From owner-freebsd-doc Sun Aug 9 01:41:05 1998 Return-Path: Received: (from majordom@localhost) by hub.freebsd.org (8.8.8/8.8.8) id BAA26302 for freebsd-doc-outgoing; Sun, 9 Aug 1998 01:41:05 -0700 (PDT) (envelope-from owner-freebsd-doc@FreeBSD.ORG) Received: from baerenklau.de.freebsd.org (baerenklau.de.freebsd.org [195.185.195.14]) by hub.freebsd.org (8.8.8/8.8.8) with ESMTP id BAA26297 for ; Sun, 9 Aug 1998 01:41:03 -0700 (PDT) (envelope-from w@panke.de.freebsd.org) Received: (from uucp@localhost) by baerenklau.de.freebsd.org (8.8.8/8.8.8) with UUCP id KAA02592; Sun, 9 Aug 1998 10:40:39 +0200 (CEST) (envelope-from w@panke.de.freebsd.org) Received: (from w@localhost) by campa.panke.de (8.8.8/8.8.8) id WAA07306; Sat, 8 Aug 1998 22:26:04 +0200 (MET DST) (envelope-from w) Message-ID: <19980808222602.A6235@panke.de> Date: Sat, 8 Aug 1998 22:26:02 +0200 From: Wolfram Schneider To: Jason Evans , freebsd-doc@FreeBSD.ORG Subject: Re: Search engine on web site flakey. References: Mime-Version: 1.0 Content-Type: text/plain; charset=us-ascii X-Mailer: Mutt 0.93.1i In-Reply-To: ; from Jason Evans on Fri, Aug 07, 1998 at 03:51:06PM -0700 Sender: owner-freebsd-doc@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org On 1998-08-07 15:51:06 -0700, Jason Evans wrote: > I have experienced poor results with the > http://www.freebsd.org/search.html search engine. In particular, I have > used it many times over the past couple of years to look for information > in the mail archives. For the last 9 months I have been subscribed to > many of the lists, and I specifically remember email messages that I can > not get the search engine to find. Here are the problems I think I'm > seeing: > > 1) Threads seem to be disjoint. If i find a message from a thread, I > can't find all of the messages that originally belonged to the thread. Not all E-Mail clients support threads. Without the mail header field References or In-Reply-to it is hard to create threads. > 2) Searches apparently don't include the mail headers in the search. This > makes a whole slew of searches impractical. I doubt. > 4) Limiting the number of results apparently causes the search engine to > stop as soon as it finds N results, then sort them. Correct. You are always searching by 'score' and the cgi script re-sort the N results by score, date, subject or author. 'sort by' changes the presentation of the results, not the search engine. > Instead, all matches > should be found, sorted, then the first N displayed. We don't have the resources for a better search engine. > I'm guessing that 4) is the root of most of the evil I'm experiencing. Well, you can set the number of results N to a high value (100 or 500). Wolfram To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-doc" in the body of the message