groups, images in SVG backend

Hello, I put a patch in the tracker that implements groups in the SVG
backend. Is that the preferred way to submit patches, or should I just
mail them?

I've also been working on images. SVG has an image tag where you can
include a external .png file. I used some code from the GTK backend to
save an image as a png, then included, as shown below.

def draw_image(self, x, y, im):
        Draw the Image instance into the current axes; x, y is the
        upper left hand corner of the image
        rows, cols, s = im.as_str()
        X = fromstring(s, UInt8)
        X.shape = rows, cols, 4
                          has_alpha=1, bits_per_sample=8,
                          width=cols, height=rows)
        try: pa = pb.get_pixels_array()
        except AttributeError: pa = pb.pixel_array

        pa[:,:,:] = X

        gc = self.new_gc()'test.png','png')
        self._draw_rawsvg('<image x=\"%f\" y=\"%f\" width=\"%f\"
height=\"%f\" xlink:href=\"test.png\"></image>' % (x, self.height-y,
cols, rows))

The above works for '' in the examples. The downside of
this approach is that it would make the SVG backend dependent on GTK.
Can anyone think of a better way?

I was thinking of naming the png files 'filename_1.png',
'filename_2.png', etc. for each image in the figure, where
'filename.svg' is the name of the main file. What do you think?