Re: [Devel] Problems with Adobe's free CID fonts

[Tom Kacvinsky]

I found the problem: the subroutines are decrypted before they are stored, but
the seed bytes (lenIV of them) are not stripped before storage.  As such, when
the charstrings are parsed and a subroutine is called, the first byte of the
subroutine that is parsed is one of the seed bytes. Oops.

I am investigating how to best strip the seed bytes right now.  It is tricky
because of how the subroutine storage is set up.

Tom

On Sun, 11 Mar 2001, Werner LEMBERG wrote:

> > What I meant is that the byte sequence for the callothersubr
> > operator is 12 + 16.  The operands are encoded differently.  So you
> > have a sequence of operand bytes (one of which is the number 12 or
> > 13, to indicate which othersubr to use), then an escape byte (12)
> > followed by the byte 16.
>
> Yes :-) I've read this piece of the PS specs.
>
> > This is what I am really after: are the problems I am seeing from
> > undocumented operators, or from a bug in FreeType?
>
> A question to Ken doesn't harm...
>
>
>     Werner
>