• Please review our updated Terms and Rules here

Laser Printer as accelerator?

G

geoffm3

Veteran Member
Joined
Oct 9, 2009
Messages
1,272
Location
Huntsville, AL
It's been said many times in the past that the early laser printers had compute power equal to or better than the host machines that they were connected to. Were there ever any attempts made to harness this extra power for applications other than printing?
 
When you mean early do you mean xerox PARC stuff from the late 60's early 70's or do you mean the IBM 3800 series from the 70's? Or are we going back to the original hp laserjet from mid 80's? ( I wish I kept mine! but that thing was huge)
 
the early Apple lasers ARE Hp lasers, they are the same as the hp LJ II and III series, same parts in most cases. I used to do minor service to hp II III and the apple ones (apple laserwriters? maybe that was thier name) at my first IT job.
I owned the original HP laserjet, it had a funny model number even though it was the first,, and it only had serial communication. Big units, real workhorses, they would still run fine today if you had one. Only drawback was 300dpi up until the hp LJ4 series.
I didnt mind, I had an HP Lj IIP because it was smaller for years, sad I gave it away.

Wells they have logic boards, they have RAM. They have IO and expansion bays. I think they were RISC based. I dont know what you would do with them. modern laserjets have far more advanced computer systems in them. Ill agree they were smarter than dot matrix or daisy wheels of the time.
 
Don Lancaster published a bunch of material about using the Postscript engine in laser printers back in the day. Perhaps a search on his name and the subject might yield results. Basically, the idea was that a laser printer with a 68K CPU was capable of more processing power than an 8088 4.77MHz PC.
 
Hmm interesting stuff. I see the argument, but we are talking postscript engines were necessary to interpret and rebuilt the code. That was almost all of what the cpu was doing. Hypothetically you could expand upon the system and implement new IO and Bios and talk to it and make it perform other tasks. The only reason would be just to say you did. It would be reinventing the wheel as there are already a couple decades worth of other devices that could do that too.
 
VERAULT said:
the early Apple lasers ARE Hp lasers, they are the same as the hp LJ II and III series, same parts in most cases. I used to do minor service to hp II III and the apple ones (apple laserwriters? maybe that was thier name) at my first IT job.
I owned the original HP laserjet, it had a funny model number even though it was the first,, and it only had serial communication. Big units, real workhorses, they would still run fine today if you had one. Only drawback was 300dpi up until the hp LJ4 series.
I didnt mind, I had an HP Lj IIP because it was smaller for years, sad I gave it away.

Wells they have logic boards, they have RAM. They have IO and expansion bays. I think they were RISC based. I dont know what you would do with them. modern laserjets have far more advanced computer systems in them. Ill agree they were smarter than dot matrix or daisy wheels of the time.
Click to expand...

I could see how it might have been tempting to use them to manipulate large datasets where it would have been much slower on the host machine. Presumably there was some way to inject your own program for them to execute rather than being stuck with what's burned into EEPROM.
 
A coworker at a previous job said that most of his master's papers were structured so that anything that could be offloaded to the PostScript engine in the printer, was. I'd imagine at the time he was using a 386 or 486 for his work.
 
glitch said:
A coworker at a previous job said that most of his master's papers were structured so that anything that could be offloaded to the PostScript engine in the printer, was. I'd imagine at the time he was using a 386 or 486 for his work.
Click to expand...

So he has a process to utilize the postscript engine at the time, or he developed a process?
 
VERAULT said:
So he has a process to utilize the postscript engine at the time, or he developed a process?
Click to expand...

Wasn't clear, I don't work with him anymore so I can't follow up, but from the story he told, he was having the PostScript engine in the printer do calculations that ended up coming out in the printed document because the printer was faster at doing it than his desktop was.
 
I still have my laserjet III with a PacificPage postscript cartridge. Traveling somewhere, I stumbled across a Postscript language reference book in of all places an airport bookstore. Somewhere I still have that book, it was quite a few hundred pages. I came away with the thought that postscript was "complicated".
 
glitch said:
Wasn't clear, I don't work with him anymore so I can't follow up, but from the story he told, he was having the PostScript engine in the printer do calculations that ended up coming out in the printed document because the printer was faster at doing it than his desktop was.
Click to expand...

I'm not saying I don't believe..it's just one of those things I'd like to see to believe
 
The key to understanding how to offload computations to a Postscript printer is realizing that Postscript is a Turing-complete programming language. When you print to a Postscript printer, what you're doing is sending the printer a Postscript program which, when executed by the printer, renders the desired output. Postscript was designed primarily for rendering graphics, but it can be used to perform arbitrary computations.

One of the advantages of Postscript vs. other contemporary schemes was that you could define an image in a manner which is independent of the resolution of whatever printer you use. If you create a Postscript program which draws a circle, that circle will be rasterized at 300 DPI on a 300 DPI printer, or at 2400 DPI on a 2400 DPI printer, for example. So, you could create vector graphics suitable for being blown up to billboard size, yet print proofs from the same file on your little 300 DPI laser printer at home.

For a very simple example of how one might offload calculations to the printer, consider these two examples:

  • Print "5" in a particular font at a particular location on the page.
  • Add 2 to 3, then print the result in a particular font at a particular location on the page.

Both of those examples produce the same printed output, but in the second example, the printer itself calculates the number to be printed.

As another example, you could declare a function in a Postscript program which draws a logo, and then call that function in a loop which scales, rotates, and translates the result in a calculated manner to render a bunch of copies of the logo in a spiral.

Here's something I found in a Google search which may shed some light on how you might compute arbitrary stuff on any Postscript printer:

http://www.mostlymaths.net/2008/12/quick-postscript-programming-tutorial.html
 
Robert Sedgewick's Algorithms books (at least the Pascal , C and C++) was done in such a fashion.
If the illustration in the book should depict a tree undergoing rotation they wrote a program in Pascal and instrumented
it such that the datastructure's current structure at an event was dumped. The tree was sent
in a symbolic form to the postscript engine in the printer which then did the layout of the picture.

The illustrations is beautiful.
 
You must log in or register to reply here.
Back
Top