>Today we generally think about BYOC at the “app” level. But can we go finer-grained than that, picking individual interface elements? Maybe will will end up r

>Today we generally think about BYOC at the “app” level. But can we go finer-grained than that, picking individual interface elements? Maybe will will end up re-inventing OLE from the 1990's for the web. In OLE 2, there was a standard OLE Compound Document file storage API/format. Different apps could all work with different parts of a document. You could even change which app you wanted to handle a certain type of feature. It was complicated, but when it worked, it was pretty neat. This was done during the time when most people had maybe 4-8MB of RAM. With much more powerful CPUs and vast amounts of RAM, we could probably come up with something even better.