[cairo] Re: Long line test
emmanuel.pacaud at univ-poitiers.fr
Wed Aug 9 15:48:02 PDT 2006
Please find attached the missing reference image.
On jeu, 2006-08-10 at 00:45 +0200, Emmanuel Pacaud wrote:
> You'll find attached a test for the long line bug.
> That's a bug we hit easily in goffice graph when using our cairo
> renderer. (A workaround we could use would be to clamp coordinates to
> something safe (it looks like it's +/-10000, without using
> transformation), but since clamping bound is quite low, that leads to
> incorrect display in some cases. Something more correct would be to clip
> our paths before sending them to cairo).
> A thing that makes this bug annoying is the fact that bounds in which
> cairo output is correct is not predictable, and seems to also depends on
> line width.
> I thought it was due to a bug in cairo_fixed_from_double, where values >
> 32767.0 are converted to -32768, but fixing that bug does not fix the
> long line bug.
-------------- next part --------------
A non-text attachment was scrubbed...
Size: 247 bytes
Desc: not available
Url : http://lists.freedesktop.org/archives/cairo/attachments/20060810/f33a31ed/long-lines-ref-0001.png
More information about the cairo