On Mar 23, 2016, at 12:03 AM, Stephen J. Turnbull wrote:
I'm with Mark here.
Okay cool, I think we're all in agreement about how this should behave.
I think the right thing to do is to file an RFE with Postorius to have the "link new address" screen "suggest" the existing display names, but not automatically add any.
Nice! We can keep the model's intention and have some nice suggestions in the ui.
The intention in the model is that the user's display name can be used as a fallback for any linked address that has no display name.
Also have a "inherit user name" value in the list. Despite the way I lean above, this could be default. But definitely EIBTI (@aditya "explicit is better than implicit") -- we should not take None for "carte blanche".
I certainly agree as far as mining for display names in linked, but not subscribed, addresses. However, I do want to keep the intended semantic that if the *subscribed* address has no display name, we fall back to the linked user record. The idea is that a user can set their display name once, in their user record, and won't need to set it every time they link (and validate) a new address to their profile.
Aditya, do you think you have enough information to finish up mr !104?
Cheers, -Barry