Re: Alignment Nazi?

Rant number two....

On Sun, 27 Apr 2008, gjmilne64 wrote:

> I find it so much easier to read when I change it to this:
>
>         final int         counts              = result.getCounts();
>         final int         periodMicros        = result.getMicros();
>         final Rate        rate                = new Rate(new
> Duration(periodMicros, Duration.MICROSECOND), counts);
>         final Current     reagentIntensity    =
> upsElectrometer.getValue();
>         final Current     productIntensity    =
> dwsElectrometer.getValue();
>         final Temperature flowtubeTemperature =
> flowtubeThermocouple.getValue();
>         final Pressure    flowtubePressure    =
> flowtubePressureGauge.getValue();
>         final FlowRate    carrierFlow         =
> reagent.getIon().isPositive() ? pcgFlowMeter.getValue() :
> nicgFlowMeter.getValue();
>         final FlowRate    sampleFlow          =
> sampleFlowMeter.getValue();
>         final Duration    reactionTime        =
> getReactionTime(carrierFlow);
>         final Lut         ICF_LUT             =
> CONFIG.getLut(DefaultConfigIfc.ConfigId.ION_ICF_LUT);
>
> Of course, this works so long as you use a monospaced font in your
> editor.

Of course, this works so long as you use a Monospaced font and use the
"use spaces not tabs" editor setting....

Sorry, that isn't Rant Number Two... This is..

Bad code is endemic in functions originating in a mathematical or engineering domain.

Why?

Because, like all Good Mathematicians, we expand the terms in an
expression as much as possible looking for further opportunities for
simplifications / cancellations / application of identities / theorems
etc.

What engineers and mathematicians _never_ do is...

Once they have the expansion as simple as possible, they forget to
reverse the expansion to make the expression as simple as
possible, invoking methods on the appropriate objects.

So instead of seeing simple, understandable, maintainable code,
scientific and mathematical computing contain dense nests of formulae
which are functions of several tens of variables.

You'd never make such as mess thinking as a Objected Oriented Designer
facing the problem as a Object Orient Design issue.

On which object do these attributes belong? On which object does this
method rightfully belong?

But somehow an expression coughed up by the Big Brain Engineer /
Mathematician / ... is sacred and is implemented, usually buggily, as
a huge monolithic, badly tested, function containing many stack
variables and subexpressions.

Rant over... (for now).

John Carter                             Phone : (64)(3) 358 6639
Tait Electronics                        Fax   : (64)(3) 359 4632
PO Box 1645 Christchurch                Email : [email protected]
New Zealand