There is something about FDS which has always seemed a bit ackward to me. This is the fact that one must first make changes to the FDS managed collection to be able to commit these changes. This in itself is fine, but this means that if I have a grid which displays the managed collection, the user must first modify this data and commit the changes to then see the changes revert back to the original values if the transaction failed. (This does not hold true for createItem(), because for some reason yet unknown to me, when one calls this method, it does not reflect on the managed collection...any ideas as to why?).
So the user sees the data displayed change, but on error, it sees it change back. To me, it would seem more logical to see the data displayed change only if the transaction was succesful. Maybe this is because I am used to RemoteObject, which calls the service, and on "result" is where one does the actual modification to the data displayed, not before. Does anyone else find this as a strange flow of events? Or are there any ways around this? One could make a temporary managed collection with the same exact data as the displayed collection, and make changes to this collection, commit them, and if successful, then make the same changes to the displayed collection. This way the user would only see the data displayed change if the commit was successful. But this seems much more trouble than its worth. Any ideas or suggestions would be appreciated. Thank you!
