🚀 go-pugleaf

<Stephen_Moye@adobeforums.com> wrote in message 
news:3bc27cf3.-1@webcrossing.la2eafNXanI...
> An author has sent us some graphics made with a program called "inkscape". 
> The eps file has this coding:
>
> /newlatin1font {findfont dup length dict copy dup
>    /Encoding ISOLatin1Encoding put definefont} def
>
> /(Times-Roman-ISOLatin1) /(Times-Roman) newlatin1font
> ...
> %%[ Error: Times-Roman-ISOLatin1 not found. Font cannot be embedded. ]%%
> ...

It is legal PostScript, but certainly with a different semantic
than what the code intends.

The "/(Times- ..." line in the code fragment above does the following:

- Push a litteral empty name on the operand stack; "/" by itself
  is legal.
- Push the string "(Times-Roman-ISOLatin1)"; note that "()" are
  delimiters, so there's no need for whitespace between "(" and the
  preceding token.
- Push a second empty litteral name.
- Push the string "(Times-Roman)".
- Execute "newlatin1font". This looks or a font named "Times-Roman",
  reencodes it, then defines a font with an empty name. Note that
  in some circumstances names and strings are interchangable, so
  "(Times-Roman)" can be used instead of "/Times-Roman".
- The 1st "/" and "(Times-Roman-ISOLatin1)" are forgotten on the stack.

A consequence of the above it that no font gets defined with the name
"Times-Roman-ISOLatin1", which later leads to the error you mention.