Lambda: the Ultimate Design Flaw

[Terry Reedy]

To summarize, instead of saying "Python has only one way to do it",

As I explained in response to Aahz, what Tim Peters wrote was that Python
'should preferably have only one obvious way to do it'. Omission of the
crucial qualifiers 'should preferably' and 'obvious' misleads any
discussion.

rather say "Python will eventually have only one way to do it",
and with such a wording, nobody will not be mislead.

The actual design principle, as opposed to the impossible
oversimplification, does not, in my opinion, mislead. It is applied to
every new proposal, most of which get rejected. What I can't tell is
whether you wish Python had added less new stuff or had already dumped more
old stuff. For myself, I wish the next version would be 3.0 and slimmed
down a bit.

Terry J. Reedy

 

[Terry Reedy]

Python has one obvious best way to do things.

More exactly, 'should preferably have' rather than 'has'.

Meaning that the most obvious and clearest way, the way that comes to
mind first, will in most cases, also be the best way.

I seem to remember reading it put in that way some place at some time.

tjr

 

[Philippa Cowderoy]

While continuations are a very interesting abstraction, the improvement
of structured programming was to be able to prove properties of your
programs in time linear to the size of the program instead of quadratic.
I don't see how giving arguments to the GOTO would help there.

By allowing you to build your own control structures, whose properties you
prove once before using them to prove properties in the programs that use
them.

 

[Ron_Adam]

More exactly, 'should preferably have' rather than 'has'.


tjr

;-)

Ron