Oh I see what you're saying but in what circumstance will a user have more than one email. You only get emailed when you make an update and you can't see the updates until you click the approval link... so I don't think that will be overly helpful.
That said, I could probably have a session or something so that if you approve an email then I'll let you make additional changes within that session without additional approval emails. You're "authenticated" as it is.
Doing session management is orthogonal to this. Still a fun idea, though. Even better, actually. But it introduces a new security concern. How quickly should the session time-out (or not?). But it will be good for UX.
But to explain the original idea:
Users can work in what sometimes might seem strange ways: A user can do multiple edits and choose to be perfectly fine not seeing the updates. The flow state is then simply doing the updates - then go to mail and do the approvals. In that flow it would be nice to only approve one mail which then covers all updates.
00:00 Edit A - Timer start
00:09 Edit B - Timer reset
00:18 Edit C - Timer reset
00:28 Timeout - send one mail for approval of A,B,C
The bulk update mail will then need a top link to simply approve all. And individual links below for each edit as mistakes might have been made during edits (if userfriendly :-)). A lot of work to to save a couple of mails - I know! But the things we choose to obsess over :-) It just might fit your definitio of minimalism.
People who want a fast turnaround of approvals one by one will be annoyed by this. But mail is "slow" and can be delayed anyway. People who want fewer clicks and mails will love it. Good design is then finding the optimum or choose who to champion.
Sub-optimisation. For sure!
So not important at all. And a very limited use-case. I just ran with the idea. As you succintly said elsewhere: This is like art. I will second that with a quote from The Dude: "That's just like, your opinion, man". Very true - and you should do you! I just hope I made clear what my impulsive idea was.
A little bit akin to how TN3270 works. Powerusers did crazy things working blindly ahead while awaiting the response from the mainframe.
if the user changes something, hits submit, thinks about what they did, hits back, changes something else, there should be both changes.
you could expire the form to prevent this, or some other thing to prevent it so you're right that it's not a problem that will get hit as implemented but just something to keep in mind if things change.
