On Fri, 2003-06-20 at 17:24, Mark Crispin wrote: > \Unmarked is the most useful status. It indicates to the client that it > can definitely skip consideration of a mailbox.
I'm not convinced I agree. A common behaviour I desire from a client is to find mailboxen which have new mail. Yet the \Unmarked flag doesn't necessarily indicate that status. The \Unmarked flag says that no new mail has been delivered since the mailbox was last SELECTed. Consider the case where my main client is issuing a LIST periodically then asking for STATUS of non-\Unmarked folders. I connect with another client, SELECT a folder and FETCH an old message from it, for some reason. The folder in question had new mail in it, which my main client had not yet observed... but now the folder doesn't have \Unmarked status since it's been SELECTed since the new mail was delivered. TBH I don't think the meaning (or indeed existence) of \Marked is particularly relevant -- it's the presence or absence of \Unmarked which in an ideal world would give the client some clue as to how to behave. I'd much rather see \Unmarked clearly defined as 'This folder contains no messages without \Seen flag', and \Marked similarly defined as 'This folder contains messages without \Seen flag.' That would seem more in line with what's really desired by real-world clients. In a different mail you also write: > It is non-constructive to make absurd interpretations of the protocol in > this mailing list. If the wording in the document is broken, then please > propose a fix to the wording. Don't damage the protocol. To a certain extent, absurd interpretations of the protocol do need to be taken into account. It's an unfortunate fact that you do have to account for the lowest common denominator -- however absurd an interpretation may be, you'll still find some nut out there will write their server to it, and hence a client author may need to work around the problem; and vice versa. -- dwmw2
