Re: [luau] Browser interface question

[Ben Beeson]

Warren and others,

I just donwloaded and installed Mozilla 1.1a, still no luck.  That 
green 
button just isn't there...

Ben 

On Monday 01 July 2002 07:35 pm, you wrote:
> The site seems to work in the latest Mozilla nightly build for me.  Which
> part doens't work?
>
> 

Re: [luau] Underclocking / Aluminum cases

[Ray Strode]

The reason for large copper bottoms is for even heating (and the reason that only the bottom is copper).  Since copper can dissipate heat very well, the heat is spread 
across the pan more evenly.  It is "large" to limit warping, provide a reserve, and still provide even heating when cold food is thrown on it.  


Interesting.  I might just have to get some copper pans now.

A copper pan of will actually cool down ... quicker then a aluminum ... pan 

of equal size. 

Hmmm.  That's what I thought before, but that article seems to say 
otherwise:



While copper is a good conductor of *heat*,
it is inferior to *aluminum* for getting rid of the *heat* it has absorbed. 

And it makes sense.  Foil is usually cool enough to grab out of a heated 
oven with my hands.  But then again, i've never tried copper foil, so 
maybe that's

not such a fair example.


Correct..
Higher number means more heat transfer

Thermal Conductivity, W/cm-K
Aluminum 2.165 
Copper 3.937 
Gold 2.913 
Iron .669 
Silver 4.173 
 

Yeah, that looks familar.   I think i remember reading that when i took 
physics.


The fins increase the surface area exposed to the air flow allowing more cooling.  Just as a soda dumped on a hot sidewalk will warm up much quicker then if it was sitting 
in a can on the same sidewalk.   


That makes perfect sense.


Here is an interesting page about copper and aluminum  (I used a Google cache 
because I could not get to the leading page from the site directly) -->

http://216.239.35.100/search?q=cache:pXcbnqmZaPkC:www.benchtest.com/alum%26copp.html+specific+heat+capacity+aluminum&hl=en&ie=UTF-8  
 

This pages logic seems strange to me.  If aluminum transfers heat slower 
than copper, then wouldn't using a copper base and aluminum fins create a
heat build up in the base?  I mean the fins wouldn't be able to receive 
the heat as fast as the copper can give it out, so wouldn't that be a 
bottleneck that

would limit efficiency?

--Ray

[luau] Wow! KDE 3.0.2

[W. Wayne Liauh]

Just installed KDE 3.0.2.  I must say . . .

WOW!

Re: [luau] Info on POS terminal

[al plant]

Jeff Zidek wrote:
> 
> >
> >
> >I just wanted to let anyone who might be in need of a cheap linux box for 
> >whatever reason that I successfully loaded both redhat 7.3 and Mandrake 8.2 
> >on an  Omron POS terminal I got at Hawaii Computer Recyclers.  It comes with 
> >a 200mhz processor a 2.1 gig hard drive a 10/100 3C905b TX-NM Nic a floppy 
> >drive and a slot where a CDROM can go.  I put one in mine to do the 
> >installs. It has two SDRAM slots (comes with 32MB ram) 2 serial ports, two 
> >USB ports, 1 parallel port a 4MB onboard video, onboard sound and can take 
> >upto a 300mhz processor if you happen to have one laying around.  They had 
> >400 when I bought mine 2 weeks ago and quoted me 105 dollars each.  25 
> >dollars more if you want a cdrom installed.  The POS box measures about 4" X 
> >13" X 15", fairly compact.  Main drawback is not very upgradable. Since one 
> >of the two pci slots is taken up by the NIC that only leave one pci slot 
> >open to ad another card.  Mandrake 8.2 installed flawlessly using graphic 
> >installer.  Redhat b!
 al!
> ked at the graphic installer and had to do a text install. Autopartitioner 
> failed to work as well but fdisk did the job.  After the initial install 
> which failed to detect the onboard audio was detected and installed by KUDZU 
> on first boot. Although a tougher install after installation Redhat 7.3 ran 
> great.  I may have to drag out a copy of Redhat 6.2 and try it with that.  
> Just thought someone might be interested.  Jeff Zidek
> >
> 
> ___
> LUAU mailing list
> [EMAIL PROTECTED]
> http://videl.ics.hawaii.edu/mailman/listinfo/luau
###


