[Pdx-pm] CPAN

Paul Rogers paulgrogers at fastmail.fm
Fri Jan 8 15:58:48 PST 2010

> Your system is your business, but if you want a lean system that has
> modern amenities, NetBSD and FreeBSD can probably deal with your

FWIW, I bought my 486 in '91 with the intent of running BSD-386 on it. I
never found a source, the first *nix it saw was RHL6.1 "Publisher's
version", Cartman, out of a book.

> hardware a lot better than a recent 2.6 kernel and environment. I'll
> leave this topic alone after this email, but I wanted to exclaim that
> you have a more options if you wish to explore them.

Yes, I'm aware of that.  Got a DVD's here from LXF with a variety of
distros.  Still, you almost quoted LFS' motto: Your system, your rules.
The advantage the LFS route has given me is knowledge of how a system
goes together.  (I back-ported the toolchain 10 years when I found a CD
with most of the software on it, and it runs everyday on my 18 year-old
486/33.  I hate to see a good horse put down when it's still healthy.  
The thing is, I could do that with OSS!)  Linux, after all, is just a 
kernel, much of the stuff it runs is GNU.  I don't even like big kernel
systems, as I never liked DOS, nor the 8080 in my IMSAI.  Sometimes
design elegance just gets overwhelmed by ubiquitous.

> I have to wonder about the electricity cost vs a $40 ~1GHz system from

Meaning for the 90% of think time it's waiting for me to press a key?
Linux turns the CPU off most of that time, but the kind of inefficiency
you're talking about comes in large part from other parts of the system
that are on all the time, whether in my old 233's or the newer salvage.

> freegeek.  Old power supplies were terribly innefficient (not that

Old?  You mean the linear in my IMSAI that had a 15# transformer in it,
capacitors the size of soup cans?  ;-)

> some newer ones aren't) and the old CPU does a lot less computation
> per kWh.

True enough.  But, as I'm sure you know, it's not a simple issue.  I'm
told by a buddy who works at a well-known local manufacturer and would
be in a position to know, the 1.4GHz Tualatin I'm running at the moment
works about as well as a 2GHz P4.  We've had some spirited discussions
about just what kind of computing power the average user needs.  An Atom
would probably do.

> (Meanwhile, pondering trade-off of deploying some really inefficient
> code vs the carbon footprint of writing a replacement.)

Wrong question, IMNSHO.  Which is more "robust" during it's lifecycle?
Paul Rogers
paulgrogers at fastmail.fm
Rogers' Second Law: "Everything you do communicates."
(I do not personally endorse any additions after this line. TANSTAAFL :-)


http://www.fastmail.fm - A fast, anti-spam email service.

More information about the Pdx-pm-list mailing list