dmd platform support - poll

dsimcha dsimcha at yahoo.com
Sun Dec 28 11:28:07 PST 2008


== Quote from Nick Sabalausky (a at a.a)'s article
 But, yes, obviously there are going to be fringe-case
> exceptions even with this, such as researchers writing custom DNA-processing
> code that's only ever going to run on their super-duper-cluster.)

I'm one of these researchers, and it would make my life a heck of a lot easier if
I could fit the entire human genome in my address space.  I don't do that much
sequence analysis, or lack of 64-bit would be a deal-breaker for D.  I mostly do
microarray analysis, which usually fits much better in 32-bit address space,
unless I start doing things like pairwise analysis among probes (some machine
learning techniques require this).  However, once in a while when my research does
take a turn into genome sequence land, the 2-gig address space limit feels like a
HUGE artificial limitation.

The alternative is to switch to nematode genomics, as their genome could have fit
into my old Pentium II's RAM.



More information about the Digitalmars-d mailing list