For review: PEP 308 - If-then-else expression

[Andrew Dalke]

Dale Strickland-Clark:
> This is such a daft argument. If the lowest common denominator is to
> dictate language complexity then we'd better all give up now.

  - this isn't 'lowest common deminator'
  - this isn't "dictating"
  - if the language gets too complex, then other languages will
appear which promote "easy for novice programmers to understand
but also usable by expert programmers, unlike that baroque Python
languages."

> To expect others to put up with a restricted language set so that a
> bunch of half-interested, non-programmers don't have to put in too
> much effort to understand it is staggering arrogance.

There are plenty of languages in the world.  My argument does not
apply to all of them.  It does not apply to C++.  It does not apply
to APL.  I have evalutated many languages as possiblities for my
target audience, and decided Python is the right one, partially
because of its simple syntax.  If it gets too complex, I will look for
other alternatives.  With this PEP, Python gets further and further
down that road.  Not complex by leaps and bounds, but by slow
steps.

> If you don't like a feature, don't use it but don't expect others to
> put up with your inadequacies.

If you read what I am saying, you will realize that I am not talking
about myself.

Should all languages get more and more complex so that people
with "inadequacies" cannot easily program?

> Read a book, for Christ's sake! You have to put in a bit of effort to
> understand something new. You can't expect to pick up a new language
> without a bit of effort.

I have.  A couple books even.  That's how I figured out how the
language is supposed to work.  That doesn't mean that I can understand
Lisp code by looking at it.  Yes, were I to spend more time on it,
I could.  But I've no desire.  We I to learn Lisp, I am convinced I would
not have that many people using my software to do new science.

Similarly, I read the Smalltalk 80 book and tried out Squeak.  Failed
misearably when I couldn't figure out how to type things into it.  Also
didn't like the language -- too objecty for my tastes, ;) and I didn't
like that new functions are added to classes, like adding a new "abs"
function to numbers.

> What do they do if they need some electronics built, some plastic
> moulded or metal worked? Do they get an expert in or have a go at
> doing it themselves?

Both.  I needed some hardware built, so got the department machinst
to do the hard parts (cutting and anodizing the aluminum) but I put
in the glassware and wires and other parts.

Some of the worst chem labs are run by grad students, who don't
have lab training and get about 2 days of safety training.

> I am all for people being self-sufficient and having a go but when
> things get too heavy for an amateur call in an expert - and learn from
> him/her.

I designed the aluminium part, but there's no way I'm going to be
a machinist.  I gave him the plans and got the part a week later.

I fail to understand your point though.  Should all labs hire an
expert programmer?  Or perhaps there should be a departmental
programmer?  At what point is the cutoff threshold to justify the
expert's cost over the relatively cheap grad students getting paid
1/4 or less the expert's wages?

                    Andrew
                    dalke at dalkescientific.com