Symmetry factor for processes with decay chains

[Frank Siegert]

When specifying a process where two identical final state particles are produced in two separate decay chains, the symmetry factor ignores those two identical particles, and only multiplies the symmetry factors at each step of the decay chains. One can of course argue that, when decays are specified, their decay products should be distinguishable. But I see the decay specification as a selection of Feynman graphs, and there I don't see any use case for ignoring that these particles are indistinguishable.

[Frank Krauss]

Strictly speaking, what we do is gauge invariant only in the narrow width approximation, and there the particles are distinguishable. So it makes sense to ignore the symmetry factor there. I do not think that this is a defect, but correct.

[Frank Siegert]

This turned out to be a bug after all, and was fixed in r19281. Targetting it for 1.4.1.

[Frank Siegert]

Committed into rel-1-4-1 (r19851).