end.

from ZConfig.

- docstrings

- add comment that we need logging support from the schema still
- revert to using a relative URL for the import, since we're not quite
  using the schema components yet

storage.  This provides a minimal test of the server close
functionality.

Close the trigger when the server exits.

Added new test to make sure this works, and ensure it does not get broken
later.

intent being to make this easier to explain.

- <sectiongroup type="foo"/> becomes <abstracttype name="foo"/>
- <sectiontype type="foo"/> becomes <sectiontype name="foo"/>
- <sectiontype ... group="foo"/> becomes <sectiontype ... implements="foo"/>
- several internal names get changed to use names consistent with the
  new terminology

(Recreating this diff after outdating the wrong revision.) :-(

AF_UNIX addresses can be parsed even on Windows.
This requires code that uses the socket-address values to change.

variables.  Updated docstrings/comments.

(i.e. via ctrl-C) before it has a chance to make the killer connect,
these zeoserver processes will eventually go away (and clean up their
temp files).

- update the sectiontype attributes

XXX The deferred name isn't perfect, but async is already taken.

When a block transaction is stopped because the client disconnects,
log a message at INFO level.

Add a message at BLATHER level when a transaction acquires the commit
lock.

- Cache the translations from severity to level and string in dicts,
  so these are normally done with a single dict.get() call rather than
  a Python function call.

- Extract the PEP 282 level first, and check it against the effective
  level of the logger, so we can take an early exit if the logger
  won't handle the event.

- Got rid of textwrap.  It slows things down, can frustrate grepping
  programs, and wasn't effective unless you were using Python 2.3
  anyway.

- Got rid of the addHandler(NullHandler()) (seems like superstition).

- Got rid of the log_format class variable (YAGNI).

the schema component library.  (The facility itself still needs to be
documented.)

The strategy classes stopped being useful, because the code had been
modified to always use one of the strategies.  The classes were left
behind because editing the code to use a single class was too much
work at the time.

Move all the logic back into ZEOStorage.  Rename non-public methods on
the class to start with _.  Use lambda: rather than getattr()() to
invoke the method that gets delayed waiting for the storage lock.

XXX Need to settle on a good name for this var and make sure the
schema and code agree on it.

- add a simple example of basic usage that should satisfy many
  applications

possibly being removed in the future.

There's now concrete test of recursively defined section types.

XXX Probably not usable until a better schema import mechanism is
implemented.

available or not, and also fixes to all the tests so they won't crap
out or complain if not.

To test whether these storages are avialable (including all package
dependencies), do:

import BDBStorage
if BDBStorage.is_available:
    # okay to use

Also, in BDBStorage/__init__.py do some cross-platform compatibility
for the bsddb module; in Python 2.3 we can just use the built-in
module, but in earlier Pythons we have to use bsddb3.  Now you can
just use "from BDBStorage import db" to get the proper db object.

available or not, and also fixes to all the tests so they won't crap
out or complain if not.

To test whether these storages are avialable (including all package
dependencies), do:

import BDBStorage
if BDBStorage.is_available:
    # okay to use

Also, in BDBStorage/__init__.py do some cross-platform compatibility
for the bsddb module; in Python 2.3 we can just use the built-in
module, but in earlier Pythons we have to use bsddb3.  Now you can
just use "from BDBStorage import db" to get the proper db object.