If it's big then code is the last thing I write. First I draw a mental map of the project [probably also a whiteboard map]
I'll need answers to:
- How much of the project could be bought off-the-shelf?
- How much needs programming resources because there are no components ready to buy/adopt?
- If off-the-shelf is very expensive or non-modifiable how much to create a custom version?
- Were there similar projects in the past [how did they fare?]
- What's changed since then that could ensure success?
- If there's tech chops needed that I don't have, then either off to the library or see which of my software colleagues is already an expert.
- Downloading and testing existing apps might open you up to legal trouble down the road so be careful
Filling in a Wiki/Project plan is probably the next step along with a whitepaper outlining goals and non-goals.
[To Be Continued]