I have installed FreeBSD 4.5 on an OMRON box from them and it worked the
first try. A good stable box for Unix.  
-- 
Aloha! Al Plant - Webmaster http://hawaiidakine.com
Providing FAST DSL Service for $28.00 /mo. Member Small Business Hawaii.
Running FreeBSD 4.5 UNIX & Caldera Linux 2.4 & RedHat 7.2
Support OPEN SOURCE in Business Computing. Phone 808-622-0043

Re: [luau] Underclocking / Aluminum cases

[yuser]

On 2 Jul 2002 at 16:33, Ray Strode wrote:

> > My feeling is that copper would tend to hold the heat, for example  a 
> > copper clad bottom on a pot or skillet.  I'm pretty certain that 
> > aluminum cooling fins are the most efficient, although aluminum car 
> > radiators aren't very good while car radiators with copper fins are 
> > common  and easy to repair. 
> 

The reason for large copper bottoms is for even heating (and the reason that 
only the bottom is copper).  Since copper can dissipate heat very well, the 
heat is spread 
across the pan more evenly.  It is "large" to limit warping, provide a reserve, 
and still provide even heating when cold food is thrown on it.  A copper pan of 
will actually cool 
down (and heat up) quicker then a aluminum or cast iron pan of equal size. 


> Actually, I read somewhere (my physics book? not sure) that copper, 
> aluminum, silver, and gold were all good conductors of heat, but the 
> order of best to worst
> is: silver, copper, gold, aluminum.  
> 

Correct..
Higher number means more heat transfer

Thermal Conductivity, W/cm-K
Aluminum 2.165 
Copper 3.937 
Gold 2.913 
Iron .669 
Silver 4.173 


> The idea is, if it's a better conductor of heat, then when it's used as 
> a heat sink, the heat will away from the chip to it. where it can be 
> cooled by the air (That's why
> heat sinks are in fins i think).

The fins increase the surface area exposed to the air flow allowing more 
cooling.  Just as a soda dumped on a hot sidewalk will warm up much quicker 
then if it was sitting 
in a can on the same sidewalk.   

 
> Anyway, i'm certainly no expert, and could probably be wrong.  Maybe one 
> of the other people on the list that are, could verify what i'm saying.
> 
> --Ray
> 

Here is an interesting page about copper and aluminum  (I used a Google cache 
because I could not get to the leading page from the site directly) -->

http://216.239.35.100/search?q=cache:pXcbnqmZaPkC:www.benchtest.com/alum%26copp.html+specific+heat+capacity+aluminum&hl=en&ie=UTF-8
  

Re: [luau] Lilo to get root shell

[MonMotha]

I had over 200 days on my laptop before I reinstalled win98 on it for 
the MPEG decoder (coindidentally, the hard drive died within 6 hours 
after installing win98...).


Of course, I didn't really do much on it, but it was impressive to be 
able to claim I could beat my school's Novell server :)


--MonMotha

[EMAIL PROTECTED] wrote:

Hey
Speaking of uptime...

Red Hat Linux release 7.1 (Seawolf)
Kernel 2.4.2-2 on an i586
login: yuser
Password:
Last login: Sun Jun 30 19:08:17 from yuser
[EMAIL PROTECTED] yuser]$ uptime
 10:33pm  up 179 days,  8:11,  1 user,  load average: 0.02, 0.01, 0.00

Not to shabby for a home server.  My previous uptime was just about 100 and I 
wacked it.  I found out my hot swap SCSI enclosure was not so hot after all.

On 2 Jul 2002 at 16:18, MonMotha wrote:


if anything just seeing the uptime reset 
might spark curiosity and the admin might find the rootkit, or whatever.


--MonMotha

Re: [luau] Lilo to get root shell

[yuser]

Hey
Speaking of uptime...

