Re: [imapext] Proposal for a new IMAP Working Group to revise CONDSTORE & QRESYNC

[Dave Cridland <[e-mail]>]

On Mar 6, 2013 10:57 AM, "Cyrus Daboo" <[e-mail]> wrote:
>
> Hi Alexey,
>
>
> --On March 5, 2013 at 2:57:25 PM +0000 Alexey Melnikov <
[e-mail]> wrote:
>
>> So please comment on whether you think this work is worth doing and on
>> the proposed charter text.
>>
>> I've also posted a new version of RFC 5162bis
>> (http://tools.ietf.org/html/draft-melnikov-5162bis-01), which addressed
>> most (but not all) of existing errata on the document. If a new WG is to
>> be formed, I propose that the document is used as one of the initial
>> documents.
>
>
> What about also doing an update to RFC4549?
>
> I think the big thing client implementors want is a single clear
specification of extensions that describe how to accurately do mailbox
synchronization. To that end I would even argue that we should be looking
at pieces to take out to achieve that goal. For example, how many clients
actually make use of the condition store piece of CONDSTORE, as opposed to
just the SELECT/SEARCH sync operations? Same goes for QRESYNC - what pieces
of it have actually been used by clients? Are there bits that could be
removed?
>
> So I'd like to see more than a simple rev of the documents to fix errata
- or at least lets not rule that out right now.

There's certainly unused pieces, like the complex per flag modseq, that can
be dropped. I'm not totally sure that general conditional store is
something that's so onerous that our needs dropping, though.

I'm in ready agreement that the QRESYNC document could use some serious
improvements to its prose, too. It was hard to write effectively, not least
because the technique was relatively new, and the reviewers were almost
entirely server side as I recall. I'm hopeful that with many more client
side reviewers we can write a clearer description of how it all works.

Dave.