Path: utzoo!utgpu!news-server.csri.toronto.edu!clyde.concordia.ca!uunet!samsung!zaphod.mps.ohio-state.edu!tut.cis.ohio-state.edu!ucbvax!SYSE.SALFORD.AC.UK!M.A.Holloway
From: M.A.Holloway@SYSE.SALFORD.AC.UK
Newsgroups: comp.sys.atari.st
Subject: (none)
Message-ID: <.2.May.90.11:24:36.A110E1@UK.AC.SALF.E>
Date: 2 May 90 10:24:36 GMT
Sender: daemon@ucbvax.BERKELEY.EDU
Organization: The Internet
Lines: 32


      Over Easter I attempted to write a mouse pointer animator.  I came
across two problems which I hope someone can help me on.  Firstly, if I took
the mouse pointer up to the menu bar and started moving it about frantically
from one menu to the next, weird bit patterns replaced some menu entries.
Secondly, when the text window scrolled it left "mouse droppings", ie the
mouse sprite was scrolled up with the text and thus not wiped out by UNDRAW
SPRITE.
At first I thought (2) was a trivial problem, accounted for by my own short-
sightedness (it still may be so, please tell me!).  My program was wiping out
the old frame and drawing the new one without knowing if the application was
scrolling or whatever.  Now the mouse handler in the VBL queue tests two
conditions before it redraws the mouse:- the byte at $2847 (in my system)
and bit 0 of $2846.  As far as I can tell the bit flag is a "has the mouse
moved since the routine last checked it?"  and $2847 is a "don't do a redraw
as the o/s is transforming the mouse sprite etc".  Is this so?  Do
applications set the byte @ $2847 (or L-A variable base - 339) to non-zero
if they want to disable the mouse?
      To eliminate the problem I tried testing $2847 and only redrawing if
it was zero, but this failed, why?
      On the subject of documentation of the system:Dissassembling the ROM
might well give you a little knowledge, and this CAN be "dangerous", as one
can find out half of what a system variable is and assume that s/he knows the
whole story of it's uses=> incompatibility with some programs.  Atari could
prevent this by providing the official documentation reasonably priced.
Atari would be the overall winner:-more money (I'm sure Abacus have made money
on their incomplete and sometimes inaccurate BUT AVAILABLE system books) and
more programmers who can find out cheaply how to program "legally"
(the little guys who write for PD/shareware are just as
important as the software houses which can afford the developers package).

                    Marc H.