On Wed, Jul 15, 2026 at 11:41 AM jian he <[email protected]> wrote: > We can document that this approach provides an error-safe way to > evaluate these to_* functions. > (After we refactored these to_* function to be error safe).
That's true. Of course, we could also just have variants of those functions that return NULL instead of signaling an error, or whatever we want the behavior to be. I don't know what the right thing to do is here. I'm not deeply opposed to this feature, but I'm not deeply in favor of it either. There's no big problem with it, but also no particular reason to do it. -- Robert Haas EDB: http://www.enterprisedb.com
