Life in Balance: Amit D. Chaudhary’s Blog
• • • • •

When discussing whether one on one or in a meeting in my role as a software developer, I noticed it has helped me to keep handy the data behind a request.

The following is a made up example:

We should check to see the rpm hang bug in Fedora Core 1 is now fixed, before using it for package verification in scripts. I remember running into it in a project around a year back.

is much better (in terms of what to check, what is the reasonging that anyone can verify) than

We should not use rpm to verify package state as rpm hangs all the time and the script will never return.

which really is the speakers’s opinion based on hearsay.
In most of these cases, it would take less time to confirm one way or the other than to discuss or speculate.
Incase of Computer Science it helps in remember, it is an exact science, it either works or not, a feature does something or not. (Well, except for those hard to reproduce bugs and estimates)

Some background articles:
1 . This is one the same lines as used by similar to Wikipedia requiring verifiable sources.

2. It is #6 of the Google principles of innovation as per Marissa Mayer, VP Google

Don’t politic, use data

Mayer discourages the use of “I like” in meetings, pushing staffers to use metrics

3. Steve Pavlina’s: Stop Debating and Start Testing

There are a number of reasonable risks to be taken where the negative consequences of failure are negligible, but the potential upside is considerable if things work out. That’s the time when it’s often a good idea to dive right in and risk making a mistake, even if it’s probable that you’re wrong.
For example, consider the decision many bloggers face of whether or not to put ads on their blogs. Some bloggers type many words discussing and debating this matter, trying to decide whether or not it’s a good idea for them. Such words are wasted. With programs like Adsense, it’s so easy to just put up some ads and test them in a matter of minutes. Then you can see what kind of feedback and income you’re generating and have all the facts in front of you to make an informed decision. Guesswork, polling, and debate are pointless when it’s so easy to test something and get the facts.