Mailing-List: contact geronimo-dev-help@incubator.apache.org; run by ezmlm
Precedence: bulk
Reply-To: geronimo-dev@incubator.apache.org
content-class: urn:content-classes:message
MIME-Version: 1.0
Content-Type: text/plain;
	charset="iso-8859-1"
Content-Transfer-Encoding: quoted-printable
Subject: RE: One True Way (TM) of handling configuration files
Date: Fri, 8 Aug 2003 13:41:33 +0300
Message-ID: <DE8281B0DD26A64E8A731430685A69C411E216@MATTI.eur.soneragroup.net>
Thread-Topic: One True Way (TM) of handling configuration files
Thread-Index: AcNdmAzbPBzrEowySbCTz443BcKJfQAACNGw
From: "Lyytinen Joonas" <joonas.lyytinen@cons.sonera.com>
To: <geronimo-dev@incubator.apache.org>

Hi Bill,

There are some good reasons why those files are needed. Firstly, for =
perfomance
and resource management reasons there probably is need to define =
somewhere e.g. how
many instances of a specific entity bean are kept in the memory. =
Secondly, some parts of the CMP
implementation need server specific configuration data, since the J2EE =
spec does not define
the object-relational mapping on specific enough level (i.e. table and =
coulumn names etc.)
(e.g. in weblogic this is defined in the weblogic-cmp-rdbms-jar.xml =
file).  Finally, there
is the question on entity bean read-write behaviour which must/should be =
configurable..

> Call me naive, but... why does a J2EE server *require* a server=20
> specific file?
>=20
> I can think of two reasons:
>=20
>   1. lock-in
>   2. an inadequate spec
>=20
> I think job one is figuring out whether Geronimo *needs* an=20
> implementation specific deployment file before deciding where it=20
> goes. For the most part these non-standard files seem to be a pita.
>=20
> Bill de h=D3ra
>=20
>=20