Eric and I pushing up some code for semi-live testing on the Herbo server. We then rolled into a discussion on the mini aggregate project for inventory quantities. We were drawing and going over ideas and options. Eric was talking about the difference between snapshots (current view of something without dates) vs normal aggregates with dates and possible roll-ups. Great conversation. We spent some time talking about flow, options, timing, and how the project would come together.

We have been calling this project the mini aggregate project, but it may end up being called the inventory snapshot project. It all deals with the quantities per location and how we are going to track it. Our final solution may end up with a process table that gets a new record when a new data or inventory snapshot is finished. We will then use the current snapshot linked with a well indexed (daily or just today's data) query of the live data transactions. That will give us the inventory quantities per location that we are looking for. It will save tons of time as the system won't have to historically go back over time to get the previous quantities.

Lots of talk about the process tables and basically waking up a built-in process. Have it check the process table and then do the requested action. When done, it makes a record in the process table and then goes back to sleep. Automated tasks and scheduled task stuff.

At the end of the meeting, John and I looked at his site and worked on setting up special top header level links and icons for his account.