I think we can close this one

Selective-max has been integrated.
So far the commandline option automatically uses LM-CUT + landmarks heuristic,
but this is just a matter of parsing.

We also think it might be best to wait with this integration until after the
first distribution, since these are mostly additions and not changes to the core
code.

One more thing to integrate - LM-A*

This is a meta-issue to keep track of the things that still need to be done for
the FDTech integration. The individual TODOs, in so far as they require more
than a little work, should be put into their own issues and referenced from this
one.

Here's the initial info from Erez:

What we still have to add from the Technion branch is:
1. Selective-Max (and some other learning stuff) - should be easy
2. Structural Patterns - requires action cost to be a double, we
should discuss this
3. LM Enriched - some changes need to be made for new landmark
implementation, but should be easy