Path: utzoo!attcan!uunet!zephyr.ens.tek.com!tekcrl!tekgvs!toma
From: toma@tekgvs.LABS.TEK.COM (Tom Almy)
Newsgroups: comp.sys.ibm.pc
Subject: Re: NNANSI doesn't work with Codeview 2.2
Message-ID: <6006@tekgvs.LABS.TEK.COM>
Date: 27 Sep 89 14:20:05 GMT
References: <768@kl-cs.UUCP>
Reply-To: toma@tekgvs.LABS.TEK.COM (Tom Almy)
Organization: Tektronix, Inc., Beaverton,  OR.
Lines: 29

In article <768@kl-cs.UUCP> jonathan@cs.keele.ac.uk (Jonathan Knight) writes:
>Hi,
>  Has anyone noticed that CodeView doesn't work with NNANSI?  It
>seems that CodeView plays with multiple screen pages, NNANSI doesn't
>seem to cooperate.  Anyone got a fix or workaround?

The problem is that NNANSI uses the "multiple screen pages" as one large
page which it scrolls -- thus its fast performance. You can compile NNANSI
with FAST=0. Then CodeView (and some other programs) will work, but you loose
most of the performance advantages.

I'm working on improvements to NNANSI, and have found some strange problems
(that were even in NANSI) -- adding declarations at the beginning of the code
causes it to crash (???) and the system crashes when using 386/MAX, attempting
to load the driver in high memory.

Improvments:

1. Option to compile without keyboard remapping (this ansi.sys feature is
   apparently used by some Trojan Horse programs).

2. Improved performance in graphic mode

3. Hook to allow proper operation with Paradise VGA cards, and others that
   set the display mode byte "wrong" when in high resolution character modes.

Tom Almy
toma@tekgvs.labs.tek.com
Standard Disclaimers Apply