-ms accent bug; sloppy fix

[Mark Turner]

In article <> mark at gargoyle.UUCP ( Turner) writes:
>Apparently, in 1983 the -ms macros were revised to
>include a .AM macro which, when placed at the head of
>a file to be formatted by nroff/troff, allowed one to use 
>various new accents and characters.  /usr/lib/tmac/tmac.s
>on our 4.2 system contains the lines
>.de AM
>.so \*(\\s.acc
>..
>meaning, I take it, read and use /usr/lib/ms/s.acc.
>Now, our /usr/lib/ms/s.acc contains the line
>
>.ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\(hy\fP\v'.25m'\h'-\*(#H'
>
>This should mean that when one types \*(d- in a file sent through
>n/troff, n/troff understands it as a lower case "eth", a character in Old
>English.  The eth should look something like a partial-differential
>sign with a dash through the stem.  The problem is: this line causes
>troff to print the partial-differential-sign, and then backspace TWICE
>to print a dash, and then proceed.  It should only backspace ONCE.
>The result is: the dash occurs one character before it should, and
>the character that should succeed the eth is printed on top of the
>partial-differential sign.  I am too stupid to read the above code.
>Can anyone tell me how to alter it so that the partial-differential
>sign and the dash are overstruck?
>Reply to . . . ihnp4!gargoyle!mark

Since no one responded, I tried my own hand at it, and produced
a sloppy fix that works:

.ds d- \h'\*(#H'\(pd\h'-\w'~'u'\v'-.25m'\f2\ \h'.27m'\(hy\fP\v'.25m'\h'-.09m'\h'-\*(#H'

>
>