P*rl in Latin, whither Python?

Peter Hansen peter at engcorp.com
Tue Nov 14 04:19:46 CET 2000

Steve Lamb wrote:
>     Heh.  Now if only you could convince the powers that be in my company of
> that.  I'm script boy and after a few weeks of Python coding I'd much prefer
> to do my work in Python than in Perl.  Problem is that they have rightfully

I beg to differ.  They are very wrong.

> pointed out one thing.  While all my coworkers agree that Python would be a
> good thing they are all working in C and when it comes time for me to move on
> and they need to replace me finding skilled Perl programmers is far easier
> than finding skilled Python programmers.

This is completely backwards, wrong, and, not to mention, incorrect.  No
offense to your respectable powers that be, but their logic is at least
half flawed.

If you write your code in Perl, they *will* need to find skilled Perl
programmers to replace you and (shudder) to maintain your Perl code. 
They have little choice there, I agree.

But if you write your code in Python, they will have a much larger
selection of potential replacements: all skilled *programmers*,
including the subset which are skilled Perl programmers.  You do *not*
go looking for skilled Python programmers; you find good programmers
with attributes other than simple knowledge of Python, you invest
roughly one week, and you have people who can begin to maintain your
codebase.  You use them for maintenance for a month or so, and now you
have skilled Python programmers who can maintain the code far better
than if it were Perl, and begin extending and enhancing as specified by
your large and growing wish list.

>     Of course my Python experience being so limited I imagine I could spend
> the next 2 years on the same projects.  6 months learning Python, 6 months to
> rework everything once I've learned it.  Another 6-12 months to rerework it
> the way I /really/ want it.  :)

I've had, let's see, a total of eleven programmers, several of them
quite junior, learn Python over the last ten months.  In that time, not
a single one of them took longer than one week to be agle to write
something useful in the language.  Several of them have advanced, after
a month or two, to significant efforts.  And interestingly enough I can
look at the code *any* of them has produced and not only understand it,
but immediately maintain it myself, should I choose to do so.  

On another thread Alex writes cogently about competitive vs. network
effects.  This is a clear case where Python has a *huge* competitive
benefit.  Let your powers that be take that as they may.

Peter Hansen

More information about the Python-list mailing list