Mega patches

[Ben Wing]

> It's this latter point that arguing for.  Mega patches should 
> be handled with a different policy.  Hopefully this 
> discussion will bring about something acceptable to all. Is 
> it holding him to a higher standard?  I don't think so, I 
> think that it's just acknowledging the different approach.
> 
> To clarify my thoughts on this some more, I'm arguing for an 
> increase in checkin latency, *not* a serialisation of 
> submissions.  I'm quite happy for Ben to post the contents of 
> his 9 remaining mega patches tomorrow but the trade off must 
> be extra time before the commit to give reviewers time to 
> digest them.  I'll concede that a month is probably too long, 
> but it got the point across.  At a minimum I'd say that the 
> delay should be two weeks, or at least two weekends.

So far there's only been one recent mega-patch, and I waited about 8 days
before checking it in.

The practical problem is that most people don't test things unless it's
checked in.  It's not clear increasing the latency even more would fix this.