Don't make assumptions.
I can do garbage me
Every day I learn more and yet I sometimes get the feeling that the more I learn, the larger the projects, the more bugs slip in because of assumptions. Right now I'm on such a project. It simply doesn't get of the ground because of assumptions. Nothing more than assumptions!
You can read it in books - and not only the books about managing software projects, but the books about management in general -, in blogpost, articles, MOC material. Don't make assumptions! And still I - and a lot of people with me for that matter - get fooled by them over and over again.
This BI project I'm working on right now for example should have been finished in two weeks, but we made a few assumptions. So it's keeping me busy for almost two months. I'll list the assumptions here and leave it up to your fantasy what went wrong.
The assumptions
We / I assumed:
- that a certain set of records in our source would not change ever;
- that if a certain record changed, the update_date field would be updated;
- that our datawarehouse would not be used by others to store some userdata that has nothing to do with it at all:
- that the data would be used for production only when the data was verified and agreed upon by our customer.
What have I learned
- Don't trust anybody: check;
- Communicate;
- Don't give somebody administrator rights on your datawarehouse because he owns the system and paid for the datawarehouse;
- Don't assume anything:
- Don't assume anything ever.
We'll probably release this project tomorrow and it will go in my books as the project where I learned the most. Because when I look back at it, it's all my own fault...