Red Hat Linux release 7.1 (Seawolf)
Kernel 2.4.2-2 on an i586
login: yuser
Password:
Last login: Sun Jun 30 19:08:17 from yuser
[EMAIL PROTECTED] yuser]$ uptime
 10:33pm  up 179 days,  8:11,  1 user,  load average: 0.02, 0.01, 0.00

Not to shabby for a home server.  My previous uptime was just about 100 and I 
wacked it.  I found out my hot swap SCSI enclosure was not so hot after all.

On 2 Jul 2002 at 16:18, MonMotha wrote:

> if anything just seeing the uptime reset 
> might spark curiosity and the admin might find the rootkit, or whatever.
> 
> --MonMotha

Re: [luau] Underclocking / Aluminum cases

[MonMotha]

The characteristic you're looking to minimize is Thermal Resistance. 
I'm no thermo expert though so possibly do some googling on what that is 
exactly.


Copper will tend to suck the heat away faster and then the objective of 
the "fins" is to maximize contact between the copper and the air (much 
larger surface area than cpu core alone) to improve conduction between 
air and copper as the copper to copper conduction will occur readily to 
transfer the heat from the area of the CPU core up to the fins.


--MonMotha

Ray Strode wrote:
My feeling is that copper would tend to hold the heat, for example  a 
copper clad bottom on a pot or skillet.  I'm pretty certain that 
aluminum cooling fins are the most efficient, although aluminum car 
radiators aren't very good while car radiators with copper fins are 
common  and easy to repair. 



Actually, I read somewhere (my physics book? not sure) that copper, 
aluminum, silver, and gold were all good conductors of heat, but the 
order of best to worst
is: silver, copper, gold, aluminum. 
The idea is, if it's a better conductor of heat, then when it's used as 
a heat sink, the heat will away from the chip to it. where it can be 
cooled by the air (That's why

heat sinks are in fins i think).

Anyway, i'm certainly no expert, and could probably be wrong.  Maybe one 
of the other people on the list that are, could verify what i'm saying.


--Ray

Re: [luau] Underclocking / Aluminum cases

[Ray Strode]

My feeling is that copper would tend to hold the heat, for example  a 
copper clad bottom on a pot or skillet.  I'm pretty certain that 
aluminum cooling fins are the most efficient, although aluminum car 
radiators aren't very good while car radiators with copper fins are 
common  and easy to repair. 


Actually, I read somewhere (my physics book? not sure) that copper, 
aluminum, silver, and gold were all good conductors of heat, but the 
order of best to worst
is: silver, copper, gold, aluminum.  

The idea is, if it's a better conductor of heat, then when it's used as 
a heat sink, the heat will away from the chip to it. where it can be 
cooled by the air (That's why

heat sinks are in fins i think).

Anyway, i'm certainly no expert, and could probably be wrong.  Maybe one 
of the other people on the list that are, could verify what i'm saying.


--Ray

Re: [luau] HCC linux course

[Warren Togami]

Is that Ho'ala Greevy teaching that course?

Ho'ala?

- Original Message -
From: "T. David Burns" <[EMAIL PROTECTED]>
To: <[EMAIL PROTECTED]>
Sent: Tuesday, July 02, 2002 3:41 PM
Subject: [luau] HCC linux course


> There will be a beginning linux course at HCC July 18 and 23, taught by
> someone named Greevy (sounds sort of familiar?). I've no idea whether it
> will be good or not. They had one scheduled in May but it was cancelled
for
> lack of students. Info at
> , second
> entry on page after the evil empire.
>
> Someone at the install-fest was interested in this, I've forgotten who.
>
> Dave

[luau] HCC linux course

[T. David Burns]

There will be a beginning linux course at HCC July 18 and 23, taught by 
someone named Greevy (sounds sort of familiar?). I've no idea whether it 
will be good or not. They had one scheduled in May but it was cancelled for 
lack of students. Info at 
, second 
entry on page after the evil empire.


