In case we find two gcodes closer than 0.01mm in Z, we estimate the layer
height from the first layers and group gcodes based on this layer height. The
logic seems fine and the code seems to work, but it's nontrivial and I'm a bit
worried that it might break things.

Splitting prontserve __init__ and start into seperate methods so that the server shutdown works more reliably.

Adding a heaptrace option for memory profiling prontserve and updating tornado in an attempt to squawsh the prontserve memory leak.

Moving all websocket communications out of the read thread callbacks (now async). Fixing threading issues with prontserve.