ZX Hard #00 31 августа 1998

Железо

Subtleties poll Kempston Rats - a glitch Kepston Mouse.

   Subtleties of polling Kempston rats;)


(C) VTS'98


                 mouse included - well,

                 not included - is bad.


                 (C) Vladimir Mayakovsky,

                 Demon, VTS.



                 WELCOME TO THE HELL,

                 DEAR USER without

                 KeMpStOnE-mouse.


                 (C) Demon (ZX-Element # 1),

                    VTS.


I want to tell you about a sad
fact - I have no kempstone-joystick'a.

-So what, and I do not kempstone-mouse,
 -I will answer (VTS) ...

Something like this started in an instructive article in the 
ZX-Elemement # 1 on how to interrogate Kempston joystick. 
Apparently, the mouse occur much more frequently and joysticks

should be for everyone, because because of
no mouse (on every wheelbarrow, where
k.joys. decrypted by # FD) magazine a little bit (to put it 
mildly) is buggy. 

Only, plzzz, do not take it for hitting! I just really wanted 
to read magazine (what, you like to go for

20km to the city to another to read it
half of 'passing' and the second with a push
on the joystick 'left' and 'right' [later
however, using a soldering iron ;-)]).

I am also sure that this bug has already caught
I am writing this nonsense to eliminate further confusion.
In fairness, I must say that
This sin is not just you.

So get down to business.

On machines with a 'firm' part-time decoding for # DF, rather 
than # 1F in the absence of mouse (the ports are in the low 
byte address # DF) instead of 'mouse' disk imaging will read 
the joystick. Well, if reading the origin of glitches is not 
happening (sorry, due to lack of entogo devaysa not familiar 
with his in-depth interviews, but the conversation not about 
that at all), then when reading keys the fun begins.

Namely: the mouse buttons are passive in 1, and
active at 0, y joystick is all the way around! Therefore when 
lie under the table joystick prog track defunct rat catching 
idiotic intent user to find the 'any key' by pianist

I mean, the press of a pressed :-)

The result is ever pressed
'Fire' and 'abolition' (good, middle button is not used for 
'lethal' ;-) functions). The effect, of course, entertaining, 
but I took the magazine and not DeMouy some nd.

Once I realized this, I
immediately understood the method of treatment: press
'Left' and 'right' to stick (good,
I have it not rychazhkovy;)) then
I dovesil pair of jumpers and revised
all buggy prog;). Nevehavshim explains the: left and right - 
Bits 0.1 joystick - Correspond to the left and right buttons

mouse (bottom - bit 2 - middle).
Just do not forget to remove these jumpers :-).

Zhelezyachny solution method - add a couple of
LL1 elements or diodes for decryption
by # 1F - not recommended, because such prog is not so much, 
and indeed, it is problem coders, not users.


Incidentally, this problem manifests itself
on KAY'e and Scorp'e. Therefore the majority of mouse pro - 
from Peter - Works correctly.


Well, now move on to dessert - the one
why you and read all of this pun, I mean 
_KAK_pravilno_izbavitsya_ _от_этого_принеприятнейшего_известия_.


Take a nd St. Petersburg myshininy
prog and run it (eg, ZX-Format).
Now pay attention to seemingly useless pribambas - test the 
joystick and mouse. Although long overdue to ask ourselves: 
where are the tests of memory, a turbo and brakes, pseudo-ROMs 
and shadow monitors ;-). We associate this fact with the 
presence of Kempston-test, for example, in the MMD. EVERYTHING!

Now you can cry "Eureka!"
If you do not Archimedes, read on.

In short, the need to test the presence of the mouse!

And if you click no, it disables the survey.

And this is not easy, but very simple.

Maybe two versions: a simple and wrong.
To start analyze incorrect.


         LD A, # FA; read

         IN A, (# DF); button.

         AND% 00000111; Masquerage them.

         INC A; # FF +1 = 0

         JP NZ, mouse_off; ...
mouse_on ...

Principle: if a mouse, its port of buttons will contain 1 to 
level 3 ... 7; after installing the bits to 1

(AND% 00000111) into the ACC. obtain # FF, and
after INC A - 0 and the flag set Z.
If the joystick, then bits 3 ... 7 reset!

Why wrong?

a). There is a mouse, unstable data bus and
from anywhere in bits nd 3 ... 7 pressure on at least one 0.
The result - the mouse is resting.

b). No mouse, Kempston (joystick) wrong (set bits 5 ... 7), user
puts pressure on all (or at least used 'fire' and
'Up') four-way controller (nazlo. ..:*))
The result - the presence of mouse (+ glitches ...).

In the second method, you should read three bytes of mouse and 
if they are all equal, then with probability 99. (9)% mouse No: 

a). read all 3 bytes of the joystick.
b). Read unstable bus.
c). Read stable bus.

The remainder (tiny) likelihood of misidentification:

a). Unstable bus had time to change
between readings.
b). Shustry user managed to change the status
joystick between the readings:-On.
c). user has managed to: *)) put myshu
so that x = y = keys !!!...

Well, that's AFM:


          LD BC, # FADF

          IN A, (C)

          INC B; B = # FA +1 = # FB

          IN D, (C)

          LD B, # FF

          IN E, (C)
; Read all 3 bytes.


          CP D

          JP NZ, mouse_on

          CP E

          JP NZ, mouse_on
mouse_off ...


Well that's all :-).

PS: Though not hurt to add a manual
correction of the test - both in the same ZF.

12.08.98 22:19 msk is-dos