[Fontinst] Bug in fontinst 1.929?

Lars Hellström lars.hellstrom at residenset.net
Thu Jun 2 19:30:07 CEST 2005

At 17.15 +0200 2005-06-02, Peter Dyballa wrote:
>James Kass' Code200x fonts it now happens that these dotless j slots
>get filled with other glyphs -- and my Perl script to adapt a glyph
>names scheme like 'uniABCD' to the usual fontinst names now fails a bit
>by making the AFM files cheat they would have a few dotless j. So that
>glyph is not taken from the fc0r9z/fc0ro9z fonts -- can this be the
>reason that fontinst refuses to add this into the mapfile?
>	fc0r9z Code2000LCDFJ <fc0r9z.pfb
>	fc0ro9z Code2000LCDFJ <fc0ro9z.pfb "0.175 SlantFont"

To be blunt: Yes, that is most likely the case --- it's not a bug in
fontinst, but a problem with the data fed to it. So why do you persist in
choosing "Bug in fontinst ..." as subject? It's not particularly
descriptive, and definitely not of any help for anyone who might actually
have the same problem and in the future browses the mailing list archives!

Less bluntly: The reason no fc0r9z entry is put in the mapfile is that
there is no \makemapentry{fc0r9z} in the .recs file. (The logic so far this
is hopefully obvious.) The reason there is no \makemapentry{fc0r9z} in the
.recs file is that none of the VPL file generated refer to fc0r9z as a base
font (as you should have no problem verifying). From the command file cited
it is clear that this font _may_ contribute to the glyph bases of several
fonts, but apparently it does not contribute anything new that t1.etx will
make use of.

As far as `dotlessj' is concerned, the answer might be that it comes
already in fc0r8r (8r has a dotlessj slot) or fc0r8p. It's impossible to
tell from the information you provided, but if you read the comments in the
VPL you should see what got placed in the slot where you expected
`dotlessj'. If you see something like

(CHARACTER D 26 (COMMENT dotlessj)
   (CHARWD R 0.296)
   (CHARHT R 0.464)
   (CHARDP R 0.205)
      (SETCHAR D 106) (COMMENT j)

then you've got the default fake (an ordinary j with the height reset so
that any accent placed on it will fall on the dot). If instead you see
something like

(CHARACTER D 26 (COMMENT dotlessj)
   (CHARWD R 2.58997)
   (CHARHT R 4.07996)
   (CHARDP R 2.68994)
   (CHARIC R 0.31995)
      (SELECTFONT D 1) (COMMENT R-zfadr+i at 10.0pt)
      (SETCHAR D 67) (COMMENT dotlessj)

then it's a glyph from some font other than the zeroth map font. Et cetera.

Lars Hellström

More information about the fontinst mailing list