Ken, I am also an engineer. At least officially. Although if you were to
argue that "software engineering" has almost nothing with "true
engineering", you wouldn't encounter any opposition from me.
I can give you an absolutely immediate example. The system that I am
working on right now has certain infrastructure layer. The engineer
responsible for that layer has decided that it has to be optimized for
performance. And optimized it was. Now, later after the most bulk of the
code was written, it became apparent that this code was better optimized
for flexibility, extensibility and maintainability. None of these
concerned the said engineer. The performance optimization was very low
hanging fruit and it was a great opportunity to do fun stuff... However,
we now have fundamental (as in building foundation that is called
"fundament" in Russian) problem in our system.
It is hard to reason about network cables for me, but I was particularly
alarmed by your saying "I HAVE to look for ANY opportunity to
optimize"... I think of optimization as of extremely sharp razor. And
the person who wields that razor has to be extremely careful when they
actually put the razor to use.
Nothing personal or nothing disrespectful intended here. As well, we can
take it off the list, as this is very much off topic.
Boris
On 2/6/2016 12:29, Bruce Walker wrote:
On Sat, Feb 6, 2016 at 2:21 AM, Ken Waller <kwal...@peoplepc.com> wrote:
"I'm an engineer, I have to look for any opportunity to optimize. " <--
that's fundamentally wrong.
How's that Boris?
I'm also an engineer & while I might not act on those opportunities, I need
to at least be aware of them.
Agreed, 110%.
A very large part of my career was built on optimizing. Cost, size,
weight, reliability, appearance, ... something. Especially cost. The
fewer lines of code the better, since bugs increase exponentially with
complexity. The fewer the components the better since that generally
drops the cost and increases MTTF.
--
PDML Pentax-Discuss Mail List
PDML@pdml.net
http://pdml.net/mailman/listinfo/pdml_pdml.net
to UNSUBSCRIBE from the PDML, please visit the link directly above and follow
the directions.
