[cairo] abort in postscript when printing

[James Cloos]

>>>>> "AJ" == Adrian Johnson <ajohnson at redneon.com> writes:

JC>> I've noticed that some apps have been crashing of late whenever I try
JC>> to print.  Printing to PDF works find, only Postscript paths fail.

AJ> This commit is the likely cause of the problem:
AJ> http://cgit.freedesktop.org/cairo/commit/?id=badf3229

Indeed, reverting that prevents the assert(3)s.

(Cute.  Of course the odds that a single problematic commit would start
with the word "bad" is only 4096:1 against.  But still. :)

After reverting the postscript ouput still is not correct, though.
Most of the images are much too small and many have bad backgrounds.
The images which are too small look to be underscaled by somewhere
between 1/4 and 1/5; 72/300 is a likely possibility, suggesting
that the main and fallback resolutions get confused at some point.

Oddly enough some of the images do get output correctly. ???

A math-heavy web page such as:

http://en.wikipedia.org/wiki/Exact_trigonometric_constants

is a good example.

-JimC
-- 
James Cloos <cloos at jhcloos.com>         OpenPGP: 1024D/ED7DAEA6