[Top] [All Lists]

Re: List of issues with Sieve notifications

2005-10-20 14:48:06

On Thu, Oct 20, 2005 at 07:30:47PM +0100, Alexey Melnikov wrote:
If I were to review such a draft, I would not accept it, unless the
scheme name or overall syntax changed.

Why? Extensions are the "IETF way".

Yes, but extensions need to be recognized as such.  Sieve uses "require",
which I like a lot.  URIs start with the scheme, and having two schemes
with the same name, one being a superset of the other, sounds very odd
to me.

You are trying to work around process issues by using different syntax. 
I think such approach is wrong.
We shouldn't talk about this unless it becomes a problem.


1). The WG needs to evaluate if mailto: is indeed appropriate.

Up to now, it is.  I have a bad feeling that may change, and gave some
examples of options I don't need, but which would mean a dead end to
URI extension, if I did.

2). credentials, message transmission modes and submission server host 
seem like a configuration issue (unless you want each user to 
authenticate as the receiver of the message), I don't think they should 
be mentioned in a Sieve script.

Suppose a SMTP to SMS gateway requires authentication, because it bills
your messages, then you needed ways to specify the credentials and
the gateway host.  But that's just imagination.  All gateways I know
work by having the recipient, not the sender, paying for the message.

Should this turn from imagination to reality, it could look like:

  notify :method "mailto:number(_at_)gateway"
         :options ["login=me","secret=geheim","transmission=encrypted"];

Using :options sounds better than :attributes to me.

I am currently building a list of German SMTP->SMS gateways to know better
what exactly they require and offer.