This is required (or maybe strongly implied) by the IStorage interface.

Also, rearrange the order of operations in _process_invalidations.

Each of these changes avoid a potential race when a connection starts
a transaction while invalidateTransaction is being called. Basically,
we don't want a connection to get a lastTransaction corresponding to
invalidations it hasn't processed.

This neither fixes nor provokes any test failures. :) The need for
this change was discovered via code inspection. It's hard to see how
to test the race without getting being insanely whitebox.

To avoid an error on tearDown.

Fix #22

Adding osx back.

Also adding back -j99, in hopes of getting quicker runs and making
travis nor hate us so much.  But then, we'll probably get more
failures. Maybe the best option will be to just make test pass even
when starved. :(

traviiiiiiiiis

A test was counting on a cache being written to, but it wasn't because
the storage got an invalidation later than expected (on suitably slow
machines, like travis and an AWS dev machine of mine).

that was making a test failure hard to interpret.

Load uses load before

  with regard to the load lock.  This allowes ZEO to work with the
  upcoming ZODB 5, which used loadbefore rather than load.

  Reimplemented load using loadBefore, this testing loadBefore
  extensively via existing tests.

Drop support for Python 2.6 and 3.2.

- 2.6 is long out-of-maintenance, and a security quagmire.

- 3.2 cannot be tested on Travis.

Add support for PyPy.

Let checkQuickVerificationWith2Clients use a longer timeaut and raise an assertion error if the timeout fails. Trying to debug intermittant failures on Travis, seen in all versions of Python.

Per @mgedmin, use a renormalizing pattern to fix the unicode string matchin in Python3. Also just print the data for the storage we expect, because the missing u prefixes change the spacing when pprint nests dicts and things don't match under Python2.

Fix the ruok protocol under Python3: a bytes/unicode problem. However, the doctest cases still don't pass because they have leading u prefixes in them.

One more place to account for the socket.error switchero on Python3. Still doesn't fix nagios/ruok though.

Make the nagios metrics come back in a deterministic order. This fixes the nagios test under PyPy, but they still fail under Python 3 for the same reason test_ruok does.

test_ruok was hanging under Python3. Now it doesn't hang but just fails with 'struct.error: unpack requires a bytes object of length 4'