Subject: Re: major hier(7) overhauls?
To: Greg A. Woods <>
From: Todd Vierling <>
List: tech-userlevel
Date: 01/26/1999 08:32:04
On Mon, 25 Jan 1999, Greg A. Woods wrote:

: I have lots more, lots older, hardware than most folks.  I've even got a
: SPARCstation-1 running happily with a 105MB drive (sans compiler and a
: few things like documentation, etc.),

Well, I have a 386 as a router with a *30MB* drive,
and a SS2 with a 52MB drive,
and an Amiga 3000 with one 52MB and one 80MB drive,
and soon a DECstation 5000/25 with a 60MB drive,
and a Shark with a 40MB internal drive,
all getting /usr/share and most /usr from my server with the only 1.2GB
drive (on a Cyrix 5x86 system) I have here.  So the system could conceivably
fit on some of the disks above, but you still have to subtract swap....

So I can't see how damn well you can claim to be retrocomputing with that
105MB drive.  I don't shell out equipment cash if I can avoid it; almost all
of the above was nearly free to free.

: On really small systems it does make sense to put stuff on a separate
: spindle, *if* you have one, and not everyone does.  If you have only one
: small spindle than it's often better to make as few separate filesystems
: as possible.  Note that on drives that small multiple FFS filesystems
: waste many kilobytes of space.

Then don't make / and /usr a separate filesystem.  I'm not forcing you to do
so.  However, /usr as a directory WILL NOT go away.

: I once tried to use a 38MB SCSI drive I have with one of my Sun-3's.
: Once upon a time this drive was "enormous" in comparison to other drives
: I had, but on the Sun-3, even with all the SCSI driver speed-ups, it was
: still too slow to use -- about 3 times slower than the network.
: Sometimes small systems just are not practical NetBSD hosts no matter
: how much we might like them to be.
: Don't try to tell *me* about old and small systems.  I have many perfect
: examples and my scheme would work fine on all of them.

You just shot down a perfectly useable setup--one almost *identical* to that
which a friend of mine has--and then claim to know what `perfect examples'
of small systems are?

`Stop making claims you can't back up and come back when you have a little
more credibility.'

-- Todd Vierling (Personal; Bus.