On Mon, 13 May 2002, Mark Keasling wrote:
> However, assuming that one client is keeping a reference to the
> deleted message in its dedicated mailbox for trash rather than the real
> message to reduce disk space usage, wouldn't keeping its trash synchronized
> be difficult for users with a large number of mailboxes who also use the
> DELETE/EXPUNGE model client deleting, undeleting and expunging the messages
> out from under the dedicated mailbox client?

It depends upon what kind of updating that you choose to implement.  I
would suggest that the quest for perfect updating is the enemy of
accomplishing good enough updating.

My suggestion is that you have "lazy updating", in which the Trash view is
updated as mailboxes are seen in normal work *or* when an explicit
operation is done on the Trash (and then only on the mailboxes identified
as having trashed messages).

The problem with trying to do something "perfect" is that it's
unrealistic; it expects something better than what you get with a
DELETE/EXPUNGE client interoperating with a Trash client.  The more modest
goal of "equivalent functionality" will get you a lot further.

> For example: what would happen
> when the DELETE/EXPUNGE client deletes the dedicated mailbox.

This isn't fatal; it can be rebuilt.  It's of less consequence than
deleting the Trash mailbox in the current situation.

One way to avoid the problem is to have well-chosen names so that this
doesn't happen by accident.

> Searching
> the trash would also be difficult to implement.

Most software does not allow searching the trash, or for that matter any
operation other than restoration and emptying.  But it isn't that bad.
You have to open for each mailbox which has a mesasge in the trash, but in
the "normal user" case this is not going to be too bad.

> Different clients would
> have different messages in what they are calling the "Trash" since they'd
> be using different dedicated mailboxes to store their information.

This is a problem.  But there is a chance that if you create a good enough
convention, others may choose to interoperate with you.

After all, there is no guarantee that they'll use the same mailbox name
for a Trash mailbox either, or have the same semantics for that file.
Consider how people disagree about the semantics of postponed
messages/draft mailboxes.

-- Mark --

http://staff.washington.edu/mrc
Science does not emerge from voting, party politics, or public debate.

Reply via email to