[Top] [All Lists]

Re: [sieve] I-D Action: draft-ietf-sieve-convert-02.txt

2011-08-08 20:15:59
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.


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).




sieve mailing list