[Zope3-dev] Space usage of unicode strings in the ZODB

[Tim Peters]

[Andreas Jung]
> Yes, I measuread the final size of the ZODB. Using UTF-8 explains why the
> internal encoding does not matter for the final size of the ZODB.

Cool!  Then everything is understood.  Note that it would be possible to
introduce a SHORT_UNICODE flavor of binary pickle encoding too, to reclaim
the difference in this test between storing Unicode directly and converting
to UTF-8 first.  (I'm not advocating that, just pointing out the
possibility.)