Adding a dependent job with --rerun results in unexpected behaviour

[sb10]

echo "sleep 10" | wr add -o 2 -i a --dep_grps a --rerun && echo "sleep 5" | wr add -o 2 -i b --dep_grps a,b --rerun && echo "sleep 2" | wr add -o 2 -i c --deps a,b --rerun

On a fresh db, c is dependent, runs last.

Previously, when repeated, a and b are added, but c is not added since it's a "duplicate". It did, however, get resurrected and start running immediately!

This is now fixed, but want to change it to use the newly added job and not claim it's a duplicate, since --rerun was used.