RussellMc wrote:
> All the above rambling may be translated into many other product areas
> with variable relative emphasis on sub-components. Knowing that things
> WILL go wrong, that Murphy cannot be beaten, but can be allowed for,
> is a good start to having things survive in the real world.

And human nature and the news media place way disproportionate emphasis on
the few cases that go wrong.  How many other cruises over how many years on
roughly similar ships were held without any such incident?  Failure can't b=
e
eliminated, only the probability reduced.  What is this failure rate,
1:1000, 1:10000, can we even tell if it only happened once?  Is 1:10000 not
acceptable?

Or put another way, before this incident occurred and it got a
disproportionate public mindshare, how much more would the passengers on
that cruise have been willing to pay to decrease the probability of such a
event happening.  Suppose you told them the chance was 1:5000.  How much
more would they pay to make it 1:10000, or 1:100000?  If it took
$1/passenger to change it from 1:5000 to 1:100000, probably many would be
willing.  What about $10, $100, $1000.  Somewhere in there the added return
isn't worth it anymore to most people.  And then it might still happen.
1:10000 means it happens every 10000 cruises or so.

Look at it yet another way, this demonstrates resiliance of the system too.
Yes, there was a failure, but the ship didn't sink, the whole ship didn't g=
o
up in flames, nobody got killed, nobody starved, and all were back safely o=
n
land a few days later.  So most likely there were some previously designed
safety systems that did their job.

We can't originally say we'll accept 1:5000 (or whatever) chance of failure=
,
then get all bent out of shape when it does go wrong once.  Who's to say th=
e
system isn't doing exactly what it was designed to do, with all the
tradeoffs carefully considered in the design?


********************************************************************
Embed Inc, Littleton Massachusetts, http://www.embedinc.com/products
(978) 742-9014.  Gold level PIC consultants since 2000.
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.