Potential issue with enabling and disabling devices

[bparks13]

While testing for #130, I ran into a potential issue that I don't understand fully. If I enable/disable some devices, I can sporadically get the memory buffer to fill up. I've stepped through the code, and cannot find any reason for it in the software, but it could be an issue with how the resets are configured throughout the plugin. Things I have verified:

• Each time an Enable button is pressed, the device it is attached to is configured and the context is reset (via oni_set_opt(ctx, ONI_OPT_RESET, 1, sizeof(oni_reg_val_t)))
• The number of devices is correct when acquiring data (i.e., the FrameReader is not wasting time looping through multiple sets of devices)
• This is not specific to any particular device, and changes if a new headstage is connected
  • If no headstages are connected, and the AuxiliaryIO device is disabled/enabled, this phenomena is pretty reliably triggered, but in the case that a Neuropixels 2.0e headstage is added it is not always true that the memory buffer fills up
  • Note that this does not occur if the AuxiliaryIO device is disabled, this only occurs when it is enabled again after being disabled
• This issue is completely resolved if the disconnect/connect button in the editor is pressed, which makes me think it is related to the resets that occur while connected

This is also possibly related to #117, where changing how the context is handled may fix this as an emergent property.