When a show has been finished for a while it gets archived. Nothing is lost, and one click loads it back in. That click used to mean a wait, and on a big show a long one. It is a lot quicker now.
We rewrote how a backup is read and written. The whole file used to be held in memory before a single row was touched. It now goes through a row at a time, which is both faster and a fraction of the memory.
That is what makes the bigger change possible. Archiving and loading used to need a large server that we started for the job and shut down again afterwards. It does not need one any more, so the servers stay up and wait for work instead. Your show starts loading the moment you click.
One more thing while we were in there. Clicking twice joins the job already running instead of starting a second one, both when you load a show and when you archive it.
Nothing changes in how you do it. Open the archived show, click OK, and it is back.