Someone at the install-fest was interested in this, I've forgotten who.

Dave

[luau] Fanless Athlon XP

[W. Wayne Liauh]

Well, since for most office tasks, clock speed provides absolutely no 
benefit once you move into the GHz range, the best desktop PC may be to 
use a couple of the mobile XPs in an SMP MB.


This effectively doubles the size of the cach to 512 KB, as well as data 
transfer rate.  And this system probably will be cheaper than doubling 
the clock speed.

Re: [luau] Lilo to get root shell

[MonMotha]

lilo has the restricted flag that I usually use to allow normal booting 
without a password, but requires a password if any parameters are used. 
 Many new BIOSes can have two separate passwords (supervisor and user) 
and only allow a floppy boot on supervisor (of course you can clear the 
CMOS, but that requires removing the case).


It is possible to make a casual attack difficult if not impossible; all 
of the methods discussed here require a reboot, which on a critical 
machine will not go unnoticed, if anything just seeing the uptime reset 
might spark curiosity and the admin might find the rootkit, or whatever.


--MonMotha

Dustin Cross wrote:

I know several ways to get root with physical access, but I didn't know
about using lilo like this or how to secure it.  This is useful if you have
linux workstations and want to make sure employees/users can't do things
they shouldn't.

Most of us know to set a bios/prom password so users can't boot floppy or
CD to get access.

This type of information won't protect us from a ruthless blackhat, but
will help keep authorized users inline.

Dusty

Re: [luau] Lilo to get root shell

[Dustin Cross]

I know several ways to get root with physical access, but I didn't know
about using lilo like this or how to secure it.  This is useful if you have
linux workstations and want to make sure employees/users can't do things
they shouldn't.

Most of us know to set a bios/prom password so users can't boot floppy or
CD to get access.

This type of information won't protect us from a ruthless blackhat, but
will help keep authorized users inline.

Dusty




> You didn't know that?
>
> There's numerous ways to root a box you haev physical access to.
> Basically, if an attacker has physical access, you're 0wn3d buddy.
>
> linux single
> linux 1
> *(Above two won't work on some distros, slack comes to mind as it's
> sulogin requires a root pw)
>
> linux init=/bin/bash (completly bypass normal startup and spawn a root
> shell, no login needed)
>
> Use a bootdisk
> Remove hard drive and mount in a different machine
>
> If the attacker has physical access, consider yourself screwed (unless
> you have an encrypted FS that requires a passphrase at startup).
>
> --MonMotha
>
> (quote cut because it was really long)
>
> ___
> LUAU mailing list
> [EMAIL PROTECTED]
> http://videl.ics.hawaii.edu/mailman/listinfo/luau

Re: [luau] Underclocking / Aluminum cases

[Robert Green]

Yeah, I'd never do it that way, I am always short of outlets. I keep
the cases closed up and save the desk fan for pointing at ME! 
I've noticed the interference when I realized I can't keep em on my
desk itself or it scrambles the monitor display.

--- MonMotha <[EMAIL PROTECTED]> wrote:
> 
> Be really careful with desk/room fans as they have really ugly
> magnetic 
> fields and can cause total EMI chaos in your comp.
> 
> ...
> 
> --MonMotha


__
Do You Yahoo!?
Sign up for SBC Yahoo! Dial - First Month Free
http://sbc.yahoo.com

Re: [luau] Lilo to get root shell

[MonMotha]

You didn't know that?

There's numerous ways to root a box you haev physical access to. 
Basically, if an attacker has physical access, you're 0wn3d buddy.


linux single
linux 1
*(Above two won't work on some distros, slack comes to mind as it's 
sulogin requires a root pw)


linux init=/bin/bash (completly bypass normal startup and spawn a root 
shell, no login needed)


Use a bootdisk
Remove hard drive and mount in a different machine

If the attacker has physical access, consider yourself screwed (unless 
you have an encrypted FS that requires a passphrase at startup).


--MonMotha

(quote cut because it was really long)

[luau] Lilo to get root shell

[Dustin Cross]

Just received this from the ITworld LINUX SECURITY news letter and thought
some would find it interesting.


