[byteio-wg] FW: BOUNCE byteio-wg at ggf.org: Non-member submission from ["Dave Berry" <daveb at nesc.ac.uk>]
Mark Morgan
mmm2a at virginia.edu
Fri Jun 3 12:29:01 CDT 2005
Mail from Dave bounced.
> -----Original Message-----
> From: owner-byteio-wg at ggf.org [mailto:owner-byteio-wg at ggf.org]
> Sent: Friday, June 03, 2005 1:24 PM
> To: owner-byteio-wg at ggf.org
> Subject: BOUNCE byteio-wg at ggf.org: Non-member submission from
> ["Dave Berry" <daveb at nesc.ac.uk>]
>
> >From owner-grdfm-byteio-wg at mailbouncer.mcs.anl.gov Fri Jun
> 3 12:24:22
> >2005
> Return-Path: <owner-grdfm-byteio-wg at mailbouncer.mcs.anl.gov>
> X-Original-To: grdfm-byteio-wg at mailbouncer.mcs.anl.gov
> Delivered-To: grdfm-byteio-wg at mailbouncer.mcs.anl.gov
> Received: from localhost (localhost.localdomain [127.0.0.1])
> by mailbouncer.mcs.anl.gov (Postfix) with ESMTP id 1330D12B50
> for <grdfm-byteio-wg at mailbouncer.mcs.anl.gov>; Fri, 3
> Jun 2005 12:24:22 -0500 (CDT)
> Received: from mailbouncer.mcs.anl.gov ([127.0.0.1])
> by localhost (mailbouncer.mcs.anl.gov [127.0.0.1])
> (amavisd-new, port 10024)
> with ESMTP id 29792-06 for
> <grdfm-byteio-wg at mailbouncer.mcs.anl.gov>;
> Fri, 3 Jun 2005 12:24:20 -0500 (CDT)
> Received: from zappa.nesc.ed.ac.uk (zappa.nesc.ed.ac.uk
> [129.215.30.10])
> by mailbouncer.mcs.anl.gov (Postfix) with ESMTP id B1C6C12B16
> for <byteio-wg at gridforum.org>; Fri, 3 Jun 2005
> 12:24:20 -0500 (CDT)
> Received: from hendrix.ad.nesc.ac.uk (hendrix [129.215.30.11])
> by zappa.nesc.ed.ac.uk (8.13.1/8.13.1) with ESMTP id
> j53HOHg4003552
> for <byteio-wg at gridforum.org>; Fri, 3 Jun 2005 18:24:18
> +0100 (BST)
> content-class: urn:content-classes:message
> MIME-Version: 1.0
> Content-Type: text/plain;
> charset="us-ascii"
> Content-Transfer-Encoding: quoted-printable
> X-MimeOLE: Produced By Microsoft Exchange V6.0.6603.0
> Subject: Why file properties should not be in ByteIO
> Date: Fri, 3 Jun 2005 18:24:17 +0100
> Message-ID:
> <5D8E46F5AA2D8A428667B8787DEF216CE1D252 at hendrix.ad.nesc.ac.uk>
> X-MS-Has-Attach:
> X-MS-TNEF-Correlator:
> Thread-Topic: Why file properties should not be in ByteIO
> Thread-Index: AcVoYRHySOceKPi9TpmVP7ZkUVG0IQ==
> From: "Dave Berry" <daveb at nesc.ac.uk>
> To: <byteio-wg at gridforum.org>
> Received-SPF: neutral (zappa.nesc.ed.ac.uk: 129.215.30.11 is
> neither permitted nor denied by domain of daveb at nesc.ac.uk)
> X-NeSC-Spam-Score: -105.771 ()
> ALL_TRUSTED,AWL,BAYES_00,USER_IN_WHITELIST
> X-NeSC-Sender-Domain: nesc.ac.uk
> X-Scanned-By: MIMEDefang 2.49 on 129.215.30.10
> X-Virus-Scanned: by amavisd-new-20030616-p10 (Debian) at
> mailbouncer.mcs.anl.gov
>
> All,
>
>
> At the London meeting is was proposed that the ByteIO
> interface should include properties such as creation time and
> modification time, which are basically file description data.
> I have two arguments against this proposal. Although we
> should develop a specification for describing files, this
> should be separate from ByteIO.
>
> First, these properties are not specific to the operation of ByteIO.
> Some ByteIO sources will not have them at all. Others may
> have these properties combined with the ByteIO interface, but
> they are not relevant to the actual IO operations.
>
> Second, these properties are needed in other interfaces. For
> example, in a replicated file system these properties may be
> stored in a metadata catalogue. =20
>
> So I think we would produce a better-factored architecture if
> we had two separate, small, specs: one for ByteIO and one for
> file description.
> Where both apply, they can be combined into the same EPR, and
> clients of that EPR will be able to access both.
>
> Dave.
>
More information about the byteio-wg
mailing list