[cups.bugs] [MOD] STR #2191: Combination of number-up and page-border does not paint border on 1st page

Kurt Pfeifle k1pfeifle at gmx.net
Tue Jan 23 10:28:49 PST 2007


On Tuesday 23 January 2007 16:24, Michael Sweet wrote:
> 
> [STR Closed w/o Resolution]
> 
> OK, let's start with the easy stuff.
> 
> The file does not conform to the Adobe Document Structuring Conventions:
> 
>     % cupstestdsc 8-page-kword-as-received-by-CUPS.ps
>     8-page-kword-as-received-by-CUPS.ps: FAIL
>         Line 1714 is longer than 255 characters (327)!
>             REF: Page 25, Line Length
>         Saw 8 lines that exceeded 255 characters!
>         Warning: file contains binary data!
>         Warning: obsolete DSC version 1.0 in file!
>         Warning: no %%EndComments comment in file!

Thanks for looking into that!

I've myself only yesterday discovered and started to play with 
cupstestdsc (and a few of the other *test* utilities that build
if you compile from SVN).

Do you have a set of standard test files you use for testing
cupstestdsc against (or something what some developers call "unit
tests")?
           (*) see below

> Also, I see a double %%EndFont comment (something cupstestdsc doesn't
> currently detect) which could indicate more problems.

Yes; I noticed (and noted?) that as well.

> Anyways, the real problem is that there is a background fill on page 1
> that overwrites the page border (which is drawn first). If you look
> closely you'll see part of the border of page 1 is still visible...

I wonder why that background fill is just on page 1 then... I'll run a
few more tests on the weekend with files from other sources. Qt is 
pretty bad and weak when it comes to PostScript generation and all
stuff that is font related (rendering/embedding) -- and I'm pretty
weak myself in being able to understand how all that font stuff works
at all. Maybe I'll find out how to teach me a bit about this during
the next 6 months or so.

> Link: http://www.cups.org/str.php?L2191
> Version: 1.1.23
> Fix Version: None


(*)
I'm trying to come up with a collection of utilities, testfiles
and test cases to test KDEPrint development against (and it'll
hopefully be useful for Gnome and other projects as well).

Also, I've now a locally working version of a "Makefile.kp" that
I include into the main Makefile. It offers various targets for
testing, the main one being "make cups_clone". This will copy
and/or symlink the newly compiled files into a separate directory,
create different config files and a startup script and allow to
run the new server in user space on a user defined port (defaults 
to 23631). A few wrapper scripts will take care of LD_PRELOAD-ing
the correct libraries.

I'm hoping this will help to find half a dozen people (from the 
CUPS lists, and from distro packager and desktop developer ranks) 
who would compile and run CUPS/SVN on a regular basis and help you 
with testing, feedback and bug reports. 

But I'll keep it more or less private if you think this leads more 
to a problem for you and distracts you with worthless reports. Other-
wise, I'll somehow publish and announce it.

So, if you think this is useful, I'd be grateful if you could include 
the "include Makefile.kp" statement (actually, it should be 
a different name) in the standard way of creating the Makefile (and 
it should also create a an empty Makefile.clone or Makefile.custom, 
or whatever the best name would be, so that builds work for everyone 
else too :-).  People would just have to copy my included file in
place and run "make readme; make cups_clone"...

Cheers,
Kurt





More information about the cups-devel mailing list