--- July 02, 2002
Published by ITworld.com -- changing the way you view IT
http://www.itworld.com/newsletters


Another Backdoor to Root Access
By Brian Hatch

In last week's article, I showed you how to enter single user mode at the
lilo prompt, ala:

lilo: linux single

or

lilo: linux 1

Both of those arguments tell init to boot into runlevel 1. If you have
sulogin set to run, then single user mode is only available if you know the
actual root password, which is a good thing. However, another method exists
for you to gain passwordless root access without using single user mode at
all.

Normally, the linux kernel will launch /sbin/init once it's finished
loading. init is responsible for starting all the programs appropriate for
your given runlevel based on the entries in the /etc/inittab file. That's
why init is always process #1 when you do a 'ps'. However, we can tell the
Linux kernel to run a different program instead of /sbin/init by using
the 'init=' option on the lilo command line:

lilo: linux init=/bin/bash

Now the kernel will launch /bin/bash as root. Viola! A root shell, no
questions asked. You could run anything you wanted, but /bin/bash is
probably the most convenient method.

When you boot Linux in this manner, you'll find that your disks are mounted
read-only[1]. Once you're at a shell though, fixing this is trivial:

# fsck /
# mount -orw,remount /

So you can see that enabling sulogin is not sufficient to prevent someone
at the console from getting a root shell; you must create password
restrictions for your kernel definitions to prevent anyone from passing
command-line arguments to the kernel. I showed you how to do this last
week, but let's recap.

Add 'restricted' and 'password' options to the relevant /etc/lilo.conf
kernel definition[2]:

image=/boot/vmlinuz
label=linux
restricted
password=suLoginIsntSufficient
read-only
root=/dev/hda7

Of course, don't forget to make the lilo.conf file unreadable by local
users:

# chmod 600 /etc/lilo.conf

And now re-run lilo when you're done:

# lilo

If you're paranoid, then you can always make lilo.conf immutable
(unchangeable) with chattr[3]

# chattr +i /etc/lilo.conf

If you ever do need to make changes, then you'll need to turn off the
immutable bit first:

# chattr -i /etc/lilo.conf
# $EDITOR /etc/lilo.conf
# chattr +i /etc/lilo.conf

So, does this mean we're completely secure now? Nope, sadly not. Other ways
remain that provide root access to the machine, such as booting from
alternate devices like a floppy/CD[4] or just pulling out the disk and
mounting it on a different machine and accessing it there directly, but
we've covered the most direct and simple methods via our lilo
configurations.

NOTES

[1] You could have the kernel mount '/' read write by specifying:

lilo: linux rw init=/bin/bash

at the lilo prompt. However, I like to fsck the drive manually and remount.
Call me paranoid.

[2] Actually, you can use restricted or password in the global section as
well, not just in an image definition. However, I like having different
passwords for each image, so I don't put 'password' in the global section.
Restricted, on the other hand, is fine if you want them all restricted.

[3] chattr only works on ext2/ext3 file systems.

[4] Most BIOS can disable or password-protect the ability to boot off other
devices. I leave that as an exercise for the reader so we can get onto more
interesting topics again next week.

About the author(s)
---
Brian Hatch is Chief Hacker at Onsight, Inc, and author of Hacking Linux
Exposed and Building Linux VPNs. Brian can be reached at
[EMAIL PROTECTED]

Re: [luau] Underclocking / Aluminum cases

[Jim Roby]

Last night on slashdot was a link to water cooling,which leads to many
interesting sites...total kit for $200 on one,also saw a page from a guy
who says the fan can be quieted with out significant proformance loss
by simply lowering the voltage from 12V to 7V...he does it by placing
diods in line,observing curent flow(two in,two out)Commercially they sell
a little inline gysmo that allows you to dial the fan RPM,believe $7.
Wonder is /. reading our list? :-)

On Tue, 2 Jul 2002, MonMotha wrote:

