Mir

Code review comment for lp:~raof/mir/1hz-rendering-always

Revision history for this message
Alan Griffiths (alan-griffiths) wrote :

Trying to summarize my understanding from the discussions going on.

1. the class FrameDroppingPolicy is really a SwapBuffersStateObserver with odd method names.

2. FrameDroppingPolicy et alia fail to place all the frame dropping algorithm(s) into the policy object (and BitsOfFrameDroppingPolicy isn't a nice name)

3. Is the policy really about "frame dropping"? Or is it really about how buffers are swapped (with frame dropping an "implementation detail" that has leaked).

~~~~

Given the amount of discussion it is causing I have to ask - why do we need a frame dropping policy? And a factory for the policy?

This uncertainty has the feel of premature generalization: Is "plugging" this particular subset of the logic actually useful for something we need to implement now (or a prerequisite for a near term goal)?

review: Needs Information

« Back to merge proposal