Python vs C++

[Chris Angelico]

On Fri, Aug 22, 2014 at 4:12 PM, dieter <dieter at handshake.de> wrote:
> Likely, you would not use Python to implement most parts of an
> operating system (where, for efficiency reasons, some parts
> are even implemented in an assembler language).
>
> I can imagine that the GNU compiler developers, too, had good
> reasons to implement them in C rather than a scripting language.
> It makes a huge difference whether you wait one or several hours
> before a large system is built.
>
> "firefox", too, seems to be implemented in C/C++. There, too, I
> see good reasons:
>   *  it is nice when your pages are rendered quickly
>
>   *  "firefox" depends on lots of external libraries, all of them
>      with C/C++ interfaces; while is is possible to create
>      Python bindings for them, this is quite some work
>
>   *  as it is, "firefox" is a huge "memory eater"; one might
>      fear that things would be worse if implemented in a
>      higher level language (with everything on the heap).
>      Though, the fear might not be justified.
>
>
> All these examples are really large projects. I like Python a lot
> for smaller projects.

Yep. But it's not so much large vs small projects; these are all
situations where it's entirely plausible to saturate a CPU core for a
while, so the performance penalty of a high level language will
actually matter. (That said, though: Firefox has a lot of
non-performance-critical code, which AIUI is implemented in a higher
level language than C. And a long 'make' run probably involves some
shell scripting and such, not pure C code. Even high performance
projects have their less-critical parts.) Similarly, I'd be pretty
worried if someone rewrote X11 in Python. But if your program's going
to spend most of its time waiting (for the user, for the network, for
the real-time clock), halving its CPU usage won't materially affect
anything, and halving development time will make a huge difference.
That's when Python is an excellent choice.

At my last job, we had some parts written in Pike, some written in PHP
(not just the web site), and one back-end engine in C++. Toward the
end, I wanted to redo the C++ engine in Python or Pike, because I
didn't think the performance hit would be at all visible, and it would
have been much easier to work on. (I'd also been progressively taking
over jobs that PHP code had been doing and giving them to Pike
processes instead, with a not-so-secret goal of eventually ripping out
the PHP engine altogether. But that for other reasons.) As a back-end
process, its performance would be hard for the end user to see
directly, so the cost of a high level language would simply have been
that the server could cope with fewer requests per hour before we need
to requisition more hardware. Alas, there wasn't development time
available for the translation (we were far too busy doing things about
which I said "YAGNI" and the boss said "Do it anyway"), but I firmly
believe that we wouldn't have been suffering from the performance hit.

ChrisA