ietf-mta-filters
[Top] [All Lists]

Re: Informal Working Group Last Call on draft-freed-sieve-date-index-04.txt

2007-04-18 08:18:13

Alexey Melnikov wrote:

> We would like to draw your attention to the following draft:

> <http://www.ietf.org/internet-drafts/draft-freed-sieve-date-index-04.txt>

> A three week informal last call of this document starts today on
> March 6th 2007 and ends on March 27th 2007 at 6 pm EST.

The informal last call has ended.

I believe that Ned has addressed all outstanding issues in
draft-freed-sieve-date-index-05.txt, so I will recommend its
publication as an RFC. Thanks to all who reviewed the document.

The one issue several people have raised is that the document needs more
examples. I'm going to push one more revision out for sure to address this. The
other (very minor) change I want to make is to allow the date test to
optionally work on non RFC2822 format date-time information.

There is one other issue that occurred to me recently which I haven't addressed
and may be worthin considering. The currentdate test either operaters in a
specified time zone or in the "local" zone, which makes sense given those are
the only possibly relevant zones. However, the date test either operates in the
"local" zone or in a specified zone as  well. This provides no way to extract
the original zone from the header.

The first question is do we care? I'm having trouble coming up with a case
where we do, but of course there's always the generic "someone might want to
know the original zone".

The second question is if we do care, how to implement it? An "original zone"
part could be added. Or it could be done with an :originalzone argument.
Or it could be done as a distinguished argument value to :sone, e.g.

   :zone "original"

I prefer the third approach personally, but could be persuaded that either
of the other two is better.

At this point I would like to see a show of hands from people who are
interested in implementing this extension.

FWIW, I have implemented both date and currentdate and we're in the process of
transitioning a lot of use of the old setdate test from an early version of the
variables draft to currentdate. (Currentdate is much more straightforward and
this makes our generated Sieves a bit tidier.)

:index is much more difficult for our implementation. I have it mostly working
but there are some intricacies involving the editheader extension I still don't
have nailed down.

Please reply to this message or directly to Ned and myself.

Yes please. And any examples you think make sense would also be welcome.

                                Ned