Real number analysts favor some other bases like Euler's constant.
Powers of one are an easy intro for the analytically challenged.
Powers of zero are pretty subjective, they depend on
where you're coming from--if you know what I mean.
Set theorists & combinatorial analysts can get funky with powers,
but they depend on the size of the set in question--I'm sure that
_never_ comes up in SQL statement processing circles.
The progression of increasing powers of numbers between zero
and one look like my checkbook balance.
The progression of powers of numbers less than -1 looks like
my checkbook balance over time (less the singularities, which
may or may not occur while I'm not paying attention, have to ask
Schrodinger about that--sorry 'bout my spelling and char set).
Powers of complex numbers are cool, kind of like OO referencing,
"it's right there, raise to power, now it's over there, raise to another
power, oh, now it's over here! ..."
I'll leave powers of Knuth's surreal numbers to another discussion.
Well, have to go renew the subscription to sci.math.ex.mathemeticians.psycho
From: Sasha Pachev <sasha@stripped>
To: monty@stripped <monty@stripped>
Cc: matthew mcglynn <mcglynn@stripped>; mysql@stripped
Date: Friday, April 02, 1999 10:24 AM
Subject: Re: powers of two in my.cnf ?
>Michael Widenius wrote:
>> >>>>> "matthew" == matthew mcglynn <mcglynn@stripped>
>> matthew> This is probably a really dumb question...
>> matthew> Every example of buffer settings I've seen
>> matthew> has used powers of two. For example, from the
>> matthew> sample configuration file:
>> matthew> set-variable = key_buffer=16M
>> matthew> set-variable = max_allowed_packet=1M
>> matthew> set-variable = thread_stack=128K
>> matthew> set-variable = max_allowed_packet=16M
>> matthew> set-variable = keybuffer=16M
>> matthew> Similarly, the default table_cache is 64, and when
>> matthew> people ask on this list about running out of file
>> matthew> descriptors, the response is usually "set table_cache
>> matthew> to 32".
>> matthew> My dumb question is: do these have to all be powers of two ?!
>> matthew> If for my application, a table_cache of 128 is insufficient,
>> matthew> can I bump it to 150 or do I need to go all the way to 256 ?
>> I just prefer to use powers as two as this gives you the best memory
>> utilization with many memory allocation packages!
>> You don't however have to use powers of two; For the table cache 150
>> should be fine!
>And to add to what Monty has said, powers of 2 are
>"holy" numbers for a programmer, just like powers of 10
>are "holy" for non-programmers.