Msg: 5586 *Conference*

12-15-94 17:45:10

From: RON WIESEN

To : ALL

Subj: TINY.CO RELOCATION INQUIRY

Does anyone know of any "intrinsic" reason why TINY.CO can't be RELOCated from
it's design TOP: of 62200?
 
Here's what I do know about TINY.CO:
. History: (C) 1986 from ACROATIX (with POWR-DOS);
.          circa 1988 public domain;
.          circa 1992 TINYLD.BA from Tracy Allen;
.          Club 100 library.
.     TOP: 62200
.     END: 62959
.     EXE: 62200
. DATATOP: 62346
. DATAEND: 62403 If RELOCated by any non overlapping interval of 200, and
keeping it's original copy at 62200, the relocated copy works but it obviously
passes control to the original copy at some point. Relocation by a non
overlapping interval of 100 does not work, so if 200 is a multiple of a smaller
interval it is not 50 (could be 40). Not keeping the original and relocating it
by a any interval of 200 does not work.  To relocate TINY.CO (length=760) from
62200 to 62000 for example, I first relocate it to 61200 and relocate that to
62000: both relocations avoid interim overlap.
 
Assuming the "intrinsic" problem can be solved and no interval is required, my
ultimate target TOP: is 62119 (81-byte Op Sys butted against MAXRAM) and this
will require a 2-step relocation.  Of course each time I use RELOC.BA on
TINY.CO (or any .CO) I take care not to let the original and relocated areas
overlap.
 
Assuming the "intrinsic" solution requires an interval, I'll live with a pair
of 40 or one 80 (I can shave 1 byte off the Op Sys) but not with 200 (119
wasted bytes).
 
Any info is appreciated.  Without any, I'm faced with long hours of code
inspection.