Skip site navigation (1)Skip section navigation (2)
Date:      Fri, 14 Jun 2002 08:51:33 +0100
From:      Nik Clayton <nik@freebsd.org>
To:        Eric Anderson <anderson@centtech.com>
Cc:        doc@freebsd.org
Subject:   Re: Search engine enhancements
Message-ID:  <20020614085133.V39690@canyon.nothing-going-on.org>
In-Reply-To: <3D08A7DB.8BE28A90@centtech.com>; from anderson@centtech.com on Thu, Jun 13, 2002 at 09:10:35AM -0500
References:  <3D08A7DB.8BE28A90@centtech.com>

next in thread | previous in thread | raw e-mail | index | archive | help

--IjgjYm7ZLtdx4pCv
Content-Type: text/plain; charset=us-ascii
Content-Disposition: inline
Content-Transfer-Encoding: quoted-printable

On Thu, Jun 13, 2002 at 09:10:35AM -0500, Eric Anderson wrote:
> I saw this on the FreeBSD documentation site, as a current project.  Is a=
nyone
> currently working on this? =20

Not that I know of.

> I would love to jump into this project see what we
> can do.  What do I need to do to get started?

Come up with a better search engine interface and backend than what
we're currently using, preferably based on software that's available in
the ports tree so that it's trivial for mirrors to set up.

"better" in this case is subjective, and I can't recall a thread in here
that's really covered a 'wishlist' of requirements for a better search
system.

So lets start one.  Off the top of my head:

The set of search operators is too small.  I'd like to be able
to limit my search to text that appears in the:

	Subject line
	Body text
	From/To address

I'd like to do queries by date, so that I can search for
messages that match only in the last 3 months.

I'd like the bug that makes it flaky when you search more than
three mailing list archives fixed.

Viewing the thread that a message comes from is painful.  Google
have solved this in a particularly nice way -- you can view all
the messages in a thread, in thread order, on a single page,
using their Google Groups interface.

An alternative way of specifying the mailing lists to search
would be nice.  Keep the checkboxes, but give me a box where I
can type in "arch,current,hackers" to limit the search to just
those lists -- I can type that much faster than I can navigate
the mouse over to three fairly small interface elements and
click.  Especially if I have to scroll the screen in order to=20
reach all the checkboxes.

Lose the requirement to specify "AND" and "OR" as connectives in
the query string.  The string
=09
	foo bar baz

should rank messages that feature all three words high in the
results.  Messages that only feature two of them should be a
little lower, and so on.  Maybe use the (fairly) standard
notation

	+foo +bar baz

to indicate that 'foo' and 'bar' are mandatory, and that baz is
optional.

It's not clear when you limit the number of search results how
the limit is done.  Does it just stop when it finds the first
'n' results?  Or does it gather all of them, order them, and
show you the first 'n'.  Better to generate a page of 'n'
results at a time, where the user can specify how many results
they want per page.

When viewing the results, highlight the terms in the search that=20
matched in the text (maybe).

Anyone else?

N
--=20
FreeBSD: The Power to Serve      http://www.freebsd.org/               (__)
FreeBSD Documentation Project    http://www.freebsd.org/docproj/    \\\'',)
                                                                      \/  \=
 ^
   --- 15B8 3FFC DDB4 34B0 AA5F  94B7 93A8 0764 2C37 E375 ---         .\._/=
_)

--IjgjYm7ZLtdx4pCv
Content-Type: application/pgp-signature
Content-Disposition: inline

-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.0.7 (FreeBSD)

iD8DBQE9CaCFk6gHZCw343URAmUpAJ9eHmHni1vWObdtW5eOkqDbcrIHCgCfcH4Z
mxhzuM51l2u+4+HG00JrQt0=
=wXXn
-----END PGP SIGNATURE-----

--IjgjYm7ZLtdx4pCv--

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-doc" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20020614085133.V39690>