Not sure who to reply to, you or kj_norway.
What kj_norway has introduced is another reason why I rarely use the Ancestry web interface for data entry. My workflow is roughly as follows:
I use Ancestry.com to locate new records, including clicking on leaves.
When I find a new record, I add it to my shoebox at Ancestry, then download the image to my PC.
I copy-n-paste the bibliographic info from Ancestry to a logsheet on which I track all my sources.
In FTM'12 I create a source record for the new source (if one doesn't already exist), and a citation detail. It's important to do this manually because Ancestry.com's source citations are so terribly, horrendously WRONG (a fact which serious genealogists have been calling Ancestry's attention to for years, to no avail), and I don't want that junk cluttering up my database.
I attach the downloaded image to either the source record or the citation detail I created, whichever seems more appropriate.
I attach the citation detail to every relevant fact.
Finally, when I'm satisfied with the result, I sync it UP to Ancestry.com. I NEVER sync down from Ancestry. I've spent way too many hours cleaning up the garbage that gets dumped in my DB.
As should be obvious, I consider the database on my hard drive my main database (so hands off, Ancestry :-) ). I sync it up to Ancestry both as a research aid and to share what I have publically, but I don't rely on the online database either as my primary DB or as a backup. The thought of having to clean that thing up gives me the heebie-jeebies.