update and clarify the choice of atom ids for angle style dipole (which is not...

update and clarify the choice of atom ids for angle style dipole (which is not really an angle potential)

doc/src/angle_dipole.txt

@@ -96,16 +96,17 @@ USER-MISC package.  See the "Making LAMMPS"_Section_start.html#start_2_3
 section for more info on packages.
 
 NOTE: In the "Angles" section of the data file, the atom ID 'j'
-corresponding to the dipole to restrain must come before the atom ID
-of the reference atom 'i'. A third atom ID 'k' must also be provided,
-although 'k' is just a 'dummy' atom which can be any atom; it may be
-useful to choose a convention (e.g., 'k'='i') and adhere to it.  For
-example, if ID=1 for the dipolar atom to restrain, and ID=2 for the
-reference atom, the corresponding line in the "Angles" section of the
-data file would read: X X 1 2 2
+defining the direction of the dipole vector to restrain must come
+before the atom ID of the reference atom 'i'. A third atom ID 'k' must
+also be provided to comply with the requirement of a valid angle
+definition. This atom ID k should be chosen to be that of an atom
+bonded to atom 'i' to avoid errors with "lost angle atoms" when running
+in parallel. Since the LAMMPS code checks for valid angle definitions,
+cannot use the same atom ID of either 'i' or 'j' (this was allowed
+and recommended with older LAMMPS versions).
 
 The "newton" command for intramolecular interactions must be "on"
-(which is the default).
+(which is the default except when using some accelerator packages).
 
 This angle style should not be used with SHAKE.