What can Nuitka do?

[Rustom Mody]

On Saturday, June 28, 2014 5:14:39 AM UTC+5:30, Paul Sokolovsky wrote:
> Hello,

> On Fri, 27 Jun 2014 16:25:02 -0700 (PDT)
> CM wrote:

> > (Trying again, simpler and cleaner post)
> > Can I use Nuitka to transform a wxPython 
> > GUI application in Python that uses several
> > 3rd party modules into a small and faster
> > compiled-to-C executable?

> Yes, you can. So, please try that, and report how that went. We're
> eager to know how that would go very much. But unlike you, we don't
> have need to transform wxPython GUI application in Python into
> an executable. So, you are in the best position to answer your question.

> And surprisingly, both Nuitka and PyPy are free, so you won't need to
> shell out few $Ks to try it, or wear your suit, hop on the plane, and
> go to another side of the country (or world) and spend many hours in
> negotiations to get an evaluation license. Just download and try.
> Unbelievable, how world has changed in some 30-40 years!


All true...
However programming is engineering and programmers can suffer severe
paradox of choice if there's too much to choose from.

Going back along a different track by a couple of decades:

20 years ago python pushed the envelope of being a sweet combination of
being a serious system programming+lightweight scripting language at the same
time.

Today there are other choices¹:

- There are new languages that outright set out to be be as
efficient as C/FORTRAN -- eg Julia.

- A whole crop of new languages that are striving to plough back (some of)
scripting convenience into the compiled language mold -- eg Swift, Go

Languages that handle modern multicore hardware appropriately -- Erlang

And finally functional programming has come of age. So eg one of Haskell's
agendas is to beat C/Fortran's performance:

| Remarkably, our benchmarks show that choosing the proper stream
| representations can result in machine code that beats
| compiler-vectorized C, and that is competitive with hand-tuned
| assembly. --
|   From  http://research.microsoft.com/en-us/um/people/simonpj/papers/ndp/haskell-beats-C.pdf

tl;dr:

If no one speaks up (with hard specific data!) for the technologies
you are considering (eg PyPy, Nuitka etc) then I would conclude that
they are not yet ready for prime-time/ your use-case


¹ Yeah that seemingly adds to the paradox of choice problem!!