Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Path: utzoo!mnetor!seismo!rutgers!sri-spam!sri-unix!hplabs!tektronix!tekcrl!vice!tekfdi!videovax!stever
From: stever@videovax.UUCP (Steven E. Rice, P.E.)
Newsgroups: net.micro.amiga
Subject: Re: Updated Diffs for vt100 version
Message-ID: <2036@videovax.UUCP>
Date: Sun, 2-Nov-86 18:32:37 EST
Article-I.D.: videovax.2036
Posted: Sun Nov  2 18:32:37 1986
Date-Received: Tue, 4-Nov-86 08:11:19 EST
References: <1502@jade.BERKELEY.EDU> <6132@decwrl.DEC.COM>
Reply-To: stever@videovax.UUCP (Steven E. Rice, P.E.)
Organization: Tektronix, Comm Group, TV R&D
Lines: 58

In article <1502@jade.BERKELEY.EDU>, Mike Meyer (mwm@eris.BERKELEY.EDU)
writes:

> In article <2009@videovax.UUCP> stever@videovax.UUCP (Steven E. Rice, P.E.)
> writes:
>
>> Lattice C is unable to cope with the construct [n+1].
> 
>> ***************
>> *** 354,360
>> !    FileItem[n].NextItem = &FileItem[n+1];
>> --- 354,361 -----
>> !    nplus1 = n+1;
>> !    FileItem[n].NextItem = &FileItem[nplus1];
> 
> This looks like utter bilge to me. I'm currently running v2.2 WITHOUT
> this patch (though it has the patches posted by DBW), and the menus
> look exactly like the code makes me expect them to. Ditto for all the
> things I've written that use this technic for setting up menus, etc.
> 
> I'm using 3.03. If 3.04 Beta has this problem, then I pray that 1) it
> gets fixed before the release, and 2) DBW won't pick up that
> particular piece of uglyness just so the Beta software will work.

I appreciate Dawn Banks (banks@viking.dec.com) writing to clarify the
origin of the [n+1] fix for the vt100 emulator.  I was about to dig out
my hard copy of her original article and type in the relevent portions.
In article <6132@decwrl.DEC.COM>, Dawn writes:

> On the subject of the [n+1] occurances in DBW's VT-100 emulator, and how
> well they sit with Lattice:

> . . .

> With Lattice 3.03, and VT-100 version 1, the occurances of [n+1] weren't
> being compiled properly.  Didn't come as any surprise, because this
> "feature" was mentioned on the network soon after Lattice 3.03 hit the
> streets.  Changing the references to "m=n+1; foo[m]= ..." did, of course,
> prevent the Guru visitations.
> 
> Since then, starting with VT-100 version 2.0, we haven't seen this problem
> reoccur, so I can only assume that it was some context sensitive bug in 
> Lattice.
> 
> I can say with all certainty that it was happening with Lattice 3.03 (and
> not 3.02), and with the original posting of the VT-100 emulator.

> . . .

Since the problem is with the current version of Lattice (V3.03), it seems
wise to avoid it -- even if it is not presently manifesting itself.  Surely,
despite the "uglyness" of the fix, it will add little to the size of the
executable.

					Steve Rice

----------------------------------------------------------------------------
{decvax | hplabs | ihnp4 | uw-beaver}!tektronix!videovax!stever