In article <vWvfd.246$D36.40@newsfe5-win.ntli.net>,
Martin 53N 1W <ml_news@ddnospamddml1dd.co.uk.dd> wrote:

Peter Smithson wrote:

I think you're missing the purpose of the FAQ.  Have a look at the other 

I wasn't trying to provide content for the FAQ, I was trying to demonstrate
why the proposed answer, and the reasoning behind it, was unsafe.

....And the answer is "no". The greater details can then be introduced 

The only safe, very simple, answer is "possibly", but that begs the question.
Giving an unqualified "yes" would scare people off.  Saying "no" is
misleading.

There are factors that will increase failure rates and factors that
will reduce them.  How the machine would be used without S@H is likely
to affect which of these will dominate.

The CPU is unlikely to be the first thing to fail, although the
rate of one failure mode (electromigration) will definitely increase
measurably[1], and S@H makes negligible demand on a disk other than
keeping it spun up.

The original FAQ oversimplifies by ignoring the fact that computers
do have wearout mechanisms.

[1] See, for example, the graph at the end of:

http://www.reliabilityanalysislab.com/tl_hd_0311_PulsedAluminumElectromigration.asp

which suggests that electromigration failure rates will double for about
every 20 to 25 degrees (C) temperature rise.