Mail Archives: geda-user/2012/10/17/15:10:54
On Wed, Oct 17, 2012 at 09:35:04PM +0400, Vladimir Zhbanov wrote:
> On Wed, Oct 17, 2012 at 10:13:19AM +0200, Gabriel Paubert wrote:
> ...
> > Unfortunately, for example the recent gsch2pdf code relies
> > on cairo 1.10 at least, which is not yet in debian-stable.
> > I also think that I hit a bug (probably in libgeda) on a
> > non-x86 machine: the output pdf scale depends on the directory
> > in which symbols are found. It looks like a wild pointer/memory
> > corruption, but I won't be able to debug more before early
> > November.
>
> As far as I can see gsch2pdf uses the same output page size for
> different schematic sizes, so if your symbols have different sizes you
> get the different scales.
No, it's a different bug: simply moving back and forth
a symbol file between the current directory and a directory
completely changes the scale of the output file (by a factor
a several hundred or several thounsand, I can't remember).
In the case I hit the bug, the frame of the titlepage occupies
at most 1mm˛ on the paper, and most of the page is white. I've traced
it to cairo_recording_surface_ink_extents() returning nonsense,
but could not go further.
> BTW, gschem can also print all your schematics from command line at once
> (blinking only once) if you change the print.scm a bit. Moreover, it can
> print EPS from command line. And if you pipe its output to a certain
> program you can get the PDF file, too ;)
I'm aware of it, but I also hit a bug in it (locally fixed).
Gabriel
- Raw text -