[aklug] Re: A developer's take on system resource consumption

On Sun, 12 Jun 2011, The Don Lachlan wrote:

> TL;DR Idiot developers write crappy code on crappy hardware just like on
> awesome hardware.
>
> Start by reading "Worse is Better" (by Richard Gabriel, not Jamie):
> http://www.dreamsongs.com/RiseOfWorseIsBetter.html

Thanks for the link, it was an interesting read. I think I differ quite a
bit from this guy in what constitutes the "Right Thing". He's definitely
focused more on functionality and flexibility than efficient use of
resources for a specific task. I'm not sure I'd characterize the UNIX
mantra of "do one thing, but do it really well" as "worse is better".

But, then, this guy is a Lisp developer. There's a reason why it's a
relatively rare affliction. ;-)

> Then, consider Moore's Law and how using a computer that is
> below-average will compare to the average computer when your software is
> released.
>
> In my opinion, you're trying to fix a process problem with a technical
> solution. Thus, you are doomed to failure. The (vague) answer is to
> structure a group where you're developing for what the average system
> will be at release time, using awesome hardware for developers and
> "real" hardware for QA/test, alongside metrics that reflect the average
> user.

:-) I have no doubt I'm doomed to failure. Your model, too, though, is
doomed to fail. What you're suggesting is, in theory, what everyone is
already trying to do, but failing miserably at. Again, testing on "average"
or low-spec hardware for "acceptable" performance is not the same as having
to live on that same hardware. What we might find acceptable in a specific
metric might be (and usually is) completely unacceptable when you're being
beat over the head with it daily.

In a nutshell your model produces products in which the developer announces
what *others* can live with, all while not having to be subjected to those
same conditions himself.... <G> kind of sounds like our form of government,
now that I think of it. If that isn't a scathing indictment, I don't know
what is.

> With respect to Mozilla, the last is their issue - their view of an
> average user is skewed, although I don't know why (I have ideas, but no
> data to support/refute).

:-) As I mentioned, determining baselines will always be rife with conflict.
I had some small hopes that the whole smart phones and embedded devices
would accomplish what I'm suggesting. When the firefox guys realized how
hard it was to shoehorn firefox onto a Nokia N800, etc., they'd have to
clean up their act. But, before we could reap all those benefits the
hardware manufacturers started coming up with multi-core smart phones, etc.
:-P We can't win.

         --Arthur Corliss
           Live Free or Die
---------
To unsubscribe, send email to <aklug-request@aklug.org>
with 'unsubscribe' in the message body.
Received on Sun Jun 12 12:00:43 2011