perl to python

[Heather Coppersmith]

On 13 May 2004 08:37:00 +0300,
Ville Vainio <ville at spammers.com> wrote:

> Because Python code is more readable and maintainable. That is a
> huge win in multi-maintainer situations (which is typical in
> production code) ...

Agreed.

> ... You can also bump up the functionality of the script while
> porting, and get better reuse.

That part makes me cringe.

Port first.  Make sure the new code still passes every *old* test
you can find.  Write new tests against the current functionality
and the new code; porting inevitably leads to new corner cases and
language idiosyncrasies.

Bump the functionality later.

If I'm the manager and I hear, "feature X, which wasn't even in
the old code, is almost working; but featyre Y, which was in the
old code, seems to be broken," then heads will roll.

Obviously, e.g., porting an OS from one hardware platform to
another (which I've done, more than once) will lead to
functionality changes, but there had better not be any bumping
until you know that the port is at least as solid as the original.

Most of the time, though, it's the other way around:  you find
bugs in the old code due to the new scrutiny, but then you have
the nasty problem of determining whether or not something else
depends on the buggy behavior (but that's a better topic for a new
thread on another newsgroup).  At that point, bumping the
functionality during the port could be extremely troublesome.

Regards,
Heather

-- 
Heather Coppersmith
That's not right; that's not even wrong. -- Wolfgang Pauli