Date   : Wed, 24 Aug 1994 11:17:45 +0000
From   : lamcw <lamcw@...>
Subject: replies...

hi everyone,

first off, i'd like to show off like a smug bastard and say that i got
arcadians to run last night, it's a truly great historic moment...(pauses
to wipe away a tear)

it's probably about 50% speed, animating around 10 frames a sec in mode 5,
but it's still very playable!

anyway, i'll do my agony aunt bit now...


"Chris L. Rae" <clr1@...> wrote:

;Also, I have discovered a bug in ADC (if a rather small one) which could 
;save me because it happens quite a lot when BASIC is displaying things. 
;When there is overflow, ADC ignored the carry! Very odd. So with C=1, 
;adding &F0 to &20 gives &10 instead of &11. I *think* this is wrong, am I 
;right? I'm going to fix it so it doesn't do it anymore, so I'll soon see 
;if it's the key...

looks like it should be &11. the BVS and BVC opcodes are probably very
infrequently used but unless implememented properly the dreaded 99% 
syndrome will exist.

;1. Errors seem to work okay (except ESCape, of course) and except the 
;"Silly" one, I know not why. However, there is a spurious character 
;before each error message. I'm not exactly sure why.

the error message should start exactly on the LHS side of the screen. if
you are getting an extra char then it could be you are pushing the
wrong address when the BRK opcode is executed. BASIC uses this address
to determine the start of the text message. so try subtracting 1 from it.
(i had this problem initially as well)

;2. (This one is *wierd*) When I type "List" a few times my emulator halts 
;with a duff instruction opcode!

could be any reason... SBC perhaps?

*************

Stephen Quan <quan@...> wrote:

;> my philosophy for handling the paged io is to do all the emulation when
;> values are STORED to sheila. therefore any reads from sheila will be 
;> correct and one does not need to 'intercept' the read.

;I did think of this, but then thought it wouldn't work.  Isn't there a
;case when the 6502 continuously polls FE4F (System VIA ORA-nh) for a key?
;Um, even if it did, I guess it is easy to get around that by writing
;directly to FE4F whenever characters come in (and the latch = 3).

my idea for the paged io seems to have been met with a general consensus
of disapproval. however it all seems to be "holding" together for the time
being


chris lam,
aston univ,
u.k.