On 02/08/2016 06:23 AM, Bron Gondwana via Cyrus-devel wrote:
Present: Nicola, ellie, Bron, Simon

Nicola
  - that email I nagged about docs.  Bron - will have time tomorrow, let's just 
do it.
  - Sieve docs are coming along very well.
  - Checksumming extension idea from FM wiki is a silly idea, let's not do 
that.  Use external tools to add headers and filter on those

Bron
  - released RFC 5490 mailbox support to FastMail today, pushed back along with 
Cassandane tests to cf/master.
  - planning to keep working on JMAP this week

Ellie
- working on tooling for getting data out of backups at the moment
- have been hoping to get code onto master but i'm hesitant to do so while the 
tests aren't passing, cause i don't want to complicate things

Simon
  - no Cyrus news

Failing tests - looks like Ken has pushed back something which is breaking the 
/.well-known/caldav discovery, and hence everything after that point is failing 
miserably. Haven't looked into it - FastMail has just branched off an earlier 
version.

I just fixed this with my latest commit.


Ken's new work (that probably broke this) is proper standards-track calendar 
sharing.  Part of bringing FastMail onto that will be making sure that our 
existing shares aren't broken.  This should allow Apple clients (and others in 
future) to work with sharing on Cyrus.  Need to find docs on it for Nicola!

The sharing work is getting close to completion. I need to implement the reply side of the invitation process. The invite side is pretty much complete. The POST is parsed and the appropriate ACL changes made on the sharer's collection and notifications sent to the sharees. The code to read shared mailboxes out of the subscriptions db and "bind" them into the home-set is there. Doing a DELETE on a shared collection just removes it from the subscriptions db.

I haven't decided if we want/should allow a sharee to actually delete a sharer's collection if it has been shared as read-write. I'm think only the owner should be allowed to DELETE a collection that they are sharing.

--
Kenneth Murchison
Principal Systems Software Engineeer
Carnegie Mellon University

Reply via email to