On Sun, Oct 17, 2010 at 12:38:10AM -0400, Brad wrote:
> The following diff disables setting the XCB backend as higher
> priority over the mature Xlib backend. The XCB backend is
> marked as experimental and unsupported for a reason! It is
> full of bugs. This isn't supposed to be enabled and we're
> the only OS I can find that does so. Not surprisingly this
> resolves the various screen corruption, images not displaying
> at the right size, URL bar highlighting looking weird and
> other weird rendering problems seen with Firefox and Midori
> and poetnailly other applications.
> 
> This does not resolve the animated GIF issue with Firefox.
> That is a separate and unrelated issue.
> 
> I bumped the major rev because it looks like there is
> an API change and I'd rather be on the safe side.
> 
> 
> Eric, stop enabling stuff that is clearly marked
> experimental and unsupported especially when you're
> not sending this stuff out for extensive testing.

Second rev.

I meant to include a patch to fix a performance related
issue with cairo which disables the use of HW gradients
since almost all of the device drivers are broken.


Index: Makefile
===================================================================
RCS file: /home/cvs/ports/graphics/cairo/Makefile,v
retrieving revision 1.28
diff -u -p -r1.28 Makefile
--- Makefile    11 Oct 2010 08:09:44 -0000      1.28
+++ Makefile    17 Oct 2010 03:48:48 -0000
@@ -3,7 +3,8 @@
 COMMENT=               vector graphics library
 
 DISTNAME=              cairo-1.10.0
-SHARED_LIBS=           cairo 10.0
+REVISION=              0
+SHARED_LIBS=           cairo 11.0
 SHARED_LIBS+=          cairo-trace 0.0
 SHARED_LIBS+=          cairo-gobject 0.0
 SHARED_LIBS+=          cairo-script-interpreter 0.0
@@ -13,7 +14,7 @@ HOMEPAGE=             http://cairographics.org/intr
 
 MAINTAINER=            Eric Faurot <[email protected]>
 
-WANTLIB=               GL X11 X11-xcb Xau Xdamage Xdmcp Xext Xfixes Xrender \
+WANTLIB=               GL X11 Xau Xdamage Xdmcp Xext Xfixes Xrender \
                        Xxf86vm drm expat fontconfig freetype glib-2.0 \
                        gobject-2.0 gthread-2.0 m pcre pixman-1 png \
                        pthread-stubs xcb xcb-render xcb-shm z
@@ -40,9 +41,7 @@ CONFIGURE_ENV=                PTHREAD_LIBS="-pthread"
 CONFIGURE_ARGS=                --enable-gl \
                        --enable-xml \
                        --enable-xcb \
-                       --enable-xlib-xcb \
                        --disable-silent-rules
-
 
 REGRESS_DEPENDS=       :ghostscript-*:print/ghostscript/gnu
 
Index: patches/patch-src_cairo-xlib-display_c
===================================================================
RCS file: patches/patch-src_cairo-xlib-display_c
diff -N patches/patch-src_cairo-xlib-display_c
--- /dev/null   1 Jan 1970 00:00:00 -0000
+++ patches/patch-src_cairo-xlib-display_c      16 Oct 2010 21:30:28 -0000
@@ -0,0 +1,15 @@
+$OpenBSD$
+--- src/cairo-xlib-display.c.orig      Sat Oct 16 13:08:38 2010
++++ src/cairo-xlib-display.c   Sat Oct 16 13:08:56 2010
+@@ -353,11 +353,7 @@ _cairo_xlib_device_create (Display *dpy)
+     /* Prior to Render 0.10, there is no protocol support for gradients and
+      * we call function stubs instead, which would silently consume the 
drawing.
+      */
+-#if RENDER_MAJOR == 0 && RENDER_MINOR < 10
+     display->buggy_gradients = TRUE;
+-#else
+-    display->buggy_gradients = FALSE;
+-#endif
+     display->buggy_pad_reflect = FALSE;
+     display->buggy_repeat = FALSE;
+ 
Index: pkg/PLIST
===================================================================
RCS file: /home/cvs/ports/graphics/cairo/pkg/PLIST,v
retrieving revision 1.8
diff -u -p -r1.8 PLIST
--- pkg/PLIST   11 Oct 2010 08:09:44 -0000      1.8
+++ pkg/PLIST   17 Oct 2010 03:05:25 -0000
@@ -38,7 +38,6 @@ lib/pkgconfig/cairo-ps.pc
 lib/pkgconfig/cairo-svg.pc
 lib/pkgconfig/cairo-xcb-shm.pc
 lib/pkgconfig/cairo-xcb.pc
-lib/pkgconfig/cairo-xlib-xcb.pc
 lib/pkgconfig/cairo-xlib-xrender.pc
 lib/pkgconfig/cairo-xlib.pc
 lib/pkgconfig/cairo-xml.pc

-- 
This message has been scanned for viruses and
dangerous content by MailScanner, and is
believed to be clean.

Reply via email to