@ocharles @sfwc Do either of you have an opinion about this?

@ocharles @sfwc Do either of you have an opinion about this?

My opinion shouldn't carry any weight, but for what it's worth: now that I see it written up, I'm not sure that having this predefined really adds any value over id $?. I think I'd just vote for adding a note to the documentation. Unless projectMaybe can be defined in some other way that leads to better SQL in some cases? (I don't know whether there's ever any important difference between id $? and maybeTable null nullify, for instance.)

While I agree that it's understandable once you understand it, having it as a documented function that includes its implementation in the haddock means that hopefully the reader will find it in the documentation or via Hoogle, notice notice that it's a really simple function, and make a considered decision whether to bother using it.

Has anyone had a chance to look at this, and should we consider other ADT-esque tables like EitherTable for symmetry?