dovecot - Mar 2015 - IMAP ANNOTATE Extension RFC5257: priority on roadmap

Hi Timo,

congrats to the merger with OX.

Currently the implementation of RFC 5257, ANNOTATE-EXPERIMENT-1, has
only low priority on http://wiki2.dovecot.org/Roadmap
I want to explain a scenario that would benefit from annotation support
to - maybe - increase the priority in your roadmap:

I'm currently working on a project to publish bank customer related
documents inside a banking webinterface (they call it 'postbox').

There are different requirements that would be satisfied very well by an
IMAP server as data storage/provider for a long term email/document
archive inside the bank.

Annotations would be necessary to tag emails with customer numbers and
document IDs to allow a delayed grouping of several emails by such
identifiers (message IDs may not be available for references so
annotations have to be used to compute references indirectly).
Same for sorting and filtering by customer number or account number.

I would highly appreciate if you at Dovecot could increase the priority
of ANNOTATE in your roadmap.

Best regards,
Florian

-- 

Agitos GmbH
Emil-Geis-Stra?e 40
D-81379 M?nchen
Telefon: +49-89-381 564 46-0
Telefax: +49-89-381 564 46-9
E-Mail: support at agitos.de
Web: http://www.agitos.de

Registergericht: Amtsgericht M?nchen, HRB 213590
Gesch?ftsf?hrer: Dipl.-Inf. (Univ.) Florian Sager
Ust-Identifikationsnummer: DE296114074

On 21 Mar 2015, at 10:14, Florian Sager <sager at agitos.de>
wrote:
> 
> Hi Timo,
> 
> congrats to the merger with OX.
> 
> Currently the implementation of RFC 5257, ANNOTATE-EXPERIMENT-1, has
> only low priority on http://wiki2.dovecot.org/Roadmap
> I want to explain a scenario that would benefit from annotation support
> to - maybe - increase the priority in your roadmap:
> 
> I'm currently working on a project to publish bank customer related
> documents inside a banking webinterface (they call it 'postbox').
> 
> There are different requirements that would be satisfied very well by an
> IMAP server as data storage/provider for a long term email/document
> archive inside the bank.
> 
> Annotations would be necessary to tag emails with customer numbers and
> document IDs to allow a delayed grouping of several emails by such
> identifiers (message IDs may not be available for references so
> annotations have to be used to compute references indirectly).
> Same for sorting and filtering by customer number or account number.
> 
> I would highly appreciate if you at Dovecot could increase the priority
> of ANNOTATE in your roadmap.
The Roadmap page hasn't been updated for a while now. ANNOTATE is something
that I'd like to implement at some point, but it's also quite a large
change and likely will have to wait until Dovecot v2.3. You could already
implement similar functionality with METADATA, although that would be kind of
kludgy. Basically use SETMETADATA folder (/private/vendor/vendor.abcdefg/key/123
"value") where 123 would be the UID. You'd of course somehow need
to explicitly delete the metadata entries when messages are expunged.

Am 2015-03-22 um 18:06 schrieb Timo Sirainen:
> On 21 Mar 2015, at 10:14, Florian Sager <sager at agitos.de> wrote:
>>
>> I would highly appreciate if you at Dovecot could increase the
>> priority of ANNOTATE in your roadmap.
>
> ANNOTATE is
> something that I'd like to implement at some point, but it's also
> quite a large change and likely will have to wait until Dovecot v2.3.
>
Ideally, and of course, I speak from my personal point of view, 
annotations should be something, that SOLR would index according to its 
own mapping, which I am in control of.

Use case: incoming mails here are not directly answered in the MUA, but 
some third party software creates outgoing emails, with a formal yet 
unique subject; inbound messages could then be /tagged/ with the same 
identifier and a single search would yield the complete correspondence?

Basically, adding a header to a message would do my job as well, but 
seems to contradict with what IMAP seems to be about: preserving 
messages unadulterated, which is a noble goal indeed.

OTOH, no hurry, client support of ANNOTATE is not there yet either (at 
least in Thunderbird).

-- 
peter