YI-RD1.THD  --- Copyright 1987 by Phil Wheeler

An original compilation of Compuserve Model 100 Forum messages for use by
Forum members only.

James Yi is a prolific M/L programmer who has uploaded some exciting new
code in recent months.  Alas the source code is written in his own assembly
language (for his own assembler).  This set of messages relates to
translating his nemonics into more standard Intel-type code.

Message range: 159383 to 159847
Dates: 10/29/87 to 11/7/87


Fm: Phil Wheeler 71266,125
To: James Yi 73327,1653

I'm looking into converting your RESEQ.200 for 100 use.  But it will have to
run in user RAM, since the M100 alt LCD buffer is not big enough to hold it!

Some questions:
 1) I assume the high RAM addresses at 65500+ are being used as a buffer, and
that I can simply use addresses in the program area itself for the M100
version.  Can't find any equivalent space in M100 system RAM (tho it may exist
).
 2) Biggest problem is that I have trouble reading your SRC code! Example: In
one place you have "hl=[de".  What is that supposed to do?  There are other
things I'm not sure of, too.  Best bet will be to get a source and let you read
it to see if it's close!
 3) Another reading problem: e=#5.  Is this MVI  E,5 ? (i.e., does #5 mean a
decimal number as it seems?).  But then what does cmp#',' mean -- CPI  ',' ?
 4) Finally (for now!), what does the "token list" at the end do? "/[136 140
."  doesn't tell me much.


Fm: James Yi 73327,1653
To: Phil Wheeler 71266,125

65500+(65362+ on T200) addr, I think are available on M100 and it's the Telcom
buffer, safely assuming, since the chrs viewed online can be seen there off
line.

hl=[de, is the undocumented op-code.  it loads hl with content of addr pointed
by de.
Ones I use frequently are:

-bc    ;8d 'Subtract BC from HL
srh    ;16d 'Shift Right HL w/o carry
sld    ;24d 'Shift Left DE
[de=hl ;217d 'store HL at addr of DE
hl=[de ;237d 'load HL from addr of DE

I use decimal, and e=#5 means MVI E, 5d
And cmp ',' or cmp #',' means
 CPI ',' or CPI 34d

"/[136 140...  is a list of numbers, the bracket indicating a string of data,
either numerical or ascii.

The numbers are Basic token codes, and it's the same on 100.
With CSA, equiv would be something like:
    DB 136d
    DB 140d
    ...
    DB 0d

And /['character' 0
means
    DM  character
    DB  $00


Hmm..  maybe I should learn to use Custom Software Assembler..
Where can I obtain it for T200?

Are the lcd buffers for both current and prev screen next to each other?
If so, you have 320+320=640 bytes!


Fm: Phil Wheeler 71266,125
To: James Yi 73327,1653

You're right!  There is a chance to stuff it in there. I'll look at the other
buffer you mention.  If the whole thing can go into high RAM, it will be a big
plus!

Another question: I've always assumed that the '/.4' (for example) is a unique
subroutine label.  But now I find several '/.1' labels (under .6,
7, .10 and .13), each seemingly an internal loop.  Maybe I understand less than
I thought!  What IS the format?


Fm: Phil Wheeler 71266,125
To: James Yi 73327,1653

Then, from reading last message, what is 'a=[de'?


Fm: James Yi 73327,1653
To: Phil Wheeler 71266,125

The compiler updates the value of a label every time it reappears.
So, if the ML code that refers to .1 is compiled before another .1 is
encountered, the code will be assigned with the value defined by the 1st .1

As an example,

A) /.1=1000
B) JMP .1
C) /.1=2000
D) JMP .1

at B, the code is equal to JMP 1000, and at D, JMP 2000.
So the same label can be used more than once, as long it is accessed before it
changes to another value.

The following will not work if you wanted to JMP to 3000.

A) JMP .1
B) /.1=3000
C) /.1=0

The compiler makes 2 passes, and on 1st pass, .1 at A is not yet defined, then
it's defined 3000 at B, and redefined at C, with 0.
Then on 2nd, code at A will be equal to JMP 0, not JMP 3000, since it's
redefined at C on the 1st pass.
If we put JMP .1 in between B and C, then it would work.

a=[de means LDAX D, and [de=a means STAX D.


Fm: Phil Wheeler 71266,125
To: James Yi 73327,1653

James, I assume that inc h means inr h and inc hl means inx h (the h means
increment the 8 bit register and the hl means increment the register pair).
Coorect?


Fm: James Yi 73327,1653
To: Phil Wheeler 71266,125

yes, that's correct..


Fm: Phil Wheeler 71266,125
To: James Yi 73327,1653

Some more questions.  First is the label line after '/.23', namely:

 /b1:/.127:/f1;call hook+1

What does the code at the end mean?  That is:

 /.221
 /j61308;use INPUT buffer
 /.224:/f1:/.225:/f1
 /.222:/f39:/.223
 /j.221

Could you give me this in a more conventional format?


Fm: James Yi 73327,1653
To: Phil Wheeler 71266,125

"/b1:/.127:/f1"-  is to set .127 the addr of 55d, of ORA #55d, in .23, by
backing up one(/b1), label it and back forward.
Because if entry is made at .23, carry is cleared by ORA, but if it's made at
.127, carry is set by STC, 55d.

The stuff at the end, /.221...  is equiv to:
jy224   equ     on/off status(61308)
jy225   equ     buffer data pointer(61309)
jy222   equ     buffer(61310)39 bytes
jy223   equ     buffer+39(61349)

or using ram at the end of program:

jy222   dm      0123456789012345678901234567890123456789        ;39 bytes
jy223   db 0
jy224   db 0
jy225   db 0