> In those cases where the mov.x is executed it backs the bombing run up
> one position. =A0A cheap way of getting little more out of the attack,
> and it does score a few more points.
Ah ok, thats the reason. Nice little trick to improve on that bombing
engine. I'm not used to that engine yet :) So I could even try
-2/-3/-4 to bomb a little more, intresting to see if the results get
better.
I'm only encoutering one problem with the bombing engine, imagine the
following warrior:
for 97
nop #-99,-99
rof
nop 2,1
nop #-99,-99
start jmp #99,#99
end start
Your bombing engine will only kill this in 2% of the finds. Because
when it encouters a nop 2,1 it will decrease and bomb it to 1,1, the
following line has been bombed so the engine stops! This shows two
problems, every 2 a-field it encounters will stop the engine + the
indirect bombing bombs the wrong line, it points to start, but its not
hit.
I'm curious why you those to use {qStep instead of *qStep (better
scores for some reason?). Fixing it will cost one line (as far as I
can see) but might improve the scores drasticly:
djn.f 1 ,*1 ;making a-field not point to dat 0,0
qPtr mov 0 ,*qStep ;not using { but *, pointing to
original location
It scored somewhat better in my tests at least :-)
Roy
|
