[cups.general] Re: incomplete PostScript print jobs

[Helge Blischke]

Thomas J Pinkl wrote:
> 
> Helge, thanks for your comments.
> 
> > Thomas J Pinkl wrote:
> >> Here's a link to one such file:
> >>
> >>   http://ftp.surfnet.nl/security/tcpwrappers/tcp_wrapper.ps.gz
> >>
> >> When I try to print it, I get the first page and nothing more.
> 
> On Tue, Nov 16, 2004 at 06:56:25AM -0500, Helge Blischke wrote:
> > Well, I got that file and - look, printing it with CUPS wasn't OK (CUPS
> > 1.1.19).
> 
> What do you mean by "wasn't OK"?
it printed an empty sheet (in duplex mode) and the rest (from the 2nd
page on)
in simplex.
> 
> > After massaging the file to be DSC conformant to a certain minimum, it
> > printed perfectly.
> > Following is a context diff of the hacked file against the original:
> 
>   [SNIP]
> 
> Did you edit the file "by hand" or do you have a program that inserts
> the DSC "comments" appropriately?
By hand
> 
> > You must know that CUPS feeds the PostScript code through the pstops
> > filter to
> > fill in the defaults defined for the respective printer and the PS code
> > that
> > implement the job options (attributes in IPP speech), and that requires
> > a minimum
> > of DSC conformance and page independence to succeed.
> >
> > As in general it is hardly possible to get all PostScript creators to
> > DSC
> > conformance, I'd suggest to convert the weird PS files to PDF (e.g.
> > using
> > Ghostscript) and printing that PDF. With a little extra effort, you
> > could
> > configure CUPS ti do that automagically for you.
> 
> It seems odd to me that CUPS' pstops filter would _require_ DSC
> conformance when modern applications such as Mozilla do not produce
> such PostScript.  Is there a CUPS option to turn off DSC conformance
> checking?
Well, nearly every application I've seen (in the lase some 5 years, I
must admit)
at least *pretend* to generate DSC conformant PostScript. And, with the
various
Linux boxes we've running here, it seems to be true. But, I must admit,
we never
tried things like ditroff here.
> 
> And having to convert PostScript to PDF and then back to PostScript
> seems wasteful at best.  I'd rather leave it as a raw queue.
Well, the only *reliable* method to digest a PostScript program that is
not 
"well-behaved" is to feed it through a PostScript interpreter, e.g.
Ghostscript.
As Ghostscript's pswrite device converts all font stuff to bitmaps but
the 
pdfwrite device is quite reliable, and the pdftops utility from the Xpdf
suite
is very well designed, I'd always prefer the reliable but CPU time
consuming 
method over others.
BTW, with raw printing, you omit all chances to configure your printout
by
job options.

Helge

> 

-- 
Helge Blischke
Softwareentwicklung
SRZ Berlin | Firmengruppe besscom
http://www.srz.de
tel: +49 30 75301-360