Bugs fixed and features added

Last revised: 5-May-2004

Release 0.8:

Added 8085 support
Fixed failure to complain about illegal addressing modes
Fixed problem with initializing processor-specifics for GUI
Accept binary (%) and octal (@) numbers for Motorola family processors
Explain binary and octal number syntax in the docs
Generate code for instructions that have bad labels
Generate bytes of 00 for post-allocation errors
Made GUI outputs the default when using the GUI.

Release 0.7:

Added 8008 support
Added 6502 support
Got rid of the ton of generated instruction classes
Fixed the problem with multiple outputs going to stdout; works OK now
Fixed the problem with opcodes of 00 being treated as 0-length
Added Intel-format numbers for 8080 and 8008 (suffix H=hex, etc)
Improved the mechanism for finding processor-dependent code
Fixed docs: run instructions, numerous small errors in syntax docs

Release 0.6:

Efficiently recognize non-existant mnemonics.
Flush the table of macros at the beginning.
A macro/if without a endm/endif just emits an error message, doesn't die.

Release 0.5:

Added macros - macro, endm, exitm
Added conditionals - if, elseif, ifdef, ifndef, ifeq, ifneq, else, endif
Added 'error' pseudo-op
Added operators - (prefix) ! - ~ (infix) % << >> < <= > >= != == & ^ |
Made s-record output all upper-case
Fixed limitation of forward referencing depth of 1
Fixed ArrayIndexOutOfBounds writing s-records when org != 0
Fixed code generation for 16-bit indirect addressing (6809 only)
Fixed recognition of trailing characters on an expression list
Fixed handling of strings as numbers, and documentation about it
Improved speed of output to GUI

Release 0.4:

Added 8080 support
Added GUI
Added Applet

Release 0.3:

Added 6800 support
fixed checksums in s-record output
added S5 and S9 records in s-record output
added 'end' pseudo-op
Included pathnames are relative to the directory of the including file (unless the pathname is absolute)
Allow for opcodes that match Java keywords
Made 'include' 6809-dependent; pseudo-op may differ for other processors

Release 0.2:

S-record output file
Shows object code in output listing
Better error reporting, somewhat more meaningful messages
Command line options for listing, symbol table, binary/s-record/no output
Changed 'incl' pseudo-op to 'include'
2-byte instruction operands no longer overwrite 2nd byte of opcode
A label on a line by itself is OK
The current-address symbol "*" is now recognized
Illegal label names are now complained about
Fixed bounds-checking for prettyShort and reallyShort index offsets
"Really short" indirect mode now generates "pretty short" code
Now generates the right opcodes for LDS/STS
Now recognizes JSR
Numerous other problems...

Known bugs:

none yet...