Hi James:
Sloopy is correct regarding his comments on the Performa; you may wish
to follow our exchange on this list.
An enterprising engineering student could take it upon himself to turn
a potentially unworkable project around to his advantage. You could
write your own C code for the kernel, build your own Nubus driver,
compile that kernel and you would have then a very specific kernel for
that machine. It's not impossible, we are talking about Linux after
all; Linux will run on anything. The problem has been of course that
the state of most public schools in the US is such that they are
stronger on anything but serious science and/or math.
James, you don't have to be an engineering student to get this done.
The discipline from such training would be helpful, of course, but you
could do this on your own. If your teachers were well grounded in
science, physics and electronics that would be great -- but even if all
you have available is a part-time librarian/clerk or a local library
open twice a week -- this can still get done.
The difficulty is really about how much you really want the features
you are looking for to exist where you are. Linux is very much about a
"build it yourself" attitude; waiting for "someone else" to write or do
it doesn't always work as a plan. Every program is not available for
everybody to do whatever they envision doing; however the foundation
for building whatever anyone could possibly want from a computer IS
within Linux and the computer languages it uses.
Perhaps it would be helpful to consider Linux like, the childhood game
of Lego blocks. It is up to you what shape these blocks have
eventually after you put them together. In that sense, programming in
Linux -- whether drivers, applications or kernels -- is very much the
same. This is a lot for any newbie, but for someone with determination
and focus it is not a major limitation. This is the kind of effort
very much worth recognition as a scientific project good for students
at all levels and worthy of the educational staff to support. Once you
do it, and if you succeed, your school will have a useful tool it can
pass on for use with and by other students. That alone could be a
powerful incentive. In my opinion, successful completion of even one
project akin to this is worthy of some sort of academic recognition.
Of course, a
Included here is an observation regarding old systems from the point of
view of power consumption:
Begin forwarded message:
From: Eric Dunbar <[EMAIL PROTECTED]>
Date: January 7, 2006 8:58:48 AM EST
To: Yellow Dog Linux General Discussion List
<[email protected]>, [EMAIL PROTECTED]
Cc: Subject: Re: [ydl-gen] Re: 4 old macs
Reply-To: Yellow Dog Linux General Discussion List
<[email protected]>
Hi, just a plug for my favourite distributed computing project,
[EMAIL PROTECTED], http://folding.stanford.edu/ (Unlike SETI [EMAIL PROTECTED]
has a number of publications to its name... though, I guess SETI was
the grand-daddy of the distributed computing projects :-)
Anyway, just a comment about the economics of running all these "old"
computers, and it's something about which I am conflicted (the
environmental damage of producing computers/disposing of them vs. the
environmental damage of the electricity used by them).
Four "old style" (hardware energy efficiency), Old World, and one
NuBus Mac (6500s are PCI, right?) running 24/7 chew up a lot of
energy. For the sake of argument (and, simplicity of calculation ;-),
let's assume each uses 100 W/h. That's 400 W/h.
Per day that's 24 h x 400 W/h = 9.6 kWh. Per year (365.25 d) that's
3506.4 kWh. At $0.10/kWh that's $350.64.
Let's assume subsidised power (most electricity is (and is fossil
fuel-produced) since the environmental costs aren't paid by
coal/fossil fuel producers or electricity users, but are "paid" by all
through the "tradgedy of the commons") at $0.075/kWh (very realistic
price in NA) and that's $262.98/a.
If you were to replace all four computers with one that'd be $65.75
and you'd save $197.24/a on the other three.
For $200 USD you can pick yourself up a 400-500 MHz G4 on eBay (not
shipped, granted), and it would do everything... run Seti faster than
all of the OldWorld Macs together (or [EMAIL PROTECTED] :-) :-) :-), act as
a server, DHCP server, firewall and wouldn't require BootX fiddling.
If your box is a dedicated DHCP server, it's far more economical to
purchase a cheap router (you can get them for $10 or $20 US). My
wireless/wired router (which will consume more power than a wired
router only) costs $10/a at $0.075/kWH if I assume that it draws the
MAXIMUM 16 W its powersupply is rated for (it's doubtful it does) vs.
my estimated $65.75 for a 100 W desktop.
Anyway, that's my environmental-economics post for the day. It's a
tragedy that "reuse" isn't always the best option in computers :-( :-(
(especially when you consider how damaging it is to produce them and
dispose of them).
Eric.
_______________________________________________
yellowdog-newbie mailing list
[email protected]
http://lists.terrasoftsolutions.com/mailman/listinfo/yellowdog-newbie
