Re: [matplotlib-devel] Significant update to backend_wx

[John H. <jdh...@ac...>]

>>>>> "Jeremy" == Jeremy O'Donoghue <je...@o-...> writes:

    Jeremy> I have attached the updated backend_wx.py. Obviously if
    Jeremy> I've broken wxGTK I'll need to look into it as a matter of
    Jeremy> urgency.

Probably is a GTK2 thing.  Do you have access to a GTK2 machine to
test?

    Jeremy> This sounds worryingly as though you do have the latest
    Jeremy> version (there was no file save dialog code at all until
    Jeremy> this morning).

Yep, same problem with the attached code.

    Jeremy> In particular, the behaviour of a wxMemoryDC (which is the
    Jeremy> context used to draw into the bitmap) seems quite
    Jeremy> different between the two platforms. The main bug I've had
    Jeremy> to deal with is that the bitmap can only be selected into
    Jeremy> a single wxMemoryDC at a time. This means that I have to
    Jeremy> take explicit care to ensure that whenever a
    Jeremy> GraphicsContextWx.new_gc() is called, I explicitly ensure
    Jeremy> that the bitmap is selected from the last wxMemoryDC.

I diagnosed why the print code doesn't save a file.  It hangs
indefinitely on the call to

  gc = self.drawable.new_gc()

    Jeremy> The other thing to look for is in the _onSize and _onPaint
    Jeremy> methods of FigureWx: the image from the bitmap is sent to
    Jeremy> the screen with a Blit() call. This seems to be necessary
    Jeremy> on Windows as the (simpler, but equivalent) DrawBitmap()
    Jeremy> call does not work properly.

Tried Blit versus DrawBitmap in both.  No help

    Jeremy> Where the printing is concerned, again, you could try
    Jeremy> uncommenting the 'self.bitmap.SaveFile()' line, which
    Jeremy> checks whether it is possible to save a 'known good'
    Jeremy> bitmap.

No help, due to the hang on new_gc

Any other ideas?

JDH