What are the changes you made?

I discovered that if I upload changes that I make to the current version, they replace that version on the server with no notice to anyone who already has it. So different people had different versions of 1.0, without any notification that they were different. I uploaded the 2.0 version to make sure everyone got the current changes. So it depended on when you downloaded the v1.0 version on how much different it is from 2.0.

I had made a lot of minor changes to the code to make it more consistent. I also changed the reporting some. I found a few ways to make things more efficient, such as band_disable() with no arguments disables all the bands, without having to iterate through each band one at a time. This makes the code somewhat faster.

The basic algorithm, though, is the same.

All future uploads will change the version number and be a child of the previous version.

Yes, sometimes, it is more efficient no to do some more, but fewer things ;)

As the detection of hydrophobics is not prefect, I use a slight change of your version to add them manually:
http://pastebin.de/7534
This example includes segments 1,2,3,4 and 5 as example to the detected ones.

I'm thinking about changing the script for also excluding segments.

I added the feature to exclude Segments.
This example includes segments 1-5 and excludes 4-10 to show that excluding is dominant:
http://pastebin.de/7536

This would help in processing if we can get the information later if a segment is hydrophobic or not.

http://pastebin.de/7539

{1,66,76} forces these segmetns also to be banded,
{75,78} excludes these hydrophilic sidechains.

Change "{1,66,76},{75,78}"to "{},{}" to get normal behaviour, and compare banding.

Copy the non-colored code of the link (bottom code) back to fold.it to test it.

With some additional information from the foldit team, I am now able to reliably tell which segments are hydrophobic (see the separately published tlaloc get_aa() function script).