🚀 go-pugleaf

>From: gymble!harvard!topaz!unirot!patwood@brillig.umd.edu (Patrick Wood)
> Since there are few people who have actually seen an Interpress and/or DDL
> printer, it's hard for anyone to say anything that isn't based on some
> manufacturer's claim or second-hand knowledge.

Now let's hear from someone who not only has seen an Interpress
printer, but worked on one during the summer of '84 writing Unix
ditroff support software for it ......... me!

> Xerox says that Interpress is designed for speed and is faster than
> PostScript.  Considering that the only version of Interpress that's
> available now doesn't do graphics, only text and scanned images (no
> vector graphics, clipping, or rotation of text), I can see why it's fast.

Interpress IS designed for speed.  Interpress files are 8 bit data and
NOT ASCII.  An Interpress command is one byte long instead of a
space-terminated string of arbitrary length.  This means faster
interpretation on the printer server end and faster transfer time since
Interpress files are smaller than, say, Postscript files.

Now, it's been awhile, but I definitely remember line graphics and I
definitely remember that that 8044 I used did them.  I also remember,
however, that it did not do circles and arcs and splines (I had to
build bitmaps and ship those to do the equivalent ditroff graphics).
But I'm almost positive that it did line graphics.  What else would you
mean by "vector graphics"?

And it certainly does rotation of text.  However, unlike Postscript,
the standard Interpress server needs to store the fonts pre-rotated.
So (although the standard does not make this restriction) no Interpress
printers are capable of *arbitrary* rotation of text:  they are capable
of certain transformations.  Adding more fonts gets more rotations.

Clipping I never tried, so I can't say anything about that.

> Read that any way you want, to me it means that
> Interpress gains speed by giving up some flexibility.

But not much flexibility.  They gave up the ability for actions on one
page to influence actions on another (nothing from one page carries
over).  It is not clear to me that such a "feature" is all that
advantageous.  The first part of every Interpress file is a "preamble".
The state of the printer at the end of the preamble is used as the
starting state for every page in the document.  So, fonts and
transformations and other global stuff is defined in the preamble and
is then available at the beginning of every page.  But one page can't
do something (either intentionally or inadvertently) that affects pages
later in the document.  This make it MUCH easier to arbitrarily
rearrange pages to your heart's content.  I understand that this
approach is *possible* in Postscript, but not *enforced*.  This makes
writing a program to rearrange any given Postscript file much harder.
Right?

			William LeFebvre
			Department of Computer Science
			Rice University
			<phil@Rice.edu>