Content-Type: text/plain;
  charset="iso-8859-2"
From: Josh Berkus <josh@agliodbs.com>
Reply-To: josh@agliodbs.com
Organization: Aglio Database Solutions
To: depesz@depesz.pl, pgsql-sql@postgresql.org
Subject: Re: tree structures in sql - my point of view (with request of
	comment from joe celko)
Date: Tue, 3 Sep 2002 11:17:38 -0700
References: <20020903132207.GC24879@depesz.pl>
In-Reply-To: <20020903132207.GC24879@depesz.pl>
MIME-Version: 1.0
Content-Transfer-Encoding: quoted-printable
Message-Id: <200209031117.38107.josh@agliodbs.com>


Hubert,

> i recently spent some time on tree-structures in sql.
> i started with simple id/parent_id approach, used by nearly everyone,
> then i stopped at joe celko's nested sets, but i found it not very
> usable.

I'll be posting an article on implementing nested set trees "real soon now".

My experieince:

Adjacency list trees are easier to understand conceptually, there are more=
=20
tools on freshmeat.net for them, and they are the most efficient form of tr=
ee=20
for graphical display.

Nested Set trees are hard to wrap your mind around, lack a lot in the way o=
f=20
code samples on freshmeat, are harder to build GUI tools for, but are much,=
=20
much faster for determining branch membership and branch parenthood.

So which model you use depends on what you intend to do with the tree.


--=20
-Josh Berkus
 Aglio Database Solutions
 San Francisco