re #1, I can't get past "1million rows". I would never dream about pulling that much data across teh wire into an application. The point of using notracking is to use EF to retrieve data that is read-only. No need to waste resources asking EF to track that data. I recall some unexecpted (to me) but expected to EF behavior where it can lazy load from objects that were queried with no tracking. I cannot remember...I'd have to go look in one of my books (yes, I have to do taht) or just try it again.
re#2: dependent on what that behavior is about notracked and lazy or possibly exlicitly loaded objects that I just mentioned.
re 3... Acch...that one I cannot answer off the top of my head. It would take some experimenting to answer. That's how I have to learn. If you have some time to do some testing on that one, I'd be interested in your discoveries. Sorry I just had a brief moment to stop by and those last two questions take a lot more thought.