Re: [tlug] Now, this is getting out of hand!

["Niels Kobschaetzki" <n.kobschaetzki@example.com>]

On Sun, Jun 8, 2008 at 1:00 PM, Curt Sampson <cjs@example.com> wrote:
> On 2008-06-05 20:24 +0900 (Thu), Stephen J. Turnbull wrote:
>
>> Josh Glover writes:
>>
>>  > I think he meant that you could tell people more about the economics
>>  > behind the idea that sometimes it is more expensive to analyse the
>>  > requirements than just try something and see if it is good enough.
>>
>> Well, I know what he was getting at, but there are limits to what any
>> science can do.  This is a matter of empirical fact, not something
>> that's particularly subject to analysis.
>
> Ok, I've gone back to my original post to see just what I was saying, and
> it turned out to be this:
>
> ] Stephen can tell you more about this I'm sure, but it turns out that
> ] this is a basic issue with buying data-centre gear: it's often more
> ] far more expensive to figure out how well something is working than it
> ] is just to (depending on your situation) to buy stuff that (might be)
> ] guaranteed to work well or to buy stuff that's cheap and hope.
>
> This seems to be a matter entirely amenable to analysis, and in dire
> need of it, really. The emperical facts are not even the topic here;
> it's the question of how much in the way of resources one should spend
> to get, or not get, those facts. To spend, say, $2000 to find out that
> a $2000 switch will do just as well in your particular situation as a
> $3000 one that's known to work is obviously counterproductive.
>
> Surely someone's done some research on good strategies for making these
> sorts of decisions. It might be the game theory guys, rather than the
> economists, but those are rather related disciplines, anyway.

Well, actually I can't really see how you would analyse that. The
problem would be that you'll have to work with different preference
sets that include risk aversion. I make here only educated guesses but
I think that in the end it would come to something like "it depends on
the risk aversion of the decision maker" and not a real "solution"
which would differ from the intuitive one.

Niels