[ale] code bloat

[Michael B. Trausch]

On Sun, 2009-10-11 at 12:20 -0400, Jim Kinney wrote:
> Why is the zipped Oracle 10g client for 32-bit Linux a 458MB file?
> 
> I don't want to replicate the entire dataset in the client! I just
> want to be able to a query and get a subset returned.
> 
> Argh! stupid code bloat.... 

Code bloat is becoming a larger and larger problem everywhere.  While it
seems to be especially comfy in proprietary software (I suppose that
just comes from having very limited resources, compared to many
community-oriented projects), it's becoming a problem with a good amount
of open/free software, too.

Sometimes, I think that developers write code to fit their own machines
and never check anything smaller.  I'd swear with some of the software
I've seen, it was written on a bloody 32-processor, 3GHz system with 1TB
of RAM---"but it made hardly a dent on the system when we tested it!"

Funny, a 500MHz Pentium III with 128 MB wouldn't make a dent on such a
system, either.

Maybe someone should go around to various places, like Microsoft and
Oracle and even Canonical, go to all the developers, and say "Here are
your new workstations" and give them 500 MHz Pentium III machines with
128 MB RAM and 4 GB disk space.

Of those that didn't quit, I wonder if any trimming down would get
done...

	--- Mike

-- 
Blog:  http://mike.trausch.us/blog/
Misc. Software:  http://mike.trausch.us/software/

“The greater danger for most of us lies not in setting our aim too
high and falling short; but in setting our aim too low, and achieving
our mark.” —Michelangelo