Re: [Q] XP and Estimation

On 29 Apr 2005 01:10:33 -0700, ashimag@xxxxxxxx (Ashima Goswamy)
wrote:

>Just finished reading "Extreme Programming Explained", second edition
>by Kent Beck and wife.
>
>XP talks of estimating early , based on skimpy user stories, and yet
>talks of estimating accurately (page 4 of the book). But how do you
>estimate?

Two basic ways:

1) estimate stories proportionate to one another, in a ratio of at
most 1:3.
2) estimate stories in hours of pair programming it will take to
accomplish them.

Then use feedback (actual performance) to compare velocity predictions
to reality and refine estimates.

> Is this a practical approash?

Quite practical. Hundreds of teams are producing software on a
predictable schedule using XP.

>I would like to hear from XP
>practitioners, as to how do they actually do the estimation. Does
>anybody really handle estimation and planning in exactly this way?

Many teams have done it as described above, to good effect. (Some
elaboration needed, of course.)
>
>[Barring this point, otherwise, I found the book to be a nice
>collection of sensible practices: small iterations, test-driven,
>simple and incremental design, etc. are really good practices.]

For further information, you might wish to ask this question on
comp.software.extreme-programming.

I'd also suggest reading Beck and Fowler's /Planning Extreme
Programming/, or Jeffries, Hendrickson, and Anderson /Extreme
Programming Installed/. Also Mike Cohn's book on User Stories has some
good planning information, though I consider much of what he
recommends to be more complex than is needed. On the other hand, he's
a big guy and could kick my ***. :)

Regards,

--
Ron Jeffries
www.XProgramming.com
I'm giving the best advice I have. You get to decide if it's true for you.
.

Quantcast