Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!linus!philabs!ron1!polyof!polyob!julerio1
From: julerio1@polyob.UUCP (A1 jose ulerio1 )
Newsgroups: net.micro.apple
Subject: Re: Re:  ProDOS bug??
Message-ID: <184@polyob.UUCP>
Date: Mon, 30-Jun-86 02:20:36 EDT
Article-I.D.: polyob.184
Posted: Mon Jun 30 02:20:36 1986
Date-Received: Fri, 4-Jul-86 00:30:51 EDT
References: <1402@pucc-j> <1749@ecsvax.UUCP>
Organization: Polytechnic Inst. of NY, Brooklyn NY
Lines: 30

> > 
> > In article <3749@decwrl.DEC.COM> you write:
> > >>I have had a problem when running in ProDOS for quite awhile...
> > >
> > >>I run a BBS...
> > >
> > >>Every now and then, when the program would CHAIN to another section, the
> > >>disk drive would just lock up!  
> > >
> > >>If it makes any difference, I am running on an enhanced //e, 128k 80col card,
> > >>two floppies, and a Sider 10meg drive (it's for sale by the way).
> > > 
> > >>		Tom Arnold -- Purdue University
> > 
	If you're running under PRODOS v1.1.1, there has been a definite and
publicly documented bug discovered in BASIC.SYSTEM.  Luckily, a solution was
made shortly after the bug was discovered.  If you're technically inclined,
the complete explanation of the bug and its fix appeared in the June and Nov.
1985 issues of C.A.L.L. Apple, respectively.  All you need to fix CHAIN
is one poke before you go on to do anything else.. this poke comes from
the April '86 issue of Open-Apple.

1 IF PEEK(49149) = 1 THEN POKE 41859,3 : REM check for V1.1.1, then fix CHAIN

	The problem arises when the length of the variable tables becomes
an exact multiple of 256.  When that length is achieved, BASIC.SYSTEM tends
to then go on and forget where the variable locs. were to begin with.
This may or may not account for your drives locking up, though I'd assume
if you were doing a disk command that involved a variable (as in the
case of reading & updating user messages) it'd blow up just as well.