On Tue, 2008-12-16 at 21:13 -0500, Ales Hvezda wrote: > >The cairo branch has seen a great deal of improvement, and I'd love some > >testing feedback. Is it fast enough.. are there any rendering artefacts > >I've missed, have any behaviours changed (which shouldn't have). > > I have updated to the latest head on the cairo branch > (e3d0a2f6a09cce161021a15ba9a3230bfeafa531) > > Most of the ddrawing artifacts are gone that I observed previously. > > Various observed behaviors though: > > - With the dark-bg color scheme, all the grid lines are solid lines.
I need to fix that grid code for dark-bg colour schemes. > - It seems that the bounding box for 90 and 270 degree rotated text > is slightly larger than the text. It is larger than the text in most cases, but if there is a specific problem with 90 and 270 degrees, I'll take a look at it. What we need to do is teach gschem to provide more realistic world bounds for the text, since the metrics of our old font doesn't match the metric of the font pango is using. > - If I zoom in close to the 72 point text in drawing_primitives.sch > gschem consumes a minimum of 200-300M of RAM (and never > decreases). Not sure this is worth worrying about. Probably cairo (or perhaps pango) being "helpful" caching glyphs. It may also be the general trend for glib and other libraries not always to free memory, but to keep it locally as a memory pool. Its a pain, but assuming the memory isn't actively being used, it ought to be paged out to disk. > - cairo gschem doesn't render as fast as gdk gschem on my 1.8GHz > laptop. Not surprising really. It could be a little quicker if I get the line-end hinting right. That keeps the rendering on the fast-path. > - Rendered schematics sure look pretty. ;) > >I've coded up a Pango renderer subclass which draws overbars (although > >it uses a magic constant for positioning them). Does this work for the > >font gschem picks up on your system? Do the overbars look OK? > > The font and overbars look mostly correct on my laptop. Thanks for testing. I feel that the invalidate/expose stuff is almost ready for git HEAD. The cairo code will need some work if we want to make it optional. (It may also want some consolidation of the various patches / fixes applied to the text-rendering path). Thanks for testing! -- Peter Clifton Electrical Engineering Division, Engineering Department, University of Cambridge, 9, JJ Thomson Avenue, Cambridge CB3 0FA Tel: +44 (0)7729 980173 - (No signal in the lab!) _______________________________________________ geda-dev mailing list geda-dev@moria.seul.org http://www.seul.org/cgi-bin/mailman/listinfo/geda-dev
