Path: utzoo!utgpu!news-server.csri.toronto.edu!cs.utexas.edu!sdd.hp.com!hp-pcd!hpcvra.cv.hp.com!rnews!hpcvbbs!akcs.falco
From: akcs.falco@hpcvbbs.UUCP (Andrey Dolgachev)
Newsgroups: comp.sys.handhelds
Subject: Re: HP48: new ML tetris in ASC format
Message-ID: <280b63a2:2787.2comp.sys.handhelds;1@hpcvbbs.UUCP>
Date: 16 Apr 91 21:40:06 GMT
References: <1991Apr15.061548.12826@colorado.edu> <280a9907:2787.1comp.sys.hand
Lines: 31

Well, I thought about taking the ML Tetris apart to see what made it tick
(and take a look at the routines).  So, I started by making the library
into a directory.  What I found was a USERLANG program, a configuration
program (17000 ATTACH), a string, a GROB, and a machine code routine. 
Now, maybe some of you have cautght on to the imortatnt part?  Well, I'll
giv you another hint, remember how the ON key does not function wile
playing Tetris?  Not yet, wel, if you look at the Start program, the
USERLANG program that is, you find that it recalls the string to the
stack, and then executes the machine code routine.  And, finally, if you
haven't caught on yet, the string can't be visited or edited because you
"can't edit null character".  Unfortuantely, I have the old version of
CHIP48, not 2.25, so I wasn't able to make sure if ML Tetris reallly is a
CHIP program or not, but it sure does seem like it, don't it?    I would
appreciate it if somebody takes a llok and sees if the string can be run
with their CHIP or if one of ttheir chip games can be run with the code
segment in ML Tetirs.  The library can be recalled as a directory by
using the RCLLIB program in HACKIT, posted by a kind soul a few days ago.
  Also, it looks like I'm going to have to delve into machine code to
speed up my Tetris.  BTW, it looks like the main speed increase is not
the speed of the actual routines, like the rotating or moving routines ,
but the speedup comes about from the way that key-presses arehandled. 
THe partial ML Tetris uses KEY, which doesn't care if you you hold down
the the ke.  Becuase of this,  there's a noticeable lag and you have to
repeatedly press the keys.  Anyways, I tried making the rotate function
do contnuous and it was very fast (you could barely see the piece when it
rotated) so it look slike I'm going to have to modify my key-handling
routine with ML.  Unless a kind soul has info on pokeing the key buffer
and directly polling the status of keys from ML or RPL, it looks like I'm
going to be spending some time manuallly dissambling the KEY routine. 
However, it's good that there is a faster Tetirs out (which means better)
, after all, competition always improves the product.
        ---Falco