the whole format looks very strange to me, what about
texnansi-raw-hum521b Humanist521BT-Bold "TeXnANSIEncoding ReEncodeFont"
Jens
It's a line from file texnansi-bitstrea-humanist.map generated by texfont (texfont --help;
see also mtexfont.pdf, mfonts.pdf and pdftex-a.pdf)
This line say, among others, that only glyphs from font hum521b.pfb (a
Type1 font) used by A.pdf document must be include into A.pdf (partial
inclusion).
Until now, I haven't found any problem with xpdf and acroread, but
postpocressing with gs sometimes fails, because some glyphs seems to
call an instructions (NP) which definitions is not included.
Hence with
texnansi-raw-hum521b Humanist521BT-Bold "TeXnANSIEncoding ReEncodeFont"
and in http://www.ghostscript.com/pipermail/gs-devel/2002-October/002043.html I find the following << #616539 - USER - gs fails for pdflatex generated document assigned: alexcher If a PDF document generated by pdflatex contains some special combinations of type1 fonts, gs fails with "Error: /undefined in NP" while acroread and xpdf have no problems to display the document. Alex determined that the PDF file is invalid, and that it includes a subsed of /CharterBT-Roman font that uses NP procedure but doesn't define it in the font header. Should we close the bug?
I'm sure that it's not a complete explanations, but I have to study the Adobe manual on Type1 fonts and encodings , and time is not on my side.... luigi