GB wrote:
"The Todal" wrote in message
...

The OP's father in law might be of an age where he doesn't want to run
computer games or anything very much other than a web browser and a
word-processing package. It ought to be possible to do these things on a
ten year old machine especially if you still use Office 97 (despite
Microsoft's petulant demand that we should upgrade) or Lotus SmartSuite.


I can confirm that. I used to run Win 98 and Office 97 on a Celeron 333MHz
machine quite comfortably, as well as IE, OE and Outlook. Although a move to
a faster processor did speed things up, it was not dramatic. Disk I/O is
about the same on new machines as on that one, and that's one of the
bottlenecks.


I can do better than that. I run Win98, Office 2000, Firefox and Thunderbird on
a 10 year old Pentium II with 64 MB of ram. I use it almost every day as my
main machine - and I'm using it now. I've added larger hard disks, an ethernet
card and a DVD burner. I don't play games so the fact that the PCI graphics
card (remember them) only has 2 MB of ram doesn't matter. It has never had a
hardware failure. The 200W PSU would be regarded as laughable by todays standards.

I have also used a state of the art games machine running XP pro. There is no
noticeable difference in speed, apart from the very slightly quicker screen
refresh - booting though was slower on the XP. The XP machine of course was
burdened with latest software and AV crap.

To follow up....it was the motherboard at fault. Having spoken to
Consumer Direct we agreed it would be pretty much impossible to
establish that there was a fault at the point of manufacture, which
would more or less require Intel to admit a problem with the
motherboard. I bought a new motherboard for 35 quid and fitted it and
it is now fine again.

wrote in message
ups.com...
To follow up....it was the motherboard at fault. Having spoken to
Consumer Direct we agreed it would be pretty much impossible to
establish that there was a fault at the point of manufacture, which
would more or less require Intel to admit a problem with the
motherboard. I bought a new motherboard for 35 quid and fitted it and
it is now fine again.

Wohoo - that's what I diagnosed! (28th Jan - from "Gorf") I should maybe do
this for a living...

I'd have thought that a fault at point of manufacture could be easily
established. A mobo is solid state, and should easily outlive other
components such as hard drives which are susceptible to shock (user) damage.

Having said that, for £30 odd, is it worth it? In the US, perhaps, where a
HDD manufacturer has been on the receiving end of a class action suit
because the GB they quoted did not match the GB after formatting. In terms
of lost capacity vs the price of the disc, I'd be surprised if it came into
whole pounds sterling, but if enough people are p*ssed off about it, there's
scope for a legal attack...

wrote in message
ups.com...
To follow up....it was the motherboard at fault. Having spoken to
Consumer Direct we agreed it would be pretty much impossible to
establish that there was a fault at the point of manufacture, which
would more or less require Intel to admit a problem with the
motherboard. I bought a new motherboard for 35 quid and fitted it and
it is now fine again.

Wohoo - that's what I diagnosed! (28th Jan - from "Gorf") I should maybe do
this for a living...

I'd have thought that a fault at point of manufacture could be easily
established. A mobo is solid state, and should easily outlive other
components such as hard drives which are susceptible to shock (user) damage.

Having said that, for £30 odd, is it worth it? In the US, perhaps, where a
HDD manufacturer has been on the receiving end of a class action suit
because the GB they quoted did not match the GB after formatting. In terms
of lost capacity vs the price of the disc, I'd be surprised if it came into
whole pounds sterling, but if enough people are p*ssed off about it, there's
scope for a legal attack...

On Mon, 12 Mar 2007 14:10:03 +0000, "
wrote:

To follow up....it was the motherboard at fault. Having spoken to
Consumer Direct we agreed it would be pretty much impossible to
establish that there was a fault at the point of manufacture, which
would more or less require Intel to admit a problem with the
motherboard. I bought a new motherboard for 35 quid and fitted it and
it is now fine again.


Thanks for the update.
--
Alex Heney, Global Villager
Pardon my driving, I'm trying to reload . . .
To reply by email, my address is alexATheneyDOTplusDOTcom

"G Hardy" wrote

I'd have thought that a fault at point of manufacture could be easily
established. A mobo is solid state, and should easily outlive other
components such as hard drives which are susceptible to shock (user)
damage.

Not at all. Systemboards (as we call them - I work for a well known OEM) do
fail. The statistics for replacement of parts will show that systemboard
failures are not that uncommon - there's a lot of components on there, and
those components do deteriorate over time.

Not at all. Systemboards (as we call them - I work for a well known OEM) do
fail. The statistics for replacement of parts will show that systemboard
failures are not that uncommon - there's a lot of components on there, and
those components do deteriorate over time.

IANAL

From personal experience (although you undoubtedly see more than me), I
haven't had a motherboard fail in less than 4-5 years - my only failure
keeled over due to the capacitor bulging problem in April/May 2006 (ECS
K7S5A).

I've had at least one system on 24/7 since '94-'95, and two systems on
24/7 since '99-'00, having swapped one base unit 4 times in that period
for a faster system each time (486 DX2-66 - 350Mhz AMD - 1Ghz AMD with
later CPU change to 2Ghz, same mobo - 3.4Ghz P4), the other twice
(300Mhz Celery - 2.4Ghz Celery). The machines were fully functional
when they were "retired", bar the one with capacitor failure (purchased
in about 2001).

"Colin Wilson" o.uk wrote
Not at all. Systemboards (as we call them - I work for a well known OEM)
do
fail. The statistics for replacement of parts will show that systemboard
failures are not that uncommon - there's a lot of components on there,
and
those components do deteriorate over time.

From personal experience (although you undoubtedly see more than me), I
haven't had a motherboard fail in less than 4-5 years

I used to support 500+ older servers and I saw 2 systemboard failures in a
year.

I've spent over a year doing hardware diagnostics over the phone as well,
and I would guesstimate systemboards to account for 5-10% of all failures.

I used to support 500+ older servers and I saw 2 systemboard failures in a
year.

I note you mention this was "older servers", which would imply fairly
good longevity for them to have lasted long enough to gain that
description :-}

I've spent over a year doing hardware diagnostics over the phone as well,
and I would guesstimate systemboards to account for 5-10% of all failures.

Was this on hardware within warranty, or older machines ? - I wonder
what the relative failure rate was as against the number of units sold
in the same period.

Somewhat off-topic...

"Colin Wilson" o.uk wrote
I used to support 500+ older servers and I saw 2 systemboard failures in
a
year.

I note you mention this was "older servers", which would imply fairly
good longevity for them to have lasted long enough to gain that
description :-}

It would more imply that I have good longevity.
Also, older computers tend to have more discrete components, so you would
get a PC with an additionnal network card, video card, SCSI card, etc. These
days, the trend is to get all these components integrated into a single
system board.
New systemboards fail more often than older ones because they are more
complex, are built to finer tolerances, and operate at higher temperatures.
The latest processors don't help either - they blow very quickly if they're
not cooled properly.

I've spent over a year doing hardware diagnostics over the phone as well,
and I would guesstimate systemboards to account for 5-10% of all
failures.

Was this on hardware within warranty

Hardware in use. Most within a 3 year warranty period, or a 4-5 year
extended warranty contract.

I wonder
what the relative failure rate was as against the number of units sold
in the same period.

So do I ;-) And so does the competition - exact figures are commercially
sensitive.

If you're interested in this sort of thing, Google did some very interesting
research on Disk Failures, which you will find at:
http://labs.google.com/papers/disk_failures.html