If you’ve been using your mailbox (and deleting mail) for a while, have over 50K mails in it now, and see (what you think is) a UID of 51950 on the most recent email, the chances that it’s “U” are extremely low, meaning there’s a gap in understanding or in implementation.
Every time I see that, I’m floored by it. The fact that message IDs in IMAP change when you delete messages has got to be one of the worst design choices in any in-use protocol. I’m flabbergasted by it.
I'm pretty sure Proton didn't invent IMAP, and from the protocol log it seems like IMAP insists on the incrementing ids. Probably thanks to it having been designed in the late eighties and early nineties.
Can you elaborate on your thinking?
If you’ve been using your mailbox (and deleting mail) for a while, have over 50K mails in it now, and see (what you think is) a UID of 51950 on the most recent email, the chances that it’s “U” are extremely low, meaning there’s a gap in understanding or in implementation.
Every time I see that, I’m floored by it. The fact that message IDs in IMAP change when you delete messages has got to be one of the worst design choices in any in-use protocol. I’m flabbergasted by it.
The sooner everyone moves to jmap the better.
7 replies →
I know right? I mean naive question, but why wouldn't they simply use actual UUIDs/GUIDs?
I'm pretty sure Proton didn't invent IMAP, and from the protocol log it seems like IMAP insists on the incrementing ids. Probably thanks to it having been designed in the late eighties and early nineties.