There has been some more discussion on that (see http://markmail.org/message/5jraw2mk6k6nvlos). In essence the maintainer seemed to be rather thinking that the joystick should not use the BTM_DIGI. Unfortunately I don't know what the solution should be and the discussion went silent after that.
Probably the best solution would be to have a new bug about the regression with the upstream but linked to that and also to try to approach Jiri and/or Dmitry to ask about the current status on that.
There has been some more discussion on that (see http:// markmail. org/message/ 5jraw2mk6k6nvlo s). In essence the maintainer seemed to be rather thinking that the joystick should not use the BTM_DIGI. Unfortunately I don't know what the solution should be and the discussion went silent after that.
Probably the best solution would be to have a new bug about the regression with the upstream but linked to that and also to try to approach Jiri and/or Dmitry to ask about the current status on that.