Daniel Rall wrote:
> On Mon, 02 Apr 2007, C. Michael Pilato wrote:
>
>> Daniel Rall wrote:
>>>> I'm actually suprised that the entire locking subsystem isn't shared -
>>>> isn't the implementation absolutely identical?
>>> If they once were, they seem to have diverged somewhat. It would be
>>> nice to share what code we can between'em, though.
>> Lock implementations are different. One uses flat files in a hashed
>> directory (FSFS), the other uses BDB tables.
>
> Would it make sense to share the interface? Or would that add
> unnecessary additional complexity?
There's no reason why the two FS lock API implementations are any more
share-able than those for other FS APIs. So, if you can find a good way to
abstract the BDB 'trail' concept so that the two backends can start sharing
more code, knock yerself out. :-)
--
C. Michael Pilato <cmpilato@collab.net>
CollabNet <> www.collab.net <> Distributed Development On Demand
Received on Wed Apr 4 19:20:39 2007