On 8/4/2011 5:09 PM, Alexey Melnikov wrote:
Hi Stephan,
Stephan Bosch wrote:
On 7/25/2011 8:18 PM, internet-drafts(_at_)ietf(_dot_)org wrote:
I gave this a quick first review. I have the following questions:
1) Do we need a mechanism to discover which conversions are
available (analogous to the IMAP CONVERSIONS command and the
AVAILABLECONVERSIONS data item)?
We can have a new test or magic variables (in a special namespace) for
that.
Do you really think that the extra complexity worth it? Maybe you have
a use case in mind.
Well, I think it would be good to be able avoid unnecessary runtime
errors (as proposed below). I guess the same considerations applied for
the enotify extension and its valid_notify_method and
notify_method_capability tests.
I don't have a particular use-case in mind.. yet. I won't be
implementing this soon, so it will be hard to foresee on my part.
2) Error handling: what happens when the convert command is issued
with invalid parameters, e.g. unknown conversion, unknown/invalid
parameter etc. ? And what happens when a conversion just fails
temporarily or due to invalid input data from the message?
I think this should be a runtime error.
Yes.
3) The interaction with subsequent disposition actions and other
convert actions is described explicitly. But what happens with
subsequent tests that evaluate the message body (e.g. header :mime) ?
E.g. for the editheader extension, subsequent header tests all see
the changes made. If tests following the convert command should see
the changes as well, applying the instructed conversions at the end
of Script execution (as suggested at the end of Section 2) will be
difficult.
Hmm. I was mostly thinking about leaf body parts conversions, but
other conversions should be possible.
I agree with applying changed "on the spot" (just like editheader).
Ok.
Regards,
Stephan.
_______________________________________________
sieve mailing list
sieve(_at_)ietf(_dot_)org
https://www.ietf.org/mailman/listinfo/sieve