How security holes happen

[Ben Finney]

Steven D'Aprano <steve at pearwood.info> writes:

> On Wed, 05 Mar 2014 08:37:42 +0200, Marko Rauhamaa wrote:
>
> > If you had tried Python 30 years ago, you'd give it up for any
> > serious work because it would be so slow and consume so much memory.
>
> /facepalm
>
> Python is only 23 years old, so it would have been a good trick to have 
> tried it 30 years ago. While it was slow back then, it used LESS memory, 
> not more.

Moreover, this is not an issue of Python the language as much as
*implementations* (the CPython implementation has improved markedly in
the intervening decades), and of *resources* very different then and
now.

The available CPU and memory resources for a language implementation is
vastly greater today than 30 years ago. You could re-implement exactly
the same compiler today as was run 30 years ago, and have its speed and
memory performance remarkably better without any change in the language.

If you'd run an implementation of *any* language of the time 30 years
ago, it would have been far slower than implementations on today's
hardware, and doubless improvements in the implementation (if the
community was motivated to improve it for that long) would account for
even greater speed differences.

None of this is argument in favour of the changing applicability of the
*language*, which is what Marko apparently wants to imply.

-- 
 \          “I got an answering machine for my phone. Now when someone |
  `\      calls me up and I'm not home, they get a recording of a busy |
_o__)                                          signal.” —Steven Wright |
Ben Finney