Met with Eric to look at some code. We chatted about some time zone offset questions and date/time stamps that a 3rd party was asking about. We then got into a small discussion about better data modeling (see below). Eric has worked with numerous companies that have built and destroyed whole systems. We talked about some pros and cons and where we are heading for fracture as well as ship A and ship B stuff. We then flipped over to the enterprise vendor catalog changes that Eric was working on. We did a small code review, light clean-up, and merged in some code.

Eric and I – talking about the data model (database stuff)

- Problem with the unique id's – too much reliance on the existing primary keys – switch it to combo primary keys

- Any time you repeat things, you may want to look at the data model

- Cross-corp issues – for example – cross-corp loyalty points

- Po/Invoice lines – database table – breaking large datasets into corp-specific tables – that's how we solved the performance issues – there may be a better way to do that

- Better DB conventions vs table naming for corp-specific tables

- Enterprise level – aggregates and query data

- Shareable data models – associated tables

- Problems with text-based fields that combine records – for example: customer phone numbers

- Lots of talk about associations and setting up those relationships – master lists and who has access to that data, association, and/or relationships

- Refactoring things as we are in there working on things – smaller nibbles – iteration process – keeping chipping away at it along the way

- If we have a multi-page flow - make sure we use a common standard vs every page being different. A list of set standards.

- ETL - extract, transform, load - This is part of data warehousing and we would love to do it along the way vs just at the very end. Let's plan it into the fracture or adilas lite project.