__/ [ nessuno@xxxxxxxxxxxxxxxxxxx ] on Wednesday 26 July 2006 07:39 \__
>> I don't see "power" in E=mc^2 ;-)
>
> What's interesting about the black hole collision is the power output
> (energy/time). The climax of the collision only lasts a few minutes,
> maybe less. (It's basically the light transit time across the black
> holes.) The stars in the rest of the universe take maybe 10^10 years
> to burn out.
>
>>
>> >I can't imagine that anyone would ever consider proprietary software
>> >for such a purpose.
>>
>> Why? What linux has to do with differential equations?
>
> Nothing, except that you need parallel processors to handle PDEs in
> many dimensions, and if you're going to be linking together lots of
> nodes, especially if you're a scientist type who always wants to know
> how things work, you don't want closed source, you don't want to go
> around paying royalties and signing EULAs every time you add a node,
> etc etc. Same reason Google chose Linux for their clusters (I'm sure
> it took them about a microsecond to make that decision).
Clearly, the troll has chosen the wrong person to address
such skepticism at. As I quickly looked for and your
homepage yesterday, I am aware that you will understand the
value of OSS versatility and distributability. Without
Linux, I'd probably be hanging about here for many years,
struggling to get computing power to carry out experiments
and get a degree. I owe to GNU/Linux a lot of what I've
achieved. Even now, I have plenty of time to spare and spend
in COLA, owing to Linux.
|
|
