optimize shared caching by configured sizes

[Reimar Bauer]

If we are in a region with a very low bandwidth we may use a local proxy or a shared cache directory

A shared cache needs the same topview and sideview geometry on each client for best results.

This needs changes to the configuration file. we need to set the default size there and to fixate windows to this size.We also could use a gui to configure it, see example .

The example image misses a second proxy url. For now we would only implement a json configuration and keep those settings guis for later.

[Reimar Bauer]

Another aspect I found is that we may need a different size parameter for the server side requested images.

[Reimar Bauer]

this is quite similar to #123

While layout sizes easily can be adjusted the size of the image requested on the server is based on some calculations. The values become retrieved from

in mpl_qtwidget

def getPlotSizePx(self):
        """Determines the size of the current figure in pixels.

        Returns the tuple width, height.
        """
        # (bounds = left, bottom, width, height)
        ax_bounds = self.ax.bbox.bounds
        width = int(round(ax_bounds[2]))
        height = int(round(ax_bounds[3]))
        return width, height

[Reimar Bauer]

initial configuration for given drawing geometry, addresses #190

→ <<cset 8a4c4a6fee33>>

[Reimar Bauer]

In the current state we don't implement that gui. May be reopen later again.