Msg: 5492 *Conference*

10-20-94 17:52:56

From: RON WIESEN

To : COMET _

Subj: REPLY TO MSG #5469 (RELOCATING M/L)

You're in luck my friend!  There are two ways to go:
 
1.  Use RELOC1.BA from the DOWNLOAD area.
    Linage: RELOC.BA-James Yi-(Orig ver), RELOC.BA in Library-Robert
Benson-(Ver 1.2), RELOC1.BA in Downloads-yours truly-(Ver 1.3).
    Ver 1.3 accounts for all op-codes except CB hex.  Maybe you can account for
it and we'll have Ver 1.4!
    Use the documentation of RELOC.BA from the Library.
    I like it's DATATOP/DATAEND feature which is fail-proof when you have all
data (non-code) in one chunk and you know the Top/End addresses of that chunk.
As a rule, I keep all data in one chunk at the end of my M/L programs.
 
2.  Read my thread re: running M/L programs in place [REPLY TO MSG# 5363
(STUFF)].
    I thought I put a follow up in the conference before, but I can't find it.
    So I'll put a message addressed to ALL here today, titled SIMPLE .CE.  I
sent this privately to Rick in April under the same title, so ask him for more
details.
    In practice, the Executive program adjusts/runs/re-adjusts the .CE program
all the while the .CE remains in-place.
    It's like having .EXE programs in the PC-DOS world!  Where the Model-T Op
Sys only accomodates absolute .CO programs (like .COM in PC-DOS), my Executive
allows relative .CE programs (like .EXE in PC-DOS).
    So HIMEM requirements disappear because the Model-T Op Sys doesn't have to
make a copy like it does for .CO programs.  Less memory, less hassle.
 
If you opt for the second way, contact me if you want the Assembly source for
my Executive.  Yes, the Executive is a .CO that works from the "Prev Screen
Buffer" so HIMEM matters not!