Hello! Thanks for raising this. These long-standing experimental features are less than ideal, and the snapd team has been trying to address this by first understanding what works needs to be done, and then scheduling this work.
We’re currently looking for feedback, which we’ll then use to prioritise and plan the work:
