Modeling hell

The heading for this post can be read in many ways: sometimes it feels like you're modeling Hell and sometimes it's hell to model something. What I'm modeling right now is something that affects everything in our system: from A to Z. And as soon I dig into something it turns out to be a new pit in hell. I'm just glad that I'm starting to tackle this now: we won't implement the first version until March. But this will take time. And my primary tools are:

1. Don't rush it. I knew this was big so I'm starting now and I'm settled on letting it take time.
2. Occupy a white board. We have plenty of space as from today so I sketch something, swear about it to the team and get some new ideas. Then I let is settle on the white board and get back to it tomorrow.
3. Don't use the current domain. I'm ignoring the rest of the domain, just focus on the problem in this case. When this feels comfortable, I will integrate the thoughts into the current domain.
4. Be open for the possibility that there are no perfect solution. It's not like this is a riddle where there is one good answer to all our questions: we will have to select which functionality we WON'T support as much as we will have to decide what to support.

Well, the day is almost up and I'm heading home for the little guy and his little girlfriend, who is spending the evening with us. I need something else on my hands and that is probably my final tip of the day: when you're in modeling hell, start thinking about something else.