🚀 go-pugleaf

There are generally two reasons why encapsulating a piece of
PostScript in this way will fail.

1. Length.  The procedure has to fit in memory, of course - many
printers have very limited memory, as little as 400K for everything.
But also a procedure has to be built on the operand stack. That has a
limit in size. And finally a procedure is an array, and arrays have a
maximum size - typically no more than 64K, sometimes less.  Only the
memory can be increased, the array limit is architectural.

2. Inline data. Any image data is likely to appear inline, and that
makes it impossible to wrap in a procedure (since the data has to be
inline at the time the procedure is run, AFTER the call, not in the
procedure).

The general problem is a difficult one. Many people have tried to
simply turn general chunks of PostScript into procedures, but found it
impractical.

If you have a level 3 printer with adequate memory, look into the
ReusableStreamDecode filter instead.  

Aandi Inston