> Robert Green wrote:
> > Your idea of just leaving the cover off is less efficient than a
> > well-designed case with good air flow via a case fan, unless you're
> > going to use a room fan to blow into the open case (or other method
> > to increase the air flow across the components).
> 
> Be really careful with desk/room fans as they have really ugly magnetic 
> fields and can cause total EMI chaos in your comp.
> 
> ...
> 
> --MonMotha
> 
> ___
> LUAU mailing list
> [EMAIL PROTECTED]
> http://videl.ics.hawaii.edu/mailman/listinfo/luau
> 

Re: [luau] Underclocking / Aluminum cases

[MonMotha]

Robert Green wrote:

Your idea of just leaving the cover off is less efficient than a
well-designed case with good air flow via a case fan, unless you're
going to use a room fan to blow into the open case (or other method
to increase the air flow across the components).


Be really careful with desk/room fans as they have really ugly magnetic 
fields and can cause total EMI chaos in your comp.


...

--MonMotha

Re: [luau] Underclocking

[Dean Fujioka]

Heheheheheheheh I've thought about doing that a few times, but to 
ventilate my own heat, not the CPU's :-D

dean

Joel wrote:

If you want to be *really* serious about increasing your airflow, you may need 
to modify your case, as described at http://www.2cooltek.com/case001.html.


;-)

--Joel

 


You don't mention if you intend to still have case fans or not.
Remember, your CPU isn't the only source of heat, and good airflow is
important (CPU will produce heat even if it doesn't need it's own fan
and that heat needs to go somewhere).  You also need to take into
account the heat produced the GPU, HDD, PSU, CD-ROM, CD-RW (these can
make more heat than you'd expect), etc.  Remember, with really good
airflow (think fan ducting), it is possible to even run a semi-recent
chip at full speed using just a heatsink (see Compaq, IBM, HP, Gateway,
etc).  I've seen numerous Pentium IIs and some Pentium IIIs running with
just a heatsink and using fan ducting to blow air across it.

When considering airflow, remember that you can considerably improve it
by rerouting cables behind drive cages (45 degree folds will get 90
degree angles in your ribbon cables.  You can also use round cables.
You might also consider using pieces of cardboard to create different
"airflow zones."  For example, you might have the 5.25" bays and PSU
completely separated from the bottom part, especially if your PSU blows
hot air inward (as reccomended by the ATX spec) rather than outward
(what most PSUs do in white boxes).  Also remember that air MOVEMENT is
not the objective, but rather air FLOW.
   



___
LUAU mailing list
[EMAIL PROTECTED]
http://videl.ics.hawaii.edu/mailman/listinfo/luau


 

[luau] (Was: Underclocking) Fanless Athlon XP

[W. Wayne Liauh]

Very interesting, just when we were talking about fanless CPUs, AMD 
started shipping Athlon XP SFF processors featuring low power consumption.:


http://www.xbitlabs.com/news/story.html?id=1025602160


Vcore=1.05 to 1.45 V, Max heat dissipation: 35W, price: $116

"Well, AMD has introduced a CPU for a relatively new market segment of 
economy PCs, which has always been the territory of VIA C3 and Transmeta 
Crusoe CPUs because of specific thermal and power requirements. Now it 
is possible to assemble power-saving and compact computers boasting the 
performance and features of their elder brothers, which is a very 
encouraging thing for those users, who need a quiet but at the same time 
fast solution.


This is indeed very exciting!

Re: [luau] Underclocking / Aluminum cases

[Robert Green]

Your idea of just leaving the cover off is less efficient than a
well-designed case with good air flow via a case fan, unless you're
going to use a room fan to blow into the open case (or other method
to increase the air flow across the components).

Cooling (heat transfer) is a function of the thermal properties of
the meterials involved (heat sink metal, and the air or water cooling
medium) combined with the flow rate of the coolant (i.e. air
circulation or water flow). 

Without a strong air flow, you rely mainly on radiant heat transfer
and natural convection. Forced convection using a fan / pump in a
closed case is much more effecient, assuming the case has good air
flow paths available.


