Refactor BlockingKernelClient

[davidbrochart]

There is a lot of code duplication between AsyncKernelClient and BlockingKernelClient. BlockingKernelClient's methods could be wrappers around AsyncKernelClient's methods, which would run the later until complete.
What do you think?

[MSeal]

I think that'd be a good idea, but maybe let's hold off making this low level infrastructure do that until 3.5 gets dropped? I'm a little worried with the hangs for 3.5 we saw in nbconvert after async change -- that one is doing a major version bump so I'm more ok with it but here we might want to wait for a major version change / 3.5 being gone before taking that risk on?

[davidbrochart]

Sure, that sounds reasonable.

[blink1073]

@davidbrochart is there any more work to be done on this front for 7.0?

[davidbrochart]

No, it has been done in #623. Closing!