Localizing Actions for Trees...

What is the proper way to localize the captions for Actions?

com.vaadin.event.Action classes have only a caption and an optional icon. This doesn’t work so well when dealing w/ localization, as the handleAction() would need to know the localization of the caption to determine which action was actually invoked…

I mean, I may be missing something here (I’m actually working on my first forray into Vaadin localization), so please clarify if you can…