Schema discovery could discover more

Jul 7, 2007 at 8:45 PM
IMHO I think a big improvement over the current implementation would be to include as much information as possible in the schema discovery, so that more information would be available if modifying the t4 templates is desired. This way validation rules, additional lookups, etc. could be accomplished much easier.
Jul 7, 2007 at 11:35 PM

katokay wrote:
IMHO I think a big improvement over the current implementation would be to include as much information as possible in the schema discovery, so that more information would be available if modifying the t4 templates is desired. This way validation rules, additional lookups, etc. could be accomplished much easier.

The current design of the SchemaDiscovery is a tradeoff between providing information and being database independent. We definitely chose on the side of db independence for most things.

What's the most useful information that you think is missing?

Thanks,

-Chris
Jul 8, 2007 at 7:07 PM
After looking at the code again, the most useful information that is missing, isn't necessarily missing from the discovery process. It just doesn't get included with the object model. If a column is a foreign key / or has foreign keys you could associate that with perhaps the column (with something other than IsForeignKey)? Also expressing the type of relationship, so related entity properties could easily be output. There is probably a tradeoff with this project I'm assuming about what is worth coding now vs. what comes out of the box with .net 3.5? It might be useful to theorize how this project will utilize what is provided with 3.5 so requests like mine might be more sympathetic to the future.