Another KEP proposal: On IMAP-metadata (annotations)

Jeroen van Meeuwen (Kolab Systems) vanmeeuwen at kolabsys.com
Fri Jul 15 11:01:57 CEST 2011


Florian v. Samson wrote:
> Hello all,
> 
> IMO the Kolab format MUST reference the correct specification for the
> IMAP-metadata (annotations) to use.
> 

I concur.

> Hence I propose this mini-KEP for the Kolab-format specification:
> 
> Kolab-clients and -servers MUST support IMAP-metadata (annotations)
> specified in RFC5464 [1].
> 

I concur.

> I see two major "pros":
> - Implementers of client-side IMAP-libraries know what to implement /
> expect from the Kolab-server.  (Specifically it took us quite a while to
> analyse this an come to a concise conclusion in the evolution-kolab
> project: [2]) - When adapting an existing PIM-client for the use with
> Kolab, it is very hard to convince the maintainers of the
> IMAP-implementation of that client to integrate an IMAP-metadata
> (annotations) extension which is based on a long outdated RFC-draft, which
> is quite incompatible to the final RFC5464.
> 
> I also see a "con":
> - The Kolab-community will have to provide patches to the IMAP-server(s)
> used in the Kolab-server (Cyrus and/or Dovecot), again.  Still, as on the
> client-side, these patches supposedly will be received much more welcome
> upstream than those implementing a long outdated RFC-draft, which is quite
> incompatible to the final RFC5464.
> 

Speaking as the release engineer for Cyrus IMAP upstream, Cyrus IMAP 2.4 and 
beyond are fully compatible with RFC5464. I do not see Cyrus IMAP 2.3.x not 
being fully RFC5464 compatible as a disadvantage, however.

With all the work needing to get done in the Cyrus project (both SASL and 
IMAP) moving forward, there's only very limited focus on support for legacy 
releases such as the 2.2 and 2.3 series. I thought I'd just say it and put it 
out there, in light of the following remark;

> I also propose enhancing the *current* Kolab format specification with a
> concise statement, which is absolutely missing, currently:
> 
> Kolab-clients and -servers MUST support IMAP-metadata (annotations)
> according to the "IMAP ANNOTATEMORE Extension"-draft-05 [3].
> 

I'm not sure what the implications are here, and I do not want to speak out of 
turn, but if one or more client-/server- implementations need significant work 
in order to gain compliancy with said draft, then perhaps a more efficient 
approach is to cut-through and do the work towards RFC 5464 (instead).

> It took us (when doing evolution-kolab) quite a while to research which
> version of the draft is implemented in the Cyrus-IMAPd as of Kolab 2.2.4,
> as this is not noted or even hinted anywhere.  We fiercely hope that we got
> that analysis right.
> 

You can always ask, as I'm supposed to know this from the top of my head -or 
at least be able to assist you in the quest for such information ;-) 

Kind regards,

Jeroen van Meeuwen

-- 
Senior Engineer, Kolab Systems AG

e: vanmeeuwen at kolabsys.com
t: +44 144 340 9500
m: +44 74 2516 3817
w: http://www.kolabsys.com

pgp: 9342 BF08




More information about the format mailing list