🚀 go-pugleaf

A question for Type-1 gurus.  (My apologies if this topic has been
discussed recently, but I've not been keeping up with this newsgroup
recently.)

I maintain a piece of software that converts Type 1 fonts into Acorn's
proprietary font format.  Because it does not incorporate a full
PostScript interpreter (or anything remotely close!) it relies on
fonts conforming to the simplified and predictable form described in
Chapter 10 of the "Adobe Type 1 Font Format" book.  Recently one of
our developers showed me a font which fails to convert.

The font is a "Bold Oblique" font, and on examining it I found that it
has been implemented as a cut-down font with no CharStrings of its
own.  Instead the whole "Bold" version of the font has been physically
included in the file, with code to avoid defining it if the printer
already knows about it.

There is some code at the bottom that copies (by reference) the
"nested" font's CharStrings across into the font being defined.

Now that I am aware of this technique I want to update my program to
cope with it.  I have some questions though:-

(1) Does Adobe condone this sort of thing, given the strictures in
Chapter 10 of the Black Book?

(2) Can ATM cope with this sort of font?  If not, then presumably they
will gradually disappear and be replaced by ones that it can cope
with.  If ATM cannot handle this technique, then I won't either.

(3) Does anyone ever nest fonts MULTIPLY?  If so, this will complicate
my code, but on the other hand I can't see why you'd want to do this.

Thanks in advance for any info!

-- Steve
-- 

				Steve Hunt	steve@acorn.co.uk