be copyrighted because it has no explicit notice, 2) it may be copyrighted anyhow, 3) it is protected by an AT&T trade secret agreement, 4) it is not a trade secret, since no effort was made to maintain its secrecy. I trust I've gotten the gist of it right? :-) (Apologies in advance!) The bottom line for me is that my access is courtesy of Cornell, and it sounds like Cornell must have signed a trade secret agreement. If it were just me I might take the chance, but I surely don't want to cause waves for Cornell - AT&T is a benefactor! So I'm going to hunt for true public domain copies. (net.lang people: help?) (Please choose one group to send followups to, depending on what you have to say.) thanks to all - garry wiegand garry%cadif-oak@cu-arpa.cs.cornell.edu