--- Joe Linux <[EMAIL PROTECTED]> wrote:
> Refrigeration systems tend to use Steel in the condenser coil (the
> hot 
> one) and aluminum or copper in the evaporative coil.
> My feeling is that copper would tend to hold the heat, for example 
> a 
> copper clad bottom on a pot or skillet.  I'm pretty certain that 
> aluminum cooling fins are the most efficient, although aluminum car
> 
> radiators aren't very good while car radiators with copper fins are
> common  and easy to repair.

This is incorrect. Copper is a better heat conduction than aluminum
or steel. The reasons steel and aluminum are in common use is cost,
weight, strength and corrosion resistance.

> Maybe wayne can get the heater coil out of a car and put that
> outside 
> his house with a small electric water pump.  Then he could pipe in
> the 
> cooled water through a copper tube forced in between the  fins of
> an 
> aluminum heat sink.

A pre-made water cooling system would probably be a better option
than a home-grown one. It can be devillishy hard to get a leak-proof
system without the proper tools, materials and practice; a small leak
in a car is not a big deal, but a spray onto a motherboard is
basically guaranteed to fry a computer.


__
Do You Yahoo!?
Sign up for SBC Yahoo! Dial - First Month Free
http://sbc.yahoo.com

Re: [luau] Underclocking / Aluminum cases

[Joe Linux]

Refrigeration systems tend to use Steel in the condenser coil (the hot 
one) and aluminum or copper in the evaporative coil.
My feeling is that copper would tend to hold the heat, for example  a 
copper clad bottom on a pot or skillet.  I'm pretty certain that 
aluminum cooling fins are the most efficient, although aluminum car 
radiators aren't very good while car radiators with copper fins are 
common  and easy to repair.


Maybe wayne can get the heater coil out of a car and put that outside 
his house with a small electric water pump.  Then he could pipe in the 
cooled water through a copper tube forced in between the  fins of an 
aluminum heat sink.


I'm not sure if aluminum cases have any practical value when it comes to 
cooling, but they are lighter weight, so they would be good if the 
computer is transported a lot.  Perhaps old heat sinks could be salvaged 
and glued onto the top and sides of a steel or aluminum case to improve 
heat dissipation..





Aluminum cases not only look cool, but are functional as well.  You 
might also look to see if anyone offers copper cases as copper is a 
really efficient medium for heat transfer.

Re: [luau] Underclocking

[Joe Linux]

I think it would be simpler to leave the cover off.

Joel wrote:

If you want to be *really* serious about increasing your airflow, you may need 
to modify your case, as described at http://www.2cooltek.com/case001.html.

Re: [luau] Underclocking / Cooling

[Joe Linux]

You could use two PC cases.  One for the mother board and associated 
cards, and the other for the drives.  That would be a cheap almost ready 
to go solution.  You could run both cases without the covers to further 
dissipate the heat statically.


A water cooled system would most likely require a pump which would be 
more noisy than a fan moving air.  You could possibly have a pond 
outside your house with a fountain pump that would circulate water 
through a small copper tube forced in between the fins of a an aluminum 
heat sink that would be rather silent.  If the tube was continuous or 
well soldered, the possibility of a leak would be minimal.  You already 
have lots of water pipes in your home under more pressure.  While a 
system such as this could work, it would most likely be cumbersome and 
make the PC difficult to service.  Not to mention the fact that the cost 
of the pond and pump and everything else associated with it might be 
more than the PC itself.


W. Wayne Liauh wrote:




3.  I am thinking about re-arranging my desktop box so that the hard 
drive(s) and both CDROM drives will be located ourside of the case 
(i.e., the IDE cables will be extended outside of the case).  Any 
potential problems? Perhaps someone is crafty enough to build a 
transparent lucite case to house those IDE drives.



6.  Water-cooled system seems interesting.  But what's the chance that 
the system may leak thus really messing up your PC?


___
LUAU mailing list
[EMAIL PROTECTED]
http://videl.ics.hawaii.edu/mailman/listinfo/luau