From: Peter Brooks <peter.h.m.brooks@gmail.com>
Subject: Re: B-tree performance
Date: Sat, 31 Aug 2013 08:47:01 -0700 (PDT)
Date: 2013-08-31T08:47:01-07:00 [thread overview]
Message-ID: <ec0c8433-fe61-482b-9b30-59c4eafbc23e@googlegroups.com> (raw)
In-Reply-To: <cc680b29-51d8-4fea-ab7d-e893aa5064e5@googlegroups.com>
On Saturday, 31 August 2013 17:26:41 UTC+2, Shark8 wrote:
> I think that the correct optimization here would be highly dependent on the information you're dealing with. If, for example, you're handling a dictionary or phone-book, you could index off of the first letter [or two] and use that as your starting-range's endpoint.
>
Yes, that's true.
I'll be working with the hash codes of RDF URIs, which should be uniformly dense - well, actually, they will be because I'll be experimenting with hash functions to find one that is.
I suppose that I'll just have to try a few options an time the result.
next prev parent reply other threads:[~2013-08-31 15:47 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2013-08-31 7:14 B-tree performance Peter Brooks
2013-08-31 15:26 ` Shark8
2013-08-31 15:47 ` Peter Brooks [this message]
2013-09-01 4:57 ` Niklas Holsti
2013-09-01 6:05 ` Peter Brooks
2013-09-01 8:05 ` Niklas Holsti
2013-09-01 8:38 ` Peter Brooks
2013-09-02 4:28 ` Niklas Holsti
2013-09-03 2:15 ` Peter Brooks
2013-09-05 8:25 ` Memory mapping and network file systems (Was: B-tree performance) Jacob Sparre Andersen
2013-09-05 10:12 ` Peter Brooks
2013-09-03 16:14 ` B-tree performance Dan'l Miller
2013-09-03 17:43 ` Peter Brooks
replies disabled
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox