And gate or Dummy script task

OK? It will technically work, but is less expressive and can be harder to understand. This increases your long term maintenance cost.

Stefan Helzle said:

less expressive and can be harder to understand

Yes, this is my usual objection to this approach.  There are cases in my older, more complex process models for instance, designed by less experienced devs, where a flow split occurs like this and I'm not quite sure whether it was intentional or accidental, since they didn't use an AND gateway to do the split.

My stance: there's very little cost to using an AND at the front of a split, and it's nearly always worth it in the long run.  If you're worried about the node performing some sort of calculation or pre-processing, the AND can simply be preceded with a script task that does this other work, with virtually zero additional cost.

Stefan Helzle said:

less expressive and can be harder to understand

Yes, this is my usual objection to this approach.  There are cases in my older, more complex process models for instance, designed by less experienced devs, where a flow split occurs like this and I'm not quite sure whether it was intentional or accidental, since they didn't use an AND gateway to do the split.

My stance: there's very little cost to using an AND at the front of a split, and it's nearly always worth it in the long run.  If you're worried about the node performing some sort of calculation or pre-processing, the AND can simply be preceded with a script task that does this other work, with virtually zero additional cost.