[ZODB-Dev] Getting rid of ZODB versions (was Re: [ZWeb] Are Versions evil?)

[Jim Fulton]

Dieter Maurer wrote:
> Jim Fulton wrote at 2004-7-15 10:00 -0400:
> 
>>...
>>Given this, I'd like to formally announce the intention to deprecate
>>versions in both Zope and ZODB.  We will formally deprecate versions when
>>we have a way to maintain long-running transactions in Zope.  I have a good
>>idea how they would work and I'd be happy to advise volunteers on how to
>>proceed. Any volunteers? Any objections?
> 
> 
> Versions are nothing else than long running transactions (which the
> additional but not essential feature to be able to join such a transaction).
> 
> Locking might be necessary for long running transactions as
> the longer the transaction the higher becomes the risk of
> conflicts and the more costly a redo of the transaction.

I don't think anyone find versions useful in cases where
locks matter.  With versions, if you get resource contention,
someone loses.

Long-running subtransactions as I'm proposing would rely on mult-version
concurrency control, rather than locks.  In either case, if you have
resource contention, you are probably out of luck.

Jim

-- 
Jim Fulton           mailto:jim at zope.com       Python Powered!
CTO                  (540) 361-1714            http://www.python.org
Zope Corporation     http://www.zope.com       http://www.zope.org