- Conflict errors didn't invalidate ZEO cache entries.

- If a ZEO client process was restarted while invalidating a ZEO cache
  entry, the cache could be left in a stage when there is data marked
  current that should be invalidated, leading to persistent conflict
  errors.

- Corrupted or invalid cache files prevented ZEO clients from
  starting. Now, bad cache files are moved aside.

- Invalidations of object records in ZEO caches, where the
  invalidation transaction ids matched the cached transaction ids
  should have been ignored.

than before).  It's possible that process shutdown after updating
lastTid but before invalidating the current record could explain some
problems we've seen on restarts.

(This change is made irrelevent by later changes.)

- BTree sets and tree sets didn't correctly check values passed to
  update or to constructors, causing Python to exit under certain
  circumstances.

- Updating blobs in save points could cause spurious "invalidations
  out of order" errors.  https://bugs.launchpad.net/zodb/+bug/509801

(Thanks to Christian Zagrodnick for chasing this down.)

- BTree sets and tree sets didn't correctly check values passed to
  update or to constructors, causing Python to exit under certain
  circumstances.

cause a ZEO server to stop accepting connections, due to a failure
to catch errors in the initial part of the connection process.

The failure to properly handle exceptions while accepting
connections is potentially problematic on other platforms.

Fixes: https://bugs.launchpad.net/zodb/+bug/135108

https://bugs.launchpad.net/zodb/+bug/435547

set object state when objects implemented _p_invalidate methods
that reloaded ther state (unghostifiable objects).

https://bugs.launchpad.net/zodb/+bug/428039

set object state when objects implemented _p_invalidate methods
that reloaded ther state (unghostifiable objects).

https://bugs.launchpad.net/zodb/+bug/428039

install and use the mkzeoinstance script. :)

optimized Python. :(

Added back the second test condition, otherwise the test wasn't failing on some platforms without the actual code change

https://bugs.launchpad.net/zodb/+bug/533015

Much thanks to Hanno Schlichting and Nikolay Kim for coming up with
the patch.

See https://bugs.launchpad.net/zodb/+bug/459889

A ZEO threading bug could cause transactions to read inconsistent
  data.  (This sometimes caused an AssertionError in
  Connection._setstate_noncurrent.)

passed a tid older than anything in changes but newer than
anything in base.

test into an automated test, and included Chris Wither's repozo fix
(as monified by Godefroid Chapelle) to avoid a deprecation warning.

Bug fixed:
- Internal ZEO errors were logged at the INFO level, rather
  than at the error level.

- zope.testing was an unnecessary non-testing dependency.

105888 when there is an adequate test.  This can be achived through a
pretty straightforward conversion of the existing manual repozo test
to an automated test.