Hi guys,
This is a small patch to fix some typos and add output recording for NEW-PAGE in the postscript backend.
When I tried to run CLIM-DEMO::POSTSCRIPT-TEST, it produced an invalid postscript file. Ghostscript failed to load it, complaining of a dictionary stack underflow. The reason was that the output from NEW-PAGE was at the top of the file, and not embedded between pages as it was meant to be. This is because WITH-OUTPUT-TO-POSTSCRIPT-STREAM uses output recording, (Perhaps this was introduced to make EPS work? Just a guess.), yet NEW-PAGE draws immediately instead of being recorded.
I attempted to add output recording for NEW-PAGE. This patch fixes the postscript test mostly -- page 7 is translated to the lower right corner for some reason, but it's at least syntactically correct. I have no understanding of this stuff, so please check whether I've done something boneheaded.
Thanks, Tim
"Tim Daly Jr." tim@tenkan.org writes:
I attempted to add output recording for NEW-PAGE. This patch fixes the postscript test mostly -- page 7 is translated to the lower right corner for some reason, but it's at least syntactically correct. I have no understanding of this stuff, so please check whether I've done something boneheaded.
I think this works "by accident", because the top-level output record is a sequence output record (rather than some complicated multidimensional data structure output record) and so the new-page-output-records in the replay happen in the same order and at the same points that they happened when drawing.
I could have sworn that I checked that postscript-test worked before checking in, but apparently not, and indeed multipage output could not possibly have worked, so I must have tested an older implementation. Sorry about that. I can't immediately see why a transformation has occurred, and I'm afraid my free seconds have run out, so I've merged the patch.
Thanks,
Christophe
-
Christophe Rhodes -
Tim Daly Jr.