Relay-Version: version B 2.10 5/3/83; site utzoo.UUCP
Posting-Version: version B 2.10.2 9/5/84; site bunker.UUCP
Path: utzoo!watmath!clyde!burl!ulysses!gamma!epsilon!zeta!sabre!petrus!bellcore!decvax!ittatc!bunker!jrh
From: jrh@bunker.UUCP (Jeff Hagen)
Newsgroups: net.micro.apple
Subject: Hackers delight - AppleII question
Message-ID: <976@bunker.UUCP>
Date: Wed, 4-Sep-85 16:51:53 EDT
Article-I.D.: bunker.976
Posted: Wed Sep  4 16:51:53 1985
Date-Received: Fri, 6-Sep-85 04:18:43 EDT
Organization: Bunker Ramo, Trumbull Ct
Lines: 33


Question, Apple II[+-], workings of 80 column cards.

Scenario:
You enter the monitor. Call -151. You type  3^P return with your 80
column card in slot 3. The monitor pokes locations $36 and $37 with
 $C300. It displays the prompt with a call through $36. $C300 gets 
executed. Among other things, the ROM code at $C300 selects the
80column video display. Your screen flips. It pokes location $36
with the usual entry point for output characters. (like 07) It
also pokes $38 and $39 for the input routine that will allow
lower case and other things. The monitor works like a champ in 80 
columns, but you have lost your hook to DOS. 3EAG fixes that.

Now here is the big if. If I have ROM selected in the upper 12K
and I do a 0^P or a PR#0 the screen gets flipped back to the 40
column mode.  If I have RAM selected it fails. The $36-$39 vectors
are set properly, but the 80 column screen is still showing.
Assume ROM and RAM have the same code in them.

Question:
Is the above scenario a hardware bug in my cheapie 80col card?
Or more generally, how does the 80 column card know when to flip
back to 40 column mode since a PR# or a 0^P only pokes the
 $36 through $39 bytes? Is it actually monitoring the bus?
Or is its input routine scanning the input?

I'm baffled.
Any ideas?
Thanks.

                     Jeff Hagen
                     decvax!ittvax!bunker!jrh