[comp.sys.ibm.pc] Parallel vs. Serial for Postscript printing

gardiner@umn-cs.cs.umn.edu (David Gardiner) (05/04/90)

I have a HP Laserjet Series II with a Pacific Page postscript emulation
cartridge, hooked up to a 386 (DOS).  I want to make the thing print
faster and believe that most of the print time is in transmission.

I read somewhere that the data transmission to the printer would be
better if I used a serial connection than with the current parallel
one.  Even though the parallel interface transmits more bits per
cycle, the serial can run at a much higher cycle rate.

Does anyone know if this is true?  I think (my manauals are at home) that
serial can transmit at 9600 baud.  What is the speed of a parallel port?

Thanks in advance.

David Gardiner

P.S.  If anyone wants a review of the Pacific Page, let me know.

-- 
--------------------------------------------------------------------
David Gardiner        University of Minnesota Computer Science Dept.
gardiner@umn-cs.cs.umn.edu

pnl@hpfinote.HP.COM (Peter Lim) (05/06/90)

> I have a HP Laserjet Series II with a Pacific Page postscript emulation
> cartridge, hooked up to a 386 (DOS).  I want to make the thing print
> faster and believe that most of the print time is in transmission.
> 
If you print a lot of bit-mapped graphics then this would be true.
Otherwise, Postscript (vector based) usually transmit a lot less code 
compared to PCL (bit-mapped).  Of course, the faster you can transfer
the merrier.

> I read somewhere that the data transmission to the printer would be
> better if I used a serial connection than with the current parallel
> one.  Even though the parallel interface transmits more bits per
> cycle, the serial can run at a much higher cycle rate.
> 
> Does anyone know if this is true?  I think (my manauals are at home) that
> serial can transmit at 9600 baud.  What is the speed of a parallel port?
> 
Don't know about serial being faster than parallel. For a 9600 baud
serial transfer to be faster than a parallel port, the parallel port has
to transfer at less than 9600 / 8 = 1200 times/second. Which means
about under a millisecond per transfer. I don't know how slow the PC's
parallel port is, I do remember building a parallel transfer system
a long time ago and the rate was like 35 K (Byte or Bit, I can't quite
remember) per second.  That was a polling system (not even interrupt 
driven !). Assuming it is 35K bit/second, it is still about 4 times the 
speed of a 9600 baud serial transfer.  So there ..... Of course I could
be wrong  :-).


On a different note; I do have second hand information that the slowness
is due to the Pacific Page postscript emulator (PPPE). A friend of mine once
tried the PPPE on a LaserJet IID (I believe this is faster than the
series II); the thing was hooked on to a UNIX machine. He dumped a short
but recursive postscript file (to print a recursive triangle/pyramid type
of pattern), it took about 5 minutes. I tried the same file on a 386-25
no cache, with 80387 and Ultrascript (a postscript interpretor). The
printer is the lowly LaserJet IIP (the half-speed low cost printer)
using parallel port. The page came out at under 2 minutes ! Sure is a lot
faster than PPPE on a faster printer ! Of course I have lots of RAM and
gave Ultrascript about 3 MB of extended memory to play with.

One thing though, when Ultrasript process the postscript file, everything
stops on the 386  :-).


Regards,                       ## Life is fast enough as it is ........
Peter Lim.                     ## .... DON'T PUSH IT !!          >>>-------,
                               ########################################### :
E-mail:  plim@hpsgwg.HP.COM     Snail-mail:  Hewlett Packard Singapore,    :
Tel:     (065)-279-2289                      (ICDS, ICS)                   |
Telnet:        520-2289                      1150 Depot Road,           __\@/__
  ... also at: pnl@hpfipnl.HP.COM            Singapore   0410.           SPLAT !