Larkin, Hobbs let's meet in the middle.

Writing software, well most of you know MS, one fix after the other,
never bug free.
I must say if John Larking gets th code right 100% bug free before
he releases whatever it is, then my hat of to him!

But who is the judge? Do you run soft to evaluate, do you
have testers, or is it yourself.
In the last case there are some question marks.

It is a tricky thing, some weeks ago I got a bug report for a
z80 disassembler I wrote in the late eighties or nineties...
Fixed it and released a new version.

As to the hardware, now after so many years I get it right first time,
but still you can think of small improvements that could be in a next version.
In the early days I got it wrong a couple of times really badly...
But I have seen so many revision PCBs, so many PCBs with wires, done
by people who claimed they were really good, almost have to laugh.


I guess it bytes you more if you make a mistake in something that will be made a
million times, then in a one of piece of scientific equipment that
can be easily updated.

But I have learned to test circuits a little bit at the time, before
using them in a bigger design.

Now, these days, with simulation it may be easier to have a look at what will happen,
but th soldering iron rules in the end.
The simulations are really good though, but the scope rules.

So, guys, somewhere in between the extreme viewpoints of 'first time perfect'
and 'many versions before it is right' is a reality - the reality
that we are neural nets that often, but not always give the correct
answer to the question-.

Even endlessly trained operators of nuclear reactors in submarines make mistakes.
Everybody makes mistakes.

So... OK Phil, I agree, after claims of perfection we should let others check too ;-)




.

Relevant Pages