(quotation+colon+URL or quotation+comma+whitespace+URL) did not
work properly (thanks to Alastair Burt for reporting and submitting
the patch).

This change includes a minor change to semantics for
reinitialization().

There are three chief performance-related changes:

    - Do not check whether logging is enabled in the actual log call.
      Instead check enviroment variables when the module is imported
      and whenever an explicit initialize() call is made.

      This might seem like it limits the user's ability to turn
      logging on and off, but I don't believe it does.  Logging is
      controlled by environment variables.  Once the Python script is
      working, the environment variables won't change.

      This change makes the case of no logging fast, instead of slow.

    - Each log call goes through a method on the stupid_log_write()
      object instead of through __call__().  It's much faster to call
      a method than an instance.

    - Use "print >> file" rather than "file.write()".  It appears to
      be a bit faster.

Replaced _set_stupid_dest() with _set_log_dest().

a diff-like output (using ndiff) of the difference
between HTML reference and generated HTML

added README

by passing self in.

Warning: this test is incomplete like the former
version because the real regression tests are not working
because the reference HTML files were generated
for the old ST - and not STXNG.

versions floating around. :/

bypassed unlinking of the commit log file.  This broke Full and
Minimal storage. :(

appearance of a leak, particularly noticible with imPermissionRole objects.

    not work to pass in a DBEnv object, and that isn't the right way
    to increase the lock size anyway (should use a DB_CONFIG file as
    per the BerkeleyDB specs).

objects as changed when OFS.DTMLMethod inserted DTMLValidate as self.validate.
This sidesteps the problem by modifying the object's __dict__ instead.

if the object is a dictionary, and do a PyDict_GetItem instead if possible.
This avoids the error machinery overhead if the key is not found in the
dictionary.  PyDict_GetItem returns a borrowed reference, so the refcount
is also bumped in these cases.

using the Indexes tab.

*sigh*

broken.