Re: [htdig3-dev] htdig-3.2.0b1 slower than htdig-3.1.3 ?


Subject: Re: [htdig3-dev] htdig-3.2.0b1 slower than htdig-3.1.3 ?
From: Geoff Hutchison (ghutchis@wso.williams.edu)
Date: Sat Feb 26 2000 - 17:13:28 PST


At 6:22 PM +0100 2/23/00, loic@ceic.com wrote:
> On the other hand it's stupid because the compression is less used and
>therefore less tested. Given the fact that we often say : turn it on, it
>would make more sense to turn it on by default. I'll do that if you think
>it's a good idea.

OK, now that it's turned on, I have a few stats. I should probably
start running the betas on htdig.org for some better testing, but in
the meantime, some entirely unscientific results (with
wordlist_compression on).

htdig-3.1.5 time: 5.245s total DB size 1760K
htdig-3.2 CVS time: 16.537s total DB size 1036K

This is for 200 documents, so it's not a good comparison, but even at
this point it's slower. I'm going to do a run with profiling tonight.

-Geoff

------------------------------------
To unsubscribe from the htdig3-dev mailing list, send a message to
htdig3-dev-unsubscribe@htdig.org
You will receive a message to confirm this.



This archive was generated by hypermail 2b28 : Sat Feb 26 2000 - 17:18:29 PST