I'm trying to get printing working with cairo by rendering into an image surface
and then printing that. It works, except that fonts are 6-7 times larger than
they're supposed to be. I've been poking around the cairo-win32-font.c code for
some time and am coming up empty handed figuring out where this scale factor
comes from. Drawing to an image surface when not printing works fine, so it's
not just all image surface font usage going wrong.
Any hints as to where to look from those familiar with that font backend?
I don't know much about cairo-win32-font.c specifically, but I'm trying to at
least understand this bug report, and so far I'm confused.
If you've got an image surface and you're printing it, that image surface just
has a bunch of pixels in it. There's nothing special about pixel values that
come from text vs. any other pixels. So if "drawing to an image surface when not
printing works fine" I can't see clearly what could be a cairo bug here.
Perhaps I'm missing something though.
And I'd be glad to see a minimal test case to resolve my confusion.
That would actually be preferred over any explanation.
Any updates here? Does this bug still happen? Test case maybe?
Bugzilla Upgrade Mass Bug Change
NEEDSINFO state was removed in Bugzilla 3.x, reopening any bugs previously listed as NEEDSINFO.
A 6-7x scale factor is reminiscent of scaling up to the dpi of the printing surface. Without a test case, my gut feeling is that is an application bug. With 1.6, Adrian Johnson introduced the win32-printing surface hopefully making the manual creation of images for printing redundant.
Please reopen if you do have a test case that we can examine.