Bruce Toback writes:

> HP publishes a 135-page catalog of differences among its PCL printers.
> This catalog does not include the HP 5000 series, which have their own
> uncataloged foibles -- and which, incidentally, are PCL 4 rather than PCL
> 5 printers.

> These differences are a huge headache for people trying to write
> graphics-oriented printer drivers. If I were in Wirt's place, I would
> have made the same decision he did.

Bah!  What's a little work? Sure it may be some work for people writing
graphics-oriented printer drivers, but I guess my supposition is that it
doesn't (and shouldn't) make that much difference to the end user.
That's why they pay us the big bucks! :)

> >
> >[snip]
> >> >  >But the
> >> >  >quality of output is significantly (to greatly) better.
> >> >
> >> >  This is misleading in the extreme.
> >>
> >> ... I was
> >> speaking about overall aesthetic image quality. There is a simple,
objective
> >> test to my statement. Count the number of books, magazines, newspapers,
> >> advertisements, catalogs, etc. that are put together using PCL -- and
> >>compare that number to PostScript-produced documents. ...
> >
> >I think this statistic is misleading.  I don't believe that the fact
> >that high end publications are produced using Postscript is directly
> >related to the quality of its output.  ... High end layout packages all
> >support Postscript because that's what the high end printing presses
> >support.  The only reason they support Postscript is because that's what
> >they've always supported.
>
> No; high-end layout packages and printing presses support PostScript
> because it's inherently device-independent. A PostScript program is a
> precise, resolution-independent mathematical description of the contents
> of a page; a PCL data stream is whatever HP says it is, meaning different
> things to different printers at different times. There have been several
> HP-sanctioned attempts at creating PCL imagesetters (essentially very
> precise, high-end printers that produce the negatives or plates for
> printing), and none have gotten very far simply because there's no point
> in sending imprecise instructions a high-precision device.

Actually, you have got me here.  I'd forgotten about this.  Yes,
Postscript's device independence is a big selling point.  This is
particularly important if you are printing on a variety of devices with
varying degrees of resolutions.  The majority of desktop publishers
proof their printouts on a desktop Postscript printer before sending the
result off to a print shop.  I don't know how much relevance this has to
someone looking to do forms.

Maybe I'll retreat a bit and say that I still think that differences
between a Postscript vs. PCL printer are largely irrelevant to the forms
marketplace unless you've got a variety of printers with differing
resolutions.  In the latter case Postscript has the edge of PCL5 (and
its earlier incarnations).

> PCL 6 does not share this failing; in fact, PCL 6 looks a lot like
> binary-encoded PostScript. It still has the resource-management issues
> that PostScript users solved a long time ago, but it's a big step toward
> a graphics system that can actually be trusted to produce the same output
> twice in a row.

This is a good point.

Wirt Atmar writes:

> Clipping paths, which are very useful for any number of graphic
constructions,
> including elaborate text constructions, are just one obvious item that's
> missing from PCL. But constraining the discussion to only those items that
are
> more likely to appear on a standard business form, I've included the
> PostScript fragment that generates our company logo. Logos are clearly a
very
> important feature of a standard business form. Our particular logo is a
> stylized Easter Island statue, incorporating a number of rounded corners,
set
> in a box with the word, "AICS", appearing at the base.

How many people program their logos in Postscript?  Wirt, I'm sure you
did it as an enjoyable programming exercise, but I don't think you
represent the majority of forms users :).  Most users will ask their
graphics arts department/marketing department for a logo.  It will
arrive in one of a variety of formats.  The vast majority of forms users
never have any contact with clipping paths.  Their presence or absence
means little to your average forms user.

> More importantly, the logo, as an object, is scalable to any size, postage
> stamp to football field, and will never appear with any greater jaggies in
the
> image than the rendering printer generates at its maximum resolution. More
> importantly yet, it is a drop-on object that does not disturb the graphics
> state -- and allows itself to be presented at any rotation, in any color
> scheme the user wishes to present it.

Again, the scaling advantage of the logo is a function of the fact that
Postscript is a vector based format.  Import a vector based graphic
(e.g. WMF, CGM, or EPS) into any forms package, you'll be able to scale
it up and down without jaggies.  This is not a problem to a forms
package user.  Yes, Postscript does scaling for the person generating
the printer code, but people generating PCL know this and scaling
(transformations of all sorts, really) are relatively simple to apply
before generating the PCL.  The end user doesn't need to concern
themselves with this fact, ergo they don't need to worry about buying a
Postscript vs. PCL printer.

[snipped anecdote proving me wrong about Postscript's origins]

I always enjoy Wirt's anecdotes.  I seem to have been mistaken about
Postscript's origins.  The Adobe red book lists its creation date as
circa 1985.  I've always assumed the Postscript migrated down from the
high end systems to the desktop printers.  Nonetheless, I still maintain
that its use on high end printers is unrelated to its inherent imaging
quality and more related to its device independence.  Take equivalent
PCL & Postscript, print them to the same device (provided the PCL was
designed for that device, I grant you), and you will get results that
most people would have trouble distinguishing between.

> PostScript on the LaserJet III didn't work well. It was slow and buggy. But
HP
> got PostScript dead right on the LaserJet IV. It took us two years to scrub
> all of the PCL code out of QueryCalc and convert it over exclusively to
> PostScript, but it's a decision I've never regretted. It has made life
> enormously easier for us and for our customers. But more than that, we've
> extremely well positioned for the future for five years now, which is very
> high quality, easy to maintain color laser printers. This is what business
> people want -- and the prices of the new printers are getting to point that
> resistance is going break as a rush of new products appears.

I agree that it has made your life as a printer driver developer easier,
but I'm afraid I still disagree that it has much effect on your end
user.  As a developer you have to make a cost/benefit choice between the
supporting multiple drivers and providing more choice for your end
users.  Part of that choice is deciding how much of a benefit providing
that the multiple drivers will be.  You obviously feel that there's not
much benefit in supporting both.  We feel that it's important for us to
support both so that our end users aren't forced to choose between us
and what their corporate infrastructure mandates (heck, we even support
Canon/Kyocera laser printers in their native formats even though they
have only a small presence in North America).

We both feel strongly about choices we've made, which is a good thing,
but as someone lives with both types of printers day in and day out, I
fear you'll never be able to convince me that Postscript vs. PCL debate
has much impact on anyone else besides those of us who generate the
Postscript/PCL, and that's how it should be :).

Rob
=======================================================
Rob McDougall            Phone:  (613)751-4800 ext.5232
JetForm Corporation      Fax:    (613)751-4864
http://www.jetform.com   mailto:[log in to unmask]
=======================================================