
-------------------------------------
TMPview v1.3
-------------------------------------

    FEAT: The default value for the source tab width is now taken from the
          input file instead of being always nine (9). [0000122]
    FEAT: "--unnamed-hex" added; when specified TMPview will output hex
          notation instead of the default decimal for un-named bastext tokens.
    FEAT: "--petscii-dec" and "--petscii-hex" added; when specified TMPview
          will output *all* PETSCII-only characters (those that cannot be
          displayed rationally in ASCII output) to bastext tokens in either
          decimal or hex notation rather than their named counterparts.
    FEAT: "--byte-dec" and "--byte-hex" apply to single character literals.
    FEAT: "--byte-columns" added; the given value sets how many values per
          line the "--byte-dec" and "--byte-hex" features output.
    FEAT: "--stats-detail" and "--stats-wild" added to display interesting
          and extremely critical statistical output, respectively.
    CHNG: "--screen" now regards values 0-63 as potential screencodes but
          only converts strings where every character qualifies.
    BUG:  corrected tabbed output when --tab-width was 0
    BUG:  bastext token {cm o} is no longer output as {cm d}  [0000123]
    BUG:  bastext token {purple} is no longer output as {156}  [0000124]
    BUG:  handle the fake '_' character properly in quoted strings and
          macro bodies.
    BUG:  .include filenames are now translated.



-------------------------------------
TMPview v1.2
-------------------------------------

    FEAT: Some error output was made more clear/useful to help debug future
          issues that may arise with additional input files
    BUG:  brk is now recognized/translated properly.
    BUG:  labels on lines with a program counter assignment were omitted.
    BUG:  parser is more tolerant of source code with an odd/corrupt label
          table - null label entries are skipped



-------------------------------------
TMPview v1.1
-------------------------------------

    FEAT: Default behaviour when translating string and character literals
           is to encode non-printable ASCII output into tokens 'compatible'
           with tok64/bastext (e.g., "{reverse off}"). Two additional options
           are available to change this behaviour: "-b"/"--byte-dec" will
           translate all string/character literals into sequences of .byte
           statements using decimal notation. "-B"/"--byte-hex" will perform
           a similar translation, outputting in hex notation. TMPview would
           formerly ignore "non-printable" ASCII output.
    FEAT: New option, "-S"/"--screen", to translate .text/.null/.shift
           statements into .screen statements. This attempts to be smart
           about identifying those string literals that appear to consist
           of screencodes instead of so-called "chrout" PETSCII.
    FEAT: "-i"/"--in" not required ("tmpview <file>" works)
    FEAT: If an input filename is not specified using "-i"/"--in", nor taken
           from a loose argument as above, input will be taken from stdin.
    FEAT: Added identification and translation of sources saved by "Turbo
           Assembler Professional v1.1lm" by Mr. Fiz/Samar, a plain turbo
           variant with sources starting at $8a00.
    FEAT: Added identification (but not translation) of sources saved by
           "Turbo Macro Cross Assembler v1.0" by Jolz/Onslaught, a
           substantial modification of plain turbo with sources starting
           at $87ff.
    FEAT: Added recognition of extra DTV v2 opcodes [0000110].
    FEAT: "--omit-source" added as a ditto for "-O".
    BUG:  bit absolute opcodes were incorrectly output as x indexed [0000091]
    BUG:  some command line options could cause abnormal termination at
           least on windows (e.g. "-o" without a corresponding filename)
           [0000090]
    BUG:  accumulator addressed opcodes (e.g. lsr, rol) now more accurately
           mimic the output from TMP/Turbo Assembler by including the 'a'
           (e.g. "lsr a" instead of just "lsr").
    BUG:  sometimes the last source code line was omitted from output.
