Linked-in tables from other diagrams

[kgish]

I would like to indicate tables visually that have been linked in from other external diagrams in such a way that they are clearly shown, either as a different color, dashed lines or whatever.

Is this possible, and if yes then how?

[Bruno.Cossi]

Hi,

you can't highlight tables from another diagram for a number of reasons, one of them being that a table (or any element for that matter) can reside on multiple diagrams. You can however highlight the tables that were brought onto your diagram from another package. Simply go to the diagram properties and make sure the option "Highlight foreign objects" is checked. Then tables from a package different than the one the diagram is stored in will be titles as package::table.

Hope this helps!
Bruno

I would like to indicate tables visually that have been linked in from other external diagrams in such a way that they are clearly shown, either as a different color, dashed lines or whatever.

Is this possible, and if yes then how?

[kgish]

Well actually, the diagrams are all from the same package, and I want to highlight those tables coming from different diagrams within the same package.

I am new to this business of modelling techniques (sorry) and was wondering if this is the standard method of dividing up a huge data model into inter-related component categories.

For instance, I have: products, contacts, subscriptions, stock, discounts, defects, tickets, orders, etc.

Should I create different data model packages (assuming I can interlink the tables across packages, or is it recommended to keep everything within the same package?

One last question: is it possible to to a kind of drilldown where at the highest level all data categories (diagrams) are shown with the inter-related links, while within each category it is possble to zoom in to the next level?

[Bruno.Cossi]

Hi,

the traditional data modeling tools have a concept of submodels, that allow you to break your model down into smaller parts (e.g. security, client etc.). You can do the same in EA by using packages. Typically what I do, I reverse engineer a database into EA, at this point I have one large diagram with all of the tables.
Then I create a package for each of the "submodels" and simply drag the tables into the relevant packages. At this point I still have one large diagram with all the tables, but it also shows me which package (i.e. submodel) each table resides in.
Then, if necessary, I create a diagram in each of the packages (submodels) and drag and drop only the relevant tables onto it. Thus I have not only one large diagram describing the whole database, but also a set of smaller diagrams describing only the relevant portions of the database.

Hope this helps!
Bruno

Well actually, the diagrams are all from the same package, and I want to highlight those tables coming from different diagrams within the same package.

I am new to this business of modelling techniques (sorry) and was wondering if this is the standard method of dividing up a huge data model into inter-related component categories.

For instance, I have: products, contacts, subscriptions, stock, discounts, defects, tickets, orders, etc.

Should I create different data model packages (assuming I can interlink the tables across packages, or is it recommended to keep everything within the same package?

One last question: is it possible to to a kind of drilldown where at the highest level all data categories (diagrams) are shown with the inter-related links, while within each category it is possble to zoom in to the next level?

[kgish]

Yes this does help alot. I will give it a try and see how it goes. Thanks a million.

[Bruno.Cossi]

No problem, let me know if you need any help!

Bruno