XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX  CTAB  XXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXXX
 
       Change Table -  a tool for programming with JAZZ for the HP49 

CTAB is a nice tool for those who like programming with JAZZ or JAZZlight and 
which are in the possession of a HP48GX and a HP49. As long as only supported 
pointers are used it makes not difference whether your program is destinated 
for the HP49 or the HP48. These will be written on the HP48 with JAZZ[light] 
and CTAB must be downloaded to the HP48. 

CTAB stores hp49tabs.lib in 'L993' and toggles it with hptabs.lib in Port 1.
Write your program with JAZZ as long as L993 is not in Port 0. Unsupported
HP49-pointers may also be written now, but HP48 ROM Pointers must either be 
replaced by their content or their equivalents must be hacked on the HP49. 
If CTAB is pressed, L993 is attached to Port 0 and hptabs48.lib in Port 1 is 
detached, or conversely. CTAB switches the tables at any time in about 1 sec. 
To port a SysRPL-program to the HP49 preserve the source string and compile it 
after switching to the hp49-table. Then tranfer the result to the HP49. Thus, 
porting a program from HP48 to the HP49 containing supported pointers only 
- these include all UsrRPL-commands -  is a child's play with CTAB. 

The targed code can often be shortened by using Flash Pointers, e.g., 3LAMBIND 
(which binds 3 stack objects to 3 NULLLAMs). The easiest way to do this is to 
use the ID 3LAMBIND in the source string and replace it after porting by the
corresponding address, e.g. '3LAMBIND' by  FPTR  6 156. These addresses can 
be found in XREFAZ.TXT by J.K.Horn. This replacement must often be done on 
the stack, not with HP49's Text editor because JAZZ has a different style of 
dissassembling and assembling. ASM2 from Library 257 is helpful so see what
is going wrong in reassembling a JAZZ-made HP49 program (Version 1.16).

A certain problem is the HP48-HP49 transfer. A nicely compiled file sent from 
the HP48 arrives as a "HPHP48..." on the HP49. Put its name on the stack and 
run <<256 ATTACH RCL ->A  26 +  A-> >>. In almost all cases the result is o.k. 
----------------------------------------------------------------------------

Wolfgang Rautenberg                    email: raut@math.fu-berlin.de  
http: www.math.fu-berlin.de/~raut      ftp.math.fu-berlin.de/pub/usr/raut