Subject: Re: HEADS UP: migration to fully dynamic linked "base" system
To: David Maxwell <>
From: Johnny Billquist <>
List: current-users
Date: 08/28/2002 17:29:37
On Wed, 28 Aug 2002, David Maxwell wrote:

> On Wed, Aug 28, 2002 at 10:40:27AM +0200, Johnny Billquist wrote:
> > I'm willing to bet people will not find the problems in /rescue until the
> > time they really need it, and by then, it's a bit late, don't you think?
> When was the last time you ran mknod? Are you _sure_ you have a good
> copy of it? Good heavens, what if it developes a bad block on the HD,
> and mknod dies every time you try to run it... You might have to resort
> to (Horror!) a... boot floppy. eeek! (Thinking of the Solaris rm
> /dev/zero and reboot case)

I *don't* have a floppy on my VAX, so I cannot have a boot floppy.
Thus, this question might be slightly more important to me than to you.
My only alternative is network booting when/if things really turn bad.
That means I'll have to get myself a somewhat recent distrib, unpack in,
get whatever machine I (hopefully) have working on the net setup to
service a network boot, and so on.
I sure hope I have some other machine readily available on the net, or
things start getting even more interesting...

Oh, and I think I last ran mknod the day before yesterday... :-)
But yes, that is not a program people usually runs *that* often. However,
there are actually ways to go around a broken mknod if you just have some
other programs working, which you are using a bit more often.

/dev/zero under solaris? Wasn't that something connected to dynamic
loaded libraries? :-)

> Previously, updating /bin and /sbin was part of an upgrade cycle. Now,
> updating /rescue is part of the cycle (well, doesn't have to be, if you
> choose to keep boot media around instead).

Updating? Yes.
I wasn't talking about updating. I was talking about really knowing that
it was fit for fight. That means testing, using, you know... More than
just building.

> Saying 'X isn't used much, so if X gets broken we wouldn't notice until
> we need X, and then it would be a pain' is about as meaningful as saying
> 'My Windows box needs to be rebooted twice a day, so I always will find
> out about boot problems within 12 hours. NetBSD is worse, since it might
> stay up for years, not being able to reboot, and I wouldn't know!'

That *is* a cause for concern. The fact that you obviously don't agree
is just another proof that we have different views (so what else is new?).
However, when the "repair" tools are programs that I do use every day, I
tend to notice if something bad has happened even though I'm not rebooting
my machines at months.

And I don't use Windows at all, thank you.


Johnny Billquist                  || "I'm on a bus
                                  ||  on a psychedelic trip
email:           ||  Reading murder books
pdp is alive!                     ||  tryin' to stay hip" - B. Idol