OK, I'm not describing this clearly. I have a workbook (W) on a server. A user provides inputs which are fed into (W) to create a new version of (W).
(W)=>(I)=>(O) But I have many users accessing (W), all with their own (I)s: |//====>(I)\ |/====>(I)\\ (W)===>(I)=(O) |\====>(I)// |\\====>(I)/ And, in fact, the (I)s are updated every few seconds. Initially, what I was doing was, every update, I would load (W) from the disk, parse it, put in the inputs, and grab the outputs. But this is more disk intensive than I would like. So instead, I loaded (W) once, and just reused it over and over again. Now, =in theory=, since the (I)s are the only fields the users can change, and they must change them all every time, there's no issue with reusing the same (W) over and over again. But I sure would feel better if I had a clean (W) that I could clone every time I got a new set of (I)s, and discard once the user had the (O)s. Does that make sense? ===Blake===
