Bi-temporal Anchor Modeling: Page 2

Bi-temporal Anchor Modeling

Roenbaeck

I've pushed a new release to test (rev. 342) so you can have a look at the generated bitemporal code there:

http://www.anchormodeling.com/modeler/test

Still not finished are bitemporal ties and the triggers, but anchors, attributes and all different views should work, as well as the insert triggers on the attributes. You have to figure out for yourself how to work with them. I'll add some tutorials when it's all completed.

Roenbaeck

I've made some mistakes that I am correcting now. I thought knots could be given a history over recording time, even if they cannot have one over changing time, but I was wrong. This effectively destroys referential integrity, and makes for an ugly theory. Knots will have no history.

Knots in bitemporal AM act as value storages, from which we can pick values, much like an enumeration. It therefore also makes sense that values we at some point recognize as being "bad" should remain, and corrections instead added in the knot as new values.

So, the code currently in test will change when it comes to knots. I also had not added the insert trigger for attributes as I thought, so once that is added I will push a new release to test.

Roenbaeck

Bitemporal Anchor Modeling is now ready for testing. Use the test version of the modeler found here to generate bitemporal implementations for SQL Server:

http://www.anchormodeling.com/modeler/test/

Let us know how testing goes! :)

« Previous Page