Execution names should be unique?

[stuartmcalpine]

When writing the tutorial for pipelines, I was thinking about how to query them, to get back the datasets.

I could start with the executions, search for it, and then get the datasets with that execution_id. I'm most likely to search on the execution name, however this at the moment is not unique, so I could get multiple results. Should we make this unique? The other option is to first search on the name, then deduce the correct ID, and use the ID from then on.