Code review comment for lp:~alan-griffiths/mir/remove-BufferInitializer

> Its one of the older classes in the code, and I agree that it hasn't gotten
> much useful customization and use, so I'm okay with its removal. This does
> break platform api, not sure what to do about that (sometimes we increment,
> sometimes we don't). I'd guess because this doesn't reach very far outside of
> the mir project we shouldn't increment. Apart from that issue though, approve.

We do need to sort out the platform API (e.g. publish the headers needed to implement a driver platform). But I think that's for another MP.