Skip site navigation (1)Skip section navigation (2)
Date:      Tue, 30 Oct 2001 13:52:14 -0500
From:      Garance A Drosihn <drosih@rpi.edu>
To:        Julian Elischer <julian@elischer.org>, Peter Wemm <peter@wemm.org>
Cc:        Bruce Evans <bde@zeta.org.au>, Nate Williams <nate@yogotech.com>, arch@FreeBSD.ORG
Subject:   Re: 64 bit times revisited..
Message-ID:  <p0510100bb804a2ab3247@[128.113.24.47]>
In-Reply-To:  <Pine.BSF.4.21.0110300949340.26174-100000@InterJet.elischer.org>
References:   <Pine.BSF.4.21.0110300949340.26174-100000@InterJet.elischer.org>

next in thread | previous in thread | raw e-mail | index | archive | help
At 9:52 AM -0800 10/30/01, Julian Elischer wrote:
>One added note however.. I am expecting that there will be embedded
>systems being deployed NOW that will be in service in 2038.

Let us say that is true.  All such systems deployed before this
second in time -> this one here <- will not work in 2038, because
freebsd at this time does not support it.

I have no objection if someone wants to do the work to change UFS to
use unsigned timestamps for that field, but I expect it'll take at
least a few months to implement that and DEBUG it for real.  That
because the only way to really debug such a change is to run your
system with a date which is AFTER 2038, and to do that for a few
months.  That's the only way to get real-world experience of what
will and what will not work if that field is changed to be unsigned.
So, not only will today's embedded systems not work, but it'll be a
few more months (or at least a few weeks) before any 2038-safe
embedded systems can be deployed.  How much longer will it be before
we have UFS2 and a more-permanent solution?  Another year?  Two?
So you are really only gaining a two-year span where embedded
systems will work after 2038, and 2038 is 37 years from now...

Still, if someone wants to do that work, then I'm all for the idea
(honest, I am!).  The only place where I really start objecting is
when we start stealing bits from other fields, and thus spreading
a single value over several different spots in a 'struct'.

Sounds like Julian is volunteering to do the work...   :-)

-- 
Garance Alistair Drosehn            =   gad@eclipse.acs.rpi.edu
Senior Systems Programmer           or  gad@freebsd.org
Rensselaer Polytechnic Institute    or  drosih@rpi.edu

To Unsubscribe: send mail to majordomo@FreeBSD.org
with "unsubscribe freebsd-arch" in the body of the message




Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?p0510100bb804a2ab3247>