Date   : Mon, 25 Jul 1994 08:37:30 EST
From   : Stephen Quan <quan@...>
Subject: Re: Anyone out there?

> Is it just me, a bug or what? Is there no traffic all of a sudden?

Hi Chris, saw your message but was wondering the same thing.
I haven't seen anything go by for 3 days, but then again who logs
on during the weekend?

My BBC has been unpacked for several days now, and I have been playing
ELITE and blasting at the constrictor!  I did, of course make use of
the hyperspace cheat a huge amount!  (I think near TIANVE and ANLE)
if you set your hyperspace to TIANVE but during the count down verify
the Data on System it will switch destination to ANLE and change the
distance.  So I have been hyperspacing to ANLE with no cost in fuel and
made a huge profit trading in an item that varies from 4.4 credits to
100 credits.  Converted the JAMESON ship to fully equipped within a
few hours.  There is a shortcut to TIANVE, you can do it in about 5-6
jumps by going down towards LEEGEES (bottom left of the map) and then
hyperspace to near the top right of the map where TIANVE is 3.6 light
years away.  If you make use of the Data on System bug it will change
the course to ANLE (8.8 light years away) and clock your fuel to 23.2.
(or somesuch)!

Has anyone thought about how you will go about timing the delays in
the vertical sync?  I am a little rustly but I think one good way is
using VDU calls in assembly.

.map_0_to_black LDA #...
                JSR &FFEE
                ...

.map_0_to_white LDA #...
                JSR &FFEE
                ...

.delay          LDX #0
                LDY #0
.loop           NOP
                DEY
                BNE loop
                DEX
                BNE loop
                RTS

.main           ... equivalent of *FX19 (wait for vertical sync)
                jsr map_0_to_white
                jsr delay
                jsr map_0_to_black
                jsr delay
                jsr map_0_to_white
                jsr delay
                jsr map_0_to_black
                jmp main

This program should be run in every graphics mode (MODE 0..6) and the idea
is you should see the background converted in black and white bars.  If you
want a rainbow you could write more map_0_to_red (whatever) and call them
into the main loop.

The actual number of instructions executed and the function calling
overheads is unpredicatable, but it doesn't matter.  Run the program
and observe how thick the black/white (or colour bars) are.  Then
insert another NOP in the delay loop (an additional 65536 NOP's) and
observe the thickness of the bars again.  The difference in thickness
is represented by 65536 NOP instructions.

We should be able to count how many additional scan lines that have
occured. (eg. might get a result like 10 scanlines / 65536 NOP's,
ie. 1 scanline takes 6553.6 * 2 cycles = 13107.2 cycles).

With mode 7 I think we can assume the refresh is near instantaneous.

Comments anyone?
-- 
Stephen Quan (quan@...                 ), SysAdmin, Analyst/Programmer.
Centre for Spatial Information Studies, University of Tasmania, Hobart.
GPO BOX 252C, Australia, 7001.  Local Tel: (002) 202898 Fax: (002) 240282
International Callers use +6102 instead of (002).