From owner-freebsd-questions  Sun Mar  2 18:24:33 1997
Return-Path: <owner-questions>
Received: (from root@localhost)
          by freefall.freebsd.org (8.8.5/8.8.5) id SAA21745
          for questions-outgoing; Sun, 2 Mar 1997 18:24:33 -0800 (PST)
Received: from dyson.iquest.net (dyson.iquest.net [198.70.144.127])
          by freefall.freebsd.org (8.8.5/8.8.5) with ESMTP id SAA21740
          for <freebsd-questions@freebsd.org>; Sun, 2 Mar 1997 18:24:29 -0800 (PST)
Received: (from root@localhost) by dyson.iquest.net (8.8.4/8.6.9) id VAA07156; Sun, 2 Mar 1997 21:24:14 -0500 (EST)
From: "John S. Dyson" <toor@dyson.iquest.net>
Message-Id: <199703030224.VAA07156@dyson.iquest.net>
Subject: Re: Lite-2 question
To: plm@xs4all.nl (Peter Mutsaers)
Date: Sun, 2 Mar 1997 21:24:14 -0500 (EST)
Cc: freebsd-questions@freebsd.org
In-Reply-To: <87pvxi38k0.fsf@plm.xs4all.nl> from "Peter Mutsaers" at Mar 2, 97 11:14:07 am
X-Mailer: ELM [version 2.4 PL24 ME8]
MIME-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Sender: owner-questions@freebsd.org
X-Loop: FreeBSD.org
Precedence: bulk

My responses are FWIW, and my opinion of expected progress for Lite/2
integration.

> 
> Could someone tell me what the Lite-2 merges bring? Does it for
> example mean that things like the union FS or LFS will work? Or are
> there other improvements?
>
It mostly brings our code up to the level of the "latest."  This will
make it more worthwhile to invest in future development on our codebase.
In a general sense, it "enables" our further development regarding
better support of filesystem layering and also has some support for
finer grained SMP locking.  It will not immediately give it to us
though.

> 
> I'm doubting what I should do: Use 3.0-current (but wait a while until
> the Lite-2 merge has stabilized a bit) or use 2.2 (when Lite-2's
> improvements are not of great use for me).
> 
For approx 1mo, I would not use Lite/2 in production of any kind.  You will
likely not see major improvements beyond where 3.0-current was before the
initial merge for about that timeframe.  I am still using a 3.0-current
kernel from early Feb for my running system.  I will likely move over to
the Lite/2 merged stuff for most of my development in about the 1mo timeframe.

The kernel is "pretty stable" right now, but still needs work, but there
is no sense in putting oneself through unnecessary pain (IMO.)

John