[phenixbb] Support for residues on symmetry elements - first reaction

[Ralf W. Grosse-Kunstleve]

Hi Felix,

> However I do not get what it is ==1

That's short for "is equal to". -- Use in many programming languages
to distinguish assignment and test for equality.

> and !=1

"is not equal to"

> or it is encoding striking back?

Yes, sorry, that's the world I live in ...

> And to the last question: yes, it is good to keep
> coordinates constrained to stay on the symmetry element. It should be
> a default.

It was the default in the past. I could make it the default again as
a special case for single-atom residues, but I found quite a few HOH
in the PDB where the water was near the symmetry axis but clearly not
on the symmetry axis. If I change phenix.refine back for single-atom
residues it will again not be compatible with other programs and many
entries in the PDB.
Ideas:
  - Using atom selections to tell phenix.refine which atoms with
    partial occupancy to constrain to a symmetry axis, e.g.
      symmetry_constraints = chain S and resid 101
  - Using negative occupancy as a flag to request symmetry constraints.
The first alternative would mean work for the user and open the door
to oversights, the second would confuse other programs (although they
could easily be adjusted).

Ralf