Adobe Reader or printer driver bug

Adobe Reader or printer driver bug?

I've run into a problem printing PDF files with Adobe Reader 9.1.2 on
Windows XP using the latest PCL drivers on both of my available
printers (a Ricoh and an HP).  A minimal example PDF file (1kb) is
here...

http://www.jjpuzzles.com/example.pdf

The PDF is created from the following postscript code using
ghostscript 8.6.4.

----------------------------------
%!PS-Adobe-3.0
100 dup scale
1 25 div setlinewidth
1 4 translate
gsave
    0 0 moveto 3 0 lineto 3 3 lineto 0 3 lineto closepath clip
    newpath 0 0 moveto 1 2 lineto 10 2 lineto stroke  % NOTE xmax=10
grestore
0 -3 translate
gsave
    0 0 moveto 3 0 lineto 3 3 lineto 0 3 lineto closepath clip
    newpath 0 0 moveto 1 2 lineto 100 2 lineto stroke % NOTE xmax=100
grestore
showpage
--------------------------------

When viewed on-screen by Adobe Reader, the top and bottom sets of line
segments look identical (as expected).  When printed on either of my
printers, the bottom set of line segments contains two additional,
partially clipped, line segments.  The printed version looks exactly
as if it was created by the following postscript code...

----------------------------------
%!PS-Adobe-3.0
100 dup scale
1 25 div setlinewidth
1 1 translate
0 0 moveto 3 0 lineto 3 3 lineto 0 3 lineto closepath clip
newpath 0 0 moveto 1 2 lineto 100 2 lineto stroke
newpath 3 2 moveto 3 0.02 lineto 0 0 lineto stroke
showpage
-----------------------------------

This problem arose for me when creating plots of portions of data sets
using Asymptote and Tikz.  Can anyone else reproduce the extra line
segments in the printed version?  Is this likely an Adobe Reader bug,
a printer driver bug, or a misunderstanding on my part?

Thank you,
James

Re: Adobe Reader or printer driver bug?

On Jul 28, 9:59=A0am, ken <k...@spamcop.net> wrote:
> I'm not sure what you mean by the extra segments, I don't see any extra
> segments in the content stream though. The output when printed via GS or
> another PDF consumer in my posession are roughtly the same as the
> Acrobat display.
>
> BTW you say you used GS 8.64, but:
>
> /Producer (GPL Ghostscript 8.61)

Thank you for your investigation and for noticing the ghostscript
version discrepency.  I had downloaded 8.64, but forgot to install
it.  I still have the same issue after the upgrade.

I tried printing from GS and I do not get the extra line segments.  I
guess there is a problem with Adobe Reader or my PCL printer drivers.

Anybody else have the extra line segments when printing with Reader
9.1.2?

James

Re: Adobe Reader or printer driver bug?

In article <3eb65935-57b1-4311-8f14-
cd5165d9f62d@b15g2000yqd.googlegroups.com>, jajohnson@parker.com says...

> Thank you for your investigation and for noticing the ghostscript
> version discrepency.  I had downloaded 8.64, but forgot to install
> it.  I still have the same issue after the upgrade.
> 
> I tried printing from GS and I do not get the extra line segments.  I
> guess there is a problem with Adobe Reader or my PCL printer drivers.
> 
> Anybody else have the extra line segments when printing with Reader
> 9.1.2?

Ah....

I printed the original file from Acrobat 7 and Acrobat 9, to a PCL 
printer driver on FILE. I then ran the two PCL files through GhostPDL to 
TIFF.

Actobat 7 looks as expected, Acrobat 9 produces two extra lines which 
close the figure. Given that its the same PC, OS, printer driver and PCL 
consumer I think that points the finger pretty firmly at Acrobat 9.

In addition the Acrobat 7 output is 21KB, the Acrobat 9 output is 2KB

		Ken

Re: Adobe Reader or printer driver bug?

On Jul 28, 11:28=A0am, ken <k...@spamcop.net> wrote:
> I printed the original file from Acrobat 7 and Acrobat 9, to a PCL
> printer driver on FILE. I then ran the two PCL files through GhostPDL to
> TIFF.
>
> Actobat 7 looks as expected, Acrobat 9 produces two extra lines which
> close the figure. Given that its the same PC, OS, printer driver and PCL
> consumer I think that points the finger pretty firmly at Acrobat 9.

Thanks again Ken.  I've posted a message at the Adobe Reader forum at
the link below...

http://forums.adobe.com/thread/468998?tstart=3D0

James

Re: Adobe Reader or printer driver bug?

Same in Reader 8.1.4, using a Samsung laserprinter

If the end point of the horizontal line in your example PS file is X, Y

where X = 3 and Y >= 29

the printer draws an extra slanted line from 0,0 to X,Y, which is clipped,
and an extra vertical line from 3,2 to the extra line, which is also clipped
(half width).

The critical value is somewhere between 28 and 29, which is in this case
approximately 1 meter on paper.

It seems that Acrobat/the Reader, or the driver, or whatever, does not
accept lines longer than ~ 1 meter, even when they would not be printed as
they are clipped in the code.

Govert

http://www.noliturbare.com

Re: Adobe Reader or printer driver bug?

> where X = 3 and Y >= 29

Obviously :-) I meant Y= 2 and X>=29

This very simplified PS code...

%!PS-Adobe-3.0
4.0 setlinewidth
100 400 moveto 200 600 lineto 10000 600 lineto stroke
showpage

...when converted to PDF, has a comparable effect. It clips at the (right)
edge of the MediaBox (A4) minus printer margin (a few mm) and draws the 2
unwanted lines.

A similar PDF, directly created with iText, has the same effect.

It also happens with the HP DesignJet 1050C driver (large size plotter),
which gives these results:

- the wanted and unwanted lines are represented as vectors
- the critical length of a line is ~ 2 meters (~ 5900 points)
- the selected page size (A4 .. A0) has no effect; only the "Oversize"
versions of these sizes have a slightly smaller (!) critical length

As the clipping occurs at/near the edge of the MediaBox and the extra lines
come with quite different drivers, I believe it's a Reader bug that causes
the extra lines.

Govert