ipfs blocks local returns _all_ local refs. For now this is one
long op. future commits will make it async.

fix listener problems

peerstream would hang when it got many temporary errors.
temp errors should not count as an error. Now, it will
only exit when an error is not temporary.

I kept the acceptErr channel because it will no longer
cause a bad hang. The goroutine is exiting, so if it
blocks until acceptErr is read, it's fine. If users
launch tons of listers and see goroutines building up,
they know they should be reading + logging those.

Instead of erroring out, which would break the listener,
we instead log a message and continue. This is not an error,
the internet is a place with lots of probing + connection
failures.

This should handle early breakages, where a failing connection
would take out the listener entirely.

There are probably other errors we should be handling here,
like secure connection failures.

refactor(core): NewIPFSNode constructor

this is a dirty hack

fix: remove merkledag import

may need to squash this commit into the merkledag move commit

NB: all commits (from master) pass tests up to this point.

fix: squashme into use core constructor

early out if no entries in wantlist

Implement ipfs ping

core: make sure to call swarm.FilterAddrs

also move network setup into its own self-contained func

linkblock size calc test