Roy Schestowitz <newsgroups@xxxxxxxxxxxxxxx> wrote:
> __/ [ Linonut ] on Saturday 03 March 2007 02:25 \__
>
>> After takin' a swig o' grog, yttrx belched out this bit o' wisdom:
>>
>>> Roy, does it really matter to you how much hardware linux runs on? Do you
>>> really have a MIPS machine lying around to play with? Or an UltraSPARC?
>>> Or an RS/6000 machine?
>>>
>>> Face it, the only thing thats important to you is that Linux runs on x86,
>>> so stop pretending.
>>
>> However, the fact that it builds and runs on a large number of platforms
>> indicates that the code has been read and reread many times.
>>
>> Heck, just getting code to compile on both gcc and Visual Studio
>> uncovers a fair number of mistakes (at least in my code!)
>
> That's a nice perspective that never crossed my mind. But what I was going to
> say (I was in a hurry so I didn't) is that you can start small and grow
> larger without code rewrites. You can use inherent scalability and
> cross-platform portability of Linux to move your -- let us say -- DNA
> sequencing program from a laptop to a 10,000 node supercomputer.
>
It's harder to port a DNA sequencing program from a laptop to a 10,000
node supercomputer, both running linux, than it is to port it from a laptop
running Linux to a laptop running Windows.
I know, it doesn't seem like it. But take into consideration the necessity
for re-writing enormous parts of an application like that to finely
internally thread, take advantage of grid/cluster memory schemes, etc.
Even if you move it from one processor architecture to another under
linux, you have to recompile the whole thing, unless it's written to
take advantage of a java or inferno virtual machine.
-----yttrx
--
http://www.yttrx.net
|
|
