Return-Path: Received: from SEGATE.SUNET.SE by xiron.pc.helsinki.fi with smtp (Linux Smail3.1.28.1 #1) id m0t8Xc6-0000ZQC; Thu, 26 Oct 95 21:05 EET Message-Id: Received: from listmail.sunet.se by SEGATE.SUNET.SE (LSMTP for OpenVMS v1.0a) with SMTP id D54BF0C2 ; Thu, 26 Oct 1995 20:05:17 +0100 Date: Thu, 26 Oct 1995 12:22:01 -0400 Reply-To: John Cowan Sender: Lojban list From: John Cowan Subject: Re: Grammar.235 X-To: Lojban List To: Veijo Vilva In-Reply-To: <199510260747.DAA20509@locke.ccil.org> from "Logical Language Group" at Oct 26, 95 03:42:41 am Content-Length: 1011 Lines: 26 la lojbab. cusku di'e > John Cowan appears not to have answered yet on this, so let me pipe in. > > The standard for Lojban grammars is to tolerate NO shift/reduce errors. > > As far as I know, our testing of grammar.235 produced no s/r errors > (and it appears from files John left around today that he tried again > and got the same result). And apparently my report never left this machine, for reasons unknown. I tested grammar.235 against several different yaccs: SunOS yacc (basically AT&T), SCO Unix yacc (AT&T), Berkeley yacc (AT&T-free), GNU Bison 1.24, and Abraxas PCYACC (a commercial MS-DOS clone). All created conflict-free parsers with no problems. > This implies that someone's version of YACC is buggy. Right. I'm not sure what the ancestry of the yacc used by {la paulos.} is, but it may have trouble handling such a large and complicated input, with its 1077 lookahead sets. -- John Cowan cowan@ccil.org e'osai ko sarji la lojban.