Subject: Re: disktab(5)
To: der Mouse <port-sparc@NetBSD.org>
From: Don Yuniskis <auryn@gci-net.com>
List: port-sparc
Date: 11/27/2001 17:19:14
Greetings and Gesticulations!

>> Greetings and Consternations!
>
>Y'know, I feel like continuing this conversation just to see what
>inventive words you'll come up with next for that signon line. :-)


<grin>  Yeah, you wouldn't believe how hard it is to try to
come up with pseudo-rhymes -- without repeating them (too often)!

Another interesting "experiment" is to try to come up with
"random" numbers.  Of course, they aren't.  But, if, say, every
hour (or every time you send email, etc.) you ask yourself
to fabricate a "random" number, you will be chagrined to see how
little imagination you truly have!  :-/   (was part of an experiment
that a friend participated in some 30 years ago...)

BTW (further off topic), emails to you (mouse) are meeting with
this warning:

Message delivery to '<mouse@Rodents.Montreal.QC.CA>' delayed
SMTP module(domain rodents.montreal.qc.ca) reports:
Sparkle.rodents.montreal.qc.ca: no response

Note that my ISP is brain-dead so it could very well be something
hosed on this end of the pipe...  :-(

>>> der mouse meowed:
>
>...or verbs for me. :-)


Yet another "randomizing" activity... :-(

>> I think the problem stems from the fact that you can probe an IDE's
>> *native* geometry as well as [its] *current* geometry
>[and NetBSD is apparently a little careless if the BIOS has set them
>different]


I tried two installs on two different i386 boxes here -- 1.5.2 and
1.5.1.  In both cases the installer messed up the drive geometry.

>>> (newfs has options to override the label's geometry info; if you use
>>> enough of them, the label's geometry _is_ ignored.)
>> OK.  So, even when "ignoring" the geometry, it is effectively using
>> command line options to imply a *new* geometry...
>
>Yes - for purposes of that filesystem only.


Ah.  I (now) see the subtlety...^^^^^^

>> Of course, a better solution might be to just not support i386??  :>
>
>The more I see of the BIOS IDE geometry hell compatability with past
>mistakes requires, the saner this option looks. :-(


Perhaps the only "technically interesting" aspect of the x86
is how mightily it has struggled to *try* to remain backward
compatible.  Though I wonder if we would tolerate that same
sort of "compatibility" if all of our automobiles ate hay...

--don