[TIP] Everybody wants a pony!

Victoria G. Laidler laidler at stsci.edu
Thu Apr 2 07:53:34 PDT 2009


C. Titus Brown wrote:
> On Wed, Apr 01, 2009 at 10:57:53AM -0600, m h wrote:
> -> Titus, why do scientists care about these platforms?  With cheap
> -> commodity hardware it seems most people would scale google style.
> -> Aren't x86 going to be cheaper/faster?  I guess if you are talking
> -> about some super optimized fortran compiler, yeah you care, but this
> -> is python we are talking about.
Just to add my two cents on this topic --
Scientists rarely abandon software and move completely to brand new 
packages. Old code never dies: it gets incremented and modified, it gets 
passed from grad student to grad student. It's code you've been using 
for 15 years so you really trust it... or, you want your new results to 
be directly comparable to your old results.

Which means there may very well be a piece of code that still needs that 
super optimized fortran compiler, or that takes advantages of 
idiosyncratic features of the oddball platform, that you're using as the 
core of your current analysis package, but you'd like to use python to 
glue pieces together, or do some pre- or post-processing.

Vicki



More information about the testing-in-python mailing list