Highly recommend two essays by Co-burn

[Doug Wang]

Hi all,
Here are two articles by Co-burn, Just-In-Time Methodology Construction
and A Methodology Per Project, which show the most important ideas of his: Projects and teams differ, so you need dynamic methodology, and one methodology per project.
The following depicts the Crystal methodologies named by color: clear, yellow, orange, red, maroon, blue, violet. The number indicates team size.

<div align="center">3-8</div>

<div align="center">10-20</div>

<div align="center">25-50</div>

<div align="center">50-100</div>

<div align="center">100-200</div>

<div align="center">200-500</div>

<div align="center">800+</div>

[ February 21, 2002: Message edited by: Doug Wang ]

[Alistair Cockburn]

Thanks for the pronunciation tip, Doug :-)
Alistair

[Doug Wang]

Hello Alistair,
Methodology-per-project is really great idea. But could you elaborate more on choosing suitable methodology among Crystal family, XP and RUP, etc. ?

[Alistair Cockburn]

Think I put it on another thread, don't know which any more (trying to consolidate some of these threads)...
My current thinking is that methodologies only live in the past: "We did this, and it worked for us, and we liked it."
Therefore, there should be a library of projects and what they did (successfully, and liked). From those, one would find the project(s) most similar to one's own, and copy / modify from that.
I could imagine a dozen or so such examples: e.g. Y2K Conversion, Submarine, invoicing, etc.