Casey,
    Re my last letter to you ( the one you're about to read) (love this mail
program).  I ran a test on Norexlax and determined that the reason that we
got negative numbers in the "char ..." output from lexdebug is that the
printf statement used was in error.  I go on at great length about this
in my last letter (to you the next one) in section "1" but here's some
important info.  The libln routing "allprint" is used to print out that
info.  The statement was "printf("\\%-3o", c)".  Apparently the intent
was to precede the value with a "\".  I wrote a little program with
a print statement like that and another without the "\\".  When called
with a positive "c" (c = 2) the print statement with the "\\" printed
"-2".  The one without the "\\" printed "2".  In other words Norabomination
couldn't deal with the attempt to print "\" and printed "-" instead.  So
those weren't really negative numbers.  I have changed the print statement
in "allprint" to "printf("%-3x", c)" and looked at the output from ntestl
run on Norpuke.  Looking at my little ebcdic card I see that the correct
hex values for the input are being printed by LEXDEBUG.  e.g. When input
is 0123 I get "char f0", "char f1", "char f2", "char f3".

 If you don't trust me you can see the test program in /usr/eus/bjs/dcon/
lextest/try.c.  Please save this letter as it will probably make more
 sense after you read the first one I sent you on the subject of nlex.

				Barry
