Re: open upgrade/downgrade, NT and UNIX

New Message Reply About this list Date view Thread view Subject view Author view Attachment view

From: RJ Atkinson (rja@inet.org)
Date: 11/18/00-07:02:50 AM Z


Message-Id: <4.3.2.7.2.20001118075827.00ba6720@10.30.15.2>
Date: Sat, 18 Nov 2000 08:02:50 -0500
From: RJ Atkinson <rja@inet.org>
Subject: Re: open upgrade/downgrade, NT and UNIX


	From a user/operator perspective, it would be better 
to have limited capabilities on a new OS platform (that is
broken wrt NFS) than to change things such that other OS 
platforms (that well work today and for many many years) 
might have increased chance of operational problems.

	I'm not sure how to translate the above into the gory
details, so I won't try and leave that as an exercise to 
the smarter folks here.

	Generally speaking, NFSv4 looks gratuitously more
complex than NFSv3 to me.  From the sidelines, it appears
this was driven by a desire to support an OS that doesn't
really support the right kinds of semantics for NFS.  Much
though I prefer NFS to SMB, it really might have been better
to leave that other OS with a somewhat limited NFS capability 
and keep the protocol simpler for the existing installed base.
I really fear having to troubleshoot byzantine corner cases
operationally that exist in NFSv4 but don't exist in NFSv3
(with existing OS platforms).

Regards,

Ran
rja@inet.org


New Message Reply About this list Date view Thread view Subject view Author view Attachment view

This archive was generated by hypermail 2.1.2 : 03/04/05-01:48:24 AM Z CST