Tim Showalter wrote:
What's wrong with saying that some actions affect delivery status,
and other actions don't?
If you're having these things in a UI, it doesn't matter what we say, we
can (we have to) assume the UI will just get it right.
But for the people designing the UI, and the people implementing scripts
using a text editor, the fewer special cases, the better.
It is much easier to explain this:
"A message that generates no actions gets filed into your INBOX."
than this:
"A mesasge that generates no delivery actions gets filed into your
INBOX. Delivery actions are ones which change the disposition of the
message, and they are reject, forward, and fileinto."
I personally think that adding this distinction NOW will vastly simplify
future additions to the language. I can think of several possible
extensions which definitely should not affect the delivery decisions
that are made elsewhere within the script. By saying that these new
actions must affect the delivery status without having a way of saying
that they don't, or having a model for such actions in the future, we
are going to make life more difficult in the future.
Consider vacation. Does it affect delivery or not? I'd say it shouldn't.
Wouldn't it be far simpler for the vacation draft to just say "The
vacation verb is not a delivery action." than to add verbiage saying
that it doesn't.
How about an extension that says "I want this message to be
cross-indexed in a fast search engine." I'd like such an extension to
also say "this action does not define a delivery action."
How about an alerting extension that says "notify my pager about this
message." The fact that a notification is sent shouldn't affect how the
message is otherwise filed.
So, yes I do think it is simpler, in the long run, to use the latter
definition you specify above.
Tony Hansen
tony(_at_)att(_dot_)com