Process Creation Process

Process Creation Process - kind of recursive :)

Being obsessed with open-source development 'methodologies' ( I hate to use BuZZ words .. but still ), here I am, again trying to do things in MY WAY ... fighitng RUP-IS-GOOD-EVEN-FOR-MY-GRANNY world.

After extensive reading (I am just 52Kgs, so your extensive might mean something else) of PeopleWare, XtremeProgramming, and Agile methods, I started to think about a 'process' that gives me a 'metric' so that I can satisfy Metric-Driven Project-Manager, still keep my mind in the tao without dumping RUP junk into it.

Process that works, and delivers bug-free software meeting every milestone and release dates is still a HolyGrail of the software development. There is no way one can Identify the Common solution for all types of software projects.

Here's how my way of Process Creation:

Steps:

• Definition
• Feasibility,
• Automation
• Training
• Implementation.

You need to have a project to create a process :) Just like that you need people to define the process, afterall , they are going to 'process'ed finally.
A good process needs to have

1. Relevance with the Project (Technically)
2. Relevance to the People(a.k.a project stake holders).
3. Right tools to automate
4. Zero interception with the core-project-tasks.
5. Real Time Metric

Definition:

First the terminology, the process needs a consistent terminology, when I say "Spade", I mean It's a "Spade", and all stakeholders understan it to be a "Spade".

After the terminology has been consistent meaning, define the metric for the process, that gives quantitave information about the project status at any given isntance. Then the next step would be to study the feasibility.

Feasibility:

Feasiblity involves simple questions:

1. Is this good for my team?
2. Is this good for my project?
3. Is this going to give me data what I expect from it, with much less effort?
4. Is this transperent enough, so that all developers/testers understand this?
5. Can your manager understand this?

if we have Yes(1), Yes(2), Yes(3) and Yes(4) then its the time to go to the next step. Yes, you are right, you can ignore the answer to the fifth question, because we know the answer even you ask the question, don't we? :)

Automation:

After defining the process and finding that it is feasible, next step is to find some good tools that augment your process and automate most possible number of steps. And evaluate them, and settle down on the best possible tools, that have near-vertical learning curve for our team. And make sure that the tools assist you generating the metric we always wanted.

Training:

The next step, explain the process to the teams, make them understant the importance of the process and the Process itself. Start training them on the tools, simulate the process once, or twice so they get a hold of it.

Implementation:

Once the training is over, setup the Process and integrate it into the project work-flow, and see the project go like breeze.

And you'll never miss to call, miss to reply, and most of all you'll never miss life :)

This the abstract guide to Process creation process .. more concrete example coming soon.

April 1st

I was looking at all the funny slashdot april 1st pranks ..

This is wonderful: Ygoohoogle :)