perm filename FORMAT[1,BGB] blob sn#064443 filedate 1973-10-01 generic text, type T, neo UTF8
∂28-SEP-73  1613		ACT,REG
 I have implemented the following extension to font files:
 In the character definition, bits 0-8 of the word which
 used to contain only character_code,,word_count+2, are
 interpreted as the "raster width" of the character.  If this
 field is 0, the raster width is the same as the character width
 (so the existing fonts will work).  Otherwise, this field
 is the actual width of character definition, distinct from
 the width that the character takes when printed adjacent to
 other characters.  The intention is to allow characters
 to be overlapped.  For example in the word "First" the top
 bar of the "F" could overlap the "i".  (In this case the
 character width is less than the raster width).  (I don't
 see much utility for the other case - which allows characters
 to be spaced farther apart than usual)