Duplicate frames close when writing on tablet

lylethorngage commented 3 years ago

I report a problem that occurred recently on rmview for rM2, although it seems to be happening also for rM1 (but not as severely).

After duplicating a frame, the duplicate is often closed by the very act of writing or tracing the pen on the tablet. The rubber tip seems to also trigger this. This does not happen very regularly: sometimes I can go on and keep writing with even more than one duplicated frames, but the frames anyway last for not more than a few minutes, before they all close due to a pen stroke.

It is very strange because it only appeared recently. Last week I could clone frames and keep writing without issues for hours. I do not remember updating anything on my machine, and this happened before the problematic update to rM 2.6.

I take advantage of this to suggest that upgrading the frame-cloning feature to target a bipartite or tripartite window would be very useful. For example one could select to have rmview show a number of windows (labeled by an index) at all times. When cloning a new frame the user could then direct the frame to window number j. Just an idea.

Thanks!

bordaigorl commented 3 years ago

Hi, I cannot really reproduce the issue, but the only thing I could think could go wrong is that there is some unwanted sharing of the image buffers. Could you try replacing the cloneViewer function at src/rmview/rmview.py line 342 with the following code:

  def cloneViewer(self):
    img = self.viewer.image()
    img = QPixmap.fromImage(img)
    img.detach()
    v = QtImageViewer()
    v.setImage(img)
    v.show()

To try it out you will need to reinstall rmview with pip install .. If you are unsure what this all means let me know (including what installation method and OS you used) and I'll try to help.