[tom_mellior]

The premise of the original simulator was to show how (reasonably shaped) icebergs actually float. This one adds melting and says: "his new code actually takes material away from the surface of the iceberg in a uniform way. It works more like you would expect melting ice to behave." But this is not how I would expect a melting iceberg to behave at all, and I'm fairly sure that -- unlike the floating part -- this is not now icebergs actually melt.

There is no reason to believe that the air and the water around the iceberg are the same temperature or otherwise in a state where melting would be completely uniform. This changes a somewhat scientific simulation into an unscientific toy.

[runarberg]

I just consider this a toy (just like the previous implementation). The melting is a super fun feature indeed. Try drawing and upside down capital lambda spanning the entire canvas:

     _    _
    | |  //
    | | //
    | |//
    |__/

Try not to draw perfect lines. At first the iceberg has the two tops standing out, which is fun, then as it melts and starts breaking a part around the thinnest parts, it is super fun to see how it reacts. Eventually you will end up with a bunch of slim topice floating on the surface.

[nonpolitic]

How far do we want to take this simulation :-) it's also 2d, which is probably a more serious concern given the way the mass distribution reacts to each increase in dimensionality.

[JoBrad]

Version 3 will explore using AR to realistically melt icebergs.

We already have version 15 in beta testing. It’s the Arctic. Melting seems to have gone too far. Should be fixed in the prod release.

[ranit]

Unfortunately this version 15 is deployed in production already and it will be very difficult to fix it :-(

[capableweb]

The initial simulation was nothing more than a toy either. The world is not 2D and to assume so when considering physics, seems not very scientific.

This is just another toy built on top of a toy.

[ineedasername]

There were also no penguins or polar bears, which might be possible in the real world but doesn't seem probably after drawing at least 10,000 large icebergs. None showed up. So, not biologically realistic either. I'm off to draw another 10,000 now to if their impact on local weather pattern is appropriately modeled.

[tinus_hn]

The original simulation featured negative weight areas.