>
>1.) Do you haved any comments concerning the use of characters ID's vs Integer ID's?
>

Integer PK's are supposed to be faster and take less space than character-based ones. However, MM's supports either. The new default is integer PKs (character-based keys were the standard before version 6.0).

>2.) I know the codebook uses lv_ and rv_ terminology for views. In mere mortals are local and remote views the primary vehicles for querying and updating data? (I use views exclusively in my desktop apps, but I am now using SPT for data querying on web apps and I am thinking of using SPT for desktop querying also -- but, of course, most of my desktop framework is based on using views)

MM's also uses lv_ and rv_ as prefixes for the views. It also supports SPT and using tables directly, and you can mix and match them in your business objects, as required. Basically, you can use whatever you're most comfortable with. The "standard" method in MM is to use views and the framework has some utilities that make it easy to upsize to a remote database (eg. SQL Server). You'd obviously lose that if you used the tables directly.