Hi, > On Tue, 30 Dec 2003, Christof Drescher wrote: > > That's the problem of the "ding-dong", not the problem of the server > > developer. > > It's a bad argument to say "don't do it, because someone could misuse it". > > Painful lessons of the past 30 years indicate that when harm is made > possible, harm will happen; and that the collateral damage from the harm > (especially to innocent third parties) greatly exceeds the effort that > could have been taken to avoid it in the first place.
This is an argument for good planning, not for completely discarding an idea. > > This is your opinion. I think delete-expunge is better as well, but I do not > > go as far as to say it is better for EVERYONE. > > A traditional source of extreme conflict between client developers and > server developers in IMAP has been whether the client or the server has > control over user interface aspects; and furthermore whether the client > can rely upon all servers to operate in a predictable way. I am deliberately for the latter, as you are. > You are proposing a major change to the server-end behavior of mailbox > semantics based upon an administrative setting which, from the client's > perspective, is unilateral and utterly unexpected. No. It is neither unilateral nor unexpected, if this information is given by the admins. If they do, everything is ok, if they don't, it is their fault, not the programmer's. Again, this is an argument for a server extension. If the client sets an option, he'll get what he wants. If he doesn't, there's no change at all. The override by an admin is not desireable, but may be necessary in certain environments. You might call a server broken which has it on by default, but you cannot call it broken simply if it allows this setting. > > If he insists, I will do it - he's the customer. > > What happens when your customer finds out that the server does not work > with many clients, and the client vendors all say it is because your > server is broken? That cannot happen. a) the customer wanted it. Keep a record of his wish, you're fine. b) there is nothing as a "broken behavior". I do not propose a change which would "break" any client, since all rfc-conforming clients would not give any error message. If it is done on a "client settable flag" (which enables the server extension by command) all is ok. If it was set by admins, someone with authorative power did it, so it is again not the fault of the programmer. Christof
