Archive of the former Yahoo!Groups mailing list: MOTM

previous by date index next by date
previous in topic topic list next in topic

Subject: Re: [motm] Ken Elhardt's Patchbook Series

From: Neil Bradley <nb@...>
Date: 2006-07-15

>> Being a professional embedded systems and app engineer for 18+ years and a
>> hobbyist programmer long before that, I will say that this is a completely
>> true statement.
> and having been in software development for a similar number of
> years, my opinion is that commercial "priorities", scheduling
> "decisions", and cost-cutting on various essential factors (such as
> QA) have a far greater influence on the quality of software in
> general than the IQ of the programmers concerned!

I've always found the "we had to ship it before it was ready" to be an
excuse for those who have no ability to push back to their program
management, and also an excuse for developers to be sloppy. I don't
believe making a schedule and making a quality product are mutually
exclusive. Cutting corners is something you can do and still make a good
product. One just has to cut the ∗RIGHT∗ corners. The problem is most
don't.

I guess I'm one of these rare people who will actually take an arrow and
ship a product slightly late or slightly defeatured to maintain quality.

Then there are the "developers" who don't check function result codes,
have intermittently/oddly behaved code that crashes, uninitialized
variables, incorrectly applied algorithms, don't check ranges, have no
concept of thread synchronization, leaving breaks off case statements,
memory leaks, no respect for system resources, and general sloppiness and
ignorance of their environment where rebooting the system is both
acceptable and expected. It's pathetic! And there are a hundred of these
types of "developers" for every one decent developer.

I'm jaded because the software I regularly release must have uptime
measured in years, and I've cleaned up so much ill written unstable code
in my life it's practically half my career. Lines of code number in the
300K+ range, and I and other team members of mine achieve this because we
pay attention to what we're doing, test what we write, and push back to
management with alternatives when they try to make us do something that
may sacrifice the quality of the product. Doesn't seem like it's
unachievable by anyone who actually cares or knows what they're doing. But
most don't, hence Kenneth's gripes, of which I completely agree with. And
don't get me started on some of the code we get from "outsourcing" houses.

Don't get me wrong - there are some VERY good programs out there. But when
I hear stories about people learning JAVA because they can't understand
pointers, it makes me yearn for a world where compilers of all languages
and assemblers are kept out of the hands of the undisciplined.

-->Neil

-------------------------------------------------------------------------------
Neil Bradley "If you owe the bank $100, it's your problem. If you
Synthcom Systems, Inc. owe them $100mil, it's the bank's problem." - JP Getty