Subject: Lines of Code and Man-power estimate for Maxima
From: dlakelan
Date: Thu, 22 Apr 2010 11:16:06 -0700
On 04/20/2010 09:55 AM, Richard Fateman wrote:
> Do I understand that this says
>
> You should be able to write Maxima in 3 years with 32 programmers at a
> cost of $13 million dollars?
>
> Somehow I don't think this works out for issues like: how long does it take
> to write a program that no one knew how to write, but figured out, in
> the course of earning a PhD?
>
> Also the idea that one can hire 32 programmers and have them do useful
> work starting
> at day 1, seems unworkable. Maybe the SLOCCount "methodology" attempts to
> account for it.
I think things like SLOCCount are generally targeting only "business"
development, where what needs to be done is for some users to figure out
what trivial tasks they want automated and programmers to figure out how
to write code to automate those trivial tasks and perhaps a few years of
back-and-forth between the users and the developers as people change
their mind about what it should do.
The idea that even if you have a perfectly exactly specified requirement
that it might take 300 years of continuous mathematical theory
development before you could start coding is totally absent :-)