Limpet Restart

1. Strategy

1.1. Re-do Limpet storage framework

1.1.1. Switch from collection of Measurables to array of doubles, with single Units attribute on collection

1.1.2. Convert/verify units as data is being stored

1.2. Produce EMF Model for storage & operations

1.2.1. ?

1.2.1.1. Can we still make operations extensible at runtime?

2. Pros

2.1. Introduce EMF foundation

2.2. Benefit from other EMF technologies

2.2.1. EMF Forms

2.2.2. Sirius graphical editor

2.3. Reduced maintenance (less code)

2.4. Introduces option to become an Eclipse project

2.4.1. Only use approved libraries

2.4.2. All contributors complete questionnaire

2.4.3. All very scary though.

2.5. Reduce UoM complexity

2.5.1. Just an attribute on a collection, not on every datum

2.6. Improve performance

2.7. Chance to overcome initial design flaws

2.7.1. Only allowing Time as independent axis

3. Cons

3.1. Cost in effort

3.2. Delivery delayed

3.3. Doesn't bring concrete benefit to Debrief users

3.4. Doesn't bring concrete benefit to Limpet users

3.5. Challenge in learning new technology

3.6. Still require architecture/design solution

3.6.1. Multi-dimensional datasets

3.6.2. Allow dimensions other than time as independent (distance for weather balloon ascent)