woo@pioneer.arc.nasa.gov (Alex Woo RAA) (04/21/91)
A few days ago, Ralph Carmichael asked if any summaries of
PostScript performance were available. The response was
underwhelming. Here is a very simple test file which Jim Sullivan
posted to comp.lang.postscript years ago. We have a limited variety of
PostScript printers and we are most interested in the performance of
the newer models, such as the HP LJ IIIsi, Kodak 716PS, etc.
We would also be interested if a particular PostScript clone fails
on this page. This includes PostScript previewers.
We have run this file on Apple LaserWriters, NeXT '030 printer, and
DEC LPS40.
If you decide to print this file, please email the timing number to
carmicha@pioneer.arc.nasa.gov or mail the output page to the address
on the sheet.
Thanks.
========================================================================
Alex Woo, MS 227-6 woo@ames.arc.nasa.gov
NASA Ames Research Center __o NASAMAIL ACWOO
Moffett Field, CA 94035-1000 -\<, SPANET 24644::WOO
(415) 604-6010 (FAX) 604-4357 .....O/ O {hplabs,decwrl,uunet}!ames!woo
========================================================================
%! PS-Adobe
%% Laser Doctor, Version 1.0.0
%% Written by Jim Sullivan, January, 1989.
%% This was written out of a need to display as much information
%% about various Postscript Laserprinters onto one page. More
%% can be added and anyone is free to modify it for their own
%% needs.
gsave
%% Beginning of definitions ********************************
/HB {/Helvetica-Bold} def
/CBO {/Courier-BoldOblique} def
/C {/Courier} def
/FF {findfont} def
/SS {scalefont setfont} def
/mt {moveto} def
/s {show} def
/fontname 30 string def
/getfont {pop fntnm cvs /fontname exch def} def
/fntnm 30 string def
/str 20 string def
/BOLD {CBO FF 10 SS} def
/NORMAL {C FF 10 SS} def
/cnt 0 def
/btime 0 def
/circleofbench
{ 15 15 345
{gsave
rotate 0 0 mt
(Benchmark) oshow
grestore
} for
} def
/oshow {true charpath stroke} def
%% End of definitions *****************************************
BOLD
200 756 mt
(Analysis by Doctor Laser, Version 1.0.0) s %Print title
20 720 mt
NORMAL (This printer is a ) s
BOLD statusdict/product get str cvs s %Gets the name of the
NORMAL ( running version ) s %printer from
BOLD version str cvs s %statusdict/product
NORMAL ( of Postscript.) s %and the version number
20 700 translate %of Postscript from the
%'version' command
0 0 mt
(Available Fonts : (PaintType)FontName:Example of font) s %Heading
0 -10 mt
(PaintType = 0\(filled\), 1\(stroked\), 2\(outlined\),) s
0 -20 mt
( 3\(mixed\), ?\(unknown\)) s
0 -30 translate
BOLD
FontDirectory {BOLD 0 0 mt %Push directory of fonts and
getfont %get the fonts one at a time.
(\() s
{fontname cvn FF /PaintType get str cvs s} stopped {%ifelse
(?) s } {} ifelse %Get the font's PaintType and print its value,
(\)) s %or if it is not supplied, print a '?'.
fontname s %Print the name of the font.
mark
fontname length 1 30 {(-) s} for %Print dashes out to the font sample column.
(:) s %Print a colon.
{fontname cvn FF 10 SS ( ABCDEFGH abcdefgh) s } stopped {%ifelse
NORMAL (Error implementing font) s } {} ifelse %Show a sample of each
cleartomark %font. If an error is
cnt 10 add %trapped, print message
/cnt exch def %indicating an error
0 -10 translate} forall %occurred. Use the
NORMAL %variable 'cnt' to count
cnt 30 add /cnt exch def %the number of fonts.
350 cnt mt %Move back up the page 'cnt' points.
(Virtual Memory Status:) s %Print the Virtual Memory Status
0 0 translate %using 'vmstatus'.
