setrofim

Please could you describe the issue in more detail? Which part is failing? What device are you running on? For rooted devices and dev boards, `su` should already be in...

@derkling @joelagnel @bjackman what is the state of this PR? Is this still needed? Are there more changes coming?

Yes, this a know issue (read: "expected behaviour"). The reason the first 32 bytes are taken as part of `__get_memo_id` is to deal with ID reallocation problem. Mutable objects is...

> Only memoize calls involving hashable objects. [...] If that highly matters to the caller, hashable types like tuples() should be used where possible. This what I meant when I...

> I think there's 3 ways to handle this: In the past, we've handled this the second way -- i.e. a separate utility collector running in a thread (e.g. see...

I believe that, on some devices at least, battery status may already be exposed via sysfs and can be collected via existing instrumentation. However, it may be worth adding a...

Cc @douglas-raillard-arm: this should be largely orthogonal to #450, I believe, but FYI, just in case.

> There is also a ConnectionBase class introduced: > d7908d1 > But it should be straightforward to merge the features of both as they don't seem to overlap. Ah cool,...

@derkling @bjackman sorry for the delay in following up on this. Is this still needed? Is it ready for merge? If so, please could you rebase onto tip of master.

> maybe WA already does that? WA has a generic "sysfs_extractor", which takes a list of sysfs locations, which it then pulls before and after each iteration, and then performs...