gswin32c.exe -dBATCH -r300 123-05-.ps Produces a bewel join, but CPSI and me believe it must be a miter join.
Created attachment 1425 [details] 123-05-.ps The test file is a reduced Genoa test.
The bug is important for the ps2write testing.
A suggested patch http://ghostscript.com/pipermail/gs-code-review/2005-August/004788.html The commitment is delayed until a review by the PCL team.
Created attachment 1625 [details] raster.zip Attaching rasters made with GS, Adobe CPSI and Adobe Acrobat Reader 4, packed into a ZIP.
Created attachment 1629 [details] patch.txt A suggested patch. It fixes the reported problem, but opens another problem - probe3.pdf renders worse due to add_round_cap appears insufficiently smart.
Created attachment 1630 [details] t2.ps A postscript file used to test CPSI about line join.
See also Bug 688269.
Patch to HEAD : http://ghostscript.com/pipermail/gs-cvs/2005-August/005646.html http://ghostscript.com/pipermail/gs-cvs/2005-August/005647.html