370 cnt 20 sub mt
(Maximum Available Bytes = ) s BOLD vmstatus str cvs s NORMAL
370 cnt 30 sub mt
(Bytes currently in use = ) s BOLD str cvs s NORMAL
370 cnt 40 sub mt
(Level of Save Nesting = ) s BOLD str cvs s NORMAL
350 cnt 60 sub mt
(Cache status \(Red book p126\):) s %Print out the cache status.
cachestatus
370 cnt 80 sub mt NORMAL (blimit = ) s BOLD str cvs s
370 cnt 90 sub mt NORMAL ( cmax = ) s BOLD str cvs s
370 cnt 100 sub mt NORMAL ( csize = ) s BOLD str cvs s
370 cnt 110 sub mt NORMAL ( mmax = ) s BOLD str cvs s
370 cnt 120 sub mt NORMAL ( msize = ) s BOLD str cvs s
370 cnt 130 sub mt NORMAL ( bmax = ) s BOLD str cvs s
370 cnt 140 sub mt NORMAL ( bsize = ) s BOLD str cvs s
usertime /btime exch def %Print a graphic (circleofbench)
HB FF 12 SS %and time how long it takes for
430 cnt 230 sub translate %the printer to interpret it.
.5 setlinewidth
circleofbench
0 0 moveto
(Benchmark Testing) true charpath
gsave 1 setgray fill grestore
stroke
-40 -80 mt
NORMAL
(Time to print) s
-40 -90 mt
(the above graphic) s
-40 -100 mt
(was ) s
BOLD
usertime btime sub 0.001 mul str cvs s %Print out the time in seconds.
NORMAL
( seconds.) s
-40 -210 mt
( Please mail this page to)s
-40 -220 mt
( Ralph Carmichael ) s
-40 -230 mt
( MS 227-2) s
-40 -240 mt
( NASA Ames Research Center) s
-40 -250 mt
( Moffett Field, CA 94035-1000) s
-40 -260 mt
( carmicha@pioneer.arc.nasa.gov) s
grestore
showpage
========================================================================
Alex Woo, MS 227-6 woo@ames.arc.nasa.gov
NASA Ames Research Center __o NASAMAIL ACWOO
Moffett Field, CA 94035-1000 -\<, SPANET 24644::WOO
(415) 604-6010 (FAX) 604-4357 .....O/ O {hplabs,decwrl,uunet}!ames!woo
========================================================================lars@SPECTRUM.CMC.COM (Lars Poulsen) (05/03/91)
(1) the "str" string is too short - our IMAGEN traps on a range check
when it tries to get the system name.
(2) Apple LaserWriter II NTX - 21.5 seconds
Imagen 5320 ImageServer - 50.0 seconds
Apple LaserWriter Plus - 115.0 seconds
Nice test...
--
/ Lars Poulsen, SMTS Software Engineer
CMC Rockwell lars@CMC.COMdavis@EE.UDEL.EDU (Michael Davis) (05/10/91)
In article <9105031718.AA15690@crayola.cs.UMD.EDU> lars@SPECTRUM.CMC.COM (Lars Poulsen) writes: >(1) the "str" string is too short - our IMAGEN traps on a range check > when it tries to get the system name. > The HPIIIsi had the same problem. The name was 30 characters long and str was defined as 20 >(2) Apple LaserWriter II NTX - 21.5 seconds > Imagen 5320 ImageServer - 50.0 seconds > Apple LaserWriter Plus - 115.0 seconds > HP IIIsi = 5.612 seconds DEC PrintServer 20 = 26.29 seconds The DEC has the same ppm rating as the HP with a price tag about 10 times higher.. mike
sherwin@royalt.enet.dec.com (Jim Sherwin) (05/15/91)
In article <9105132000.AA02334@crayola.cs.UMD.EDU>, davis@EE.UDEL.EDU (Michael Davis) writes... > >In article <9105031718.AA15690@crayola.cs.UMD.EDU> lars@SPECTRUM.CMC.COM (Lars Poulsen) writes: > >HP IIIsi = 5.612 seconds >DEC PrintServer 20 = 26.29 seconds Given that our PrintServer 20 has been on the market for approx. 2 years, it's understandable that more recently introduced products would include impovements in throughput, among other things. I'm confident that the PrintServer 20 is not the last PrintServer product we will bring to market. >The DEC has the same ppm rating as the HP with a price tag about 10 times >higher.. These products have different speed ratings. The PrintServer 20 is rated at 20 ppm whereas the H-P LJIII Si is rated at 17 ppm. There is a price differential. IT IS NOT, HOWEVER, ANYWHERE NEAR 10x. I think it's reasonable for readers of this news group to expect that assertions made by submitters be accurate. <>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<> <> "When things are going well, \ Jim Sherwin a.k.a. <> <> someone will inevitably \ sherwin@royalt.enet.dec.com <> <> experiment detrimentally." \ !decwrl!royalt.enet.dec.com!sherwin <> <> Boyle's Second Law \ sherwin%royalt.enet@decwrl.dec.com <> <>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>-<>