[Spambayes] Outlook addin using bsddb
Alex Polite
m2 at plusseven.com
Fri Feb 7 11:55:44 EST 2003
On Fri, Feb 07, 2003 at 04:46:30PM +1100, Mark Hammond wrote:
> [Resending - appears to have hit the bit-bucket!]
>
> I have just checked in a change to the Outlook plugin that will use a bsddb
> style database if a reasonable implementation can be found. Currently, a
> "reasonable" implementation means:
> * A bsddb3 module can be imported.
Maybe you want need bsddb3. My tests indicate that dumbdbm is faster
then bsddb3. And dumbdbm is all python and included in the standard
library. I was quite surprised by this and put a mail out here to see
if anyone could corroborate it but it didn't seem to spark any
interest.
Here's what I wrote.
alex> I moved from spamcan to spambayes today and wasted a couple
alex> hours profiling hammie.py
alex> <snip> profile.run("spambayes.hammiebulk.main()",
alex> '/tmp/stats') <snip>
alex> I ran this on approximately 2000 messages and aggregated the
alex> stats. The entire run was 496 CPU seconds.
alex> When looking at the profiling information I realized that I
alex> was using dumbdbm, which is supposed to very slow. I
alex> installed bsddb3, rebuilt my db and rerun the profiling
alex> tests.
alex> The entire run was now 520 CPU seconds, a 4.8% increase.
alex> So it seems like "stupid beats smart" goes for speed
alex> optimizations to.
alex> Can anyone corroborate this?
alex
--
Alex Polite
http://plusseven.com/gpg
More information about the Spambayes
mailing list