Path: utzoo!attcan!uunet!tut.cis.ohio-state.edu!zaphod.mps.ohio-state.edu!sdd.hp.com!elroy.jpl.nasa.gov!decwrl!shlump.nac.dec.com!tkou02.enet.dec.com!diamond From: diamond@tkou02.enet.dec.com (diamond@tkovoa) Newsgroups: comp.sources.games.bugs Subject: Re: problem running dungeon Message-ID: <1763@tkou02.enet.dec.com> Date: 4 Jun 90 02:08:01 GMT References: Reply-To: diamond@tkou02.enet.dec.com (diamond@tkovoa) Distribution: comp Organization: Digital Equipment Corporation Japan , Tokyo Lines: 32 In article interran@lurch.Stanford.EDU (John Interrante) writes: >lurch/interran 8 % /usr/games/dungeon >Welcome to Dungeon. This version created 17-SEP-87. >eof/uio: Error 0 >apparent state: unit 2 named /usr/games/lib/dunlib/dtext.dat >last format: >lately reading direct unformatted external IO I saw the same error message on a competitor's machine before I was hired by DEC. (If I thought this were DEC's bug, I probably wouldn't be allowed to comment on it.) Anyway, I think I remember a combination of three reasons: 1. The record number is negative (zero might have the same effect). 2. The company that wrote the Fortran I/O library forgot to check for this possible error, so they can't figure out what went wrong. (Thus error 0 and no explanation.) 3. The dungeon program and data files (particularly dindx.dat in this case) were originally designed for integer*2 data. When patched and/or compiled differently with integer*4 data, the code must be more careful in operations on the sign bit, or on what used to be the sign bit, so that other bits will be stripped out as necessary. Someone forgot a few of these cases. Sorry, I don't remember how I fixed it in an older version of dungeon. -- Norman Diamond, Nihon DEC diamond@tkou02.enet.dec.com Proposed group comp.networks.load-reduction: send your "yes" vote to /dev/null.