Thunderbird:Summit 2014/minutes

From MozillaWiki
Jump to: navigation, search

Contacts Session 15october2014 morning

Action Items:

  1. mconley: review bug 841603, is it the right thing to do? does it match mailnews?
  2. review existing Mozilla (and non Mozilla) CardDAv implementations with an eye to see if they are usable in Thunderbird
  3. once 1 and 2 are done, how do we get CardDav into existing Address Book including existing UI? see if this a "google summer of code"-size project?

Contact Session 16october2014

Compose (jcranmer)

Joshua has some rewrites planned for nsMsgSend. We discussed needs, and agreed to investigate creating the ability in TB 38 to allow existing compose and send code to be replaced by an extension, which would be used as a test ground for the new compose and send implementation.

New Account Types / skinkglue (rkent)

  • There was general agreement that implementing the capability provided by SkinkGlue should be a target for TB
  • We need to test existing js XPCOM wrappers to see of the complexity of the C++ override in SkingGlue is needed


  • We need a dns SRV implementation available, either in core or in the messaging dns web server (see gozer)
  • We should be adding calendar and chat (XMPP) information in autolookup

OAuth / Gmail

  • There are a number of unresolved issues with OAuth usability, including private key exposure and auto lookup
  • GMail now requires multiple steps for new users to access IMAP without using OAuth, which in an unacceptable user experience.
  • We need to implement, at least for GMail, OAuth option for TB 38. [roland] Could we please have it sooner :-) ? e.g. TB 35, TB 36?



Testing in Real Environments

  • Docker is a viable option for testing in linux systems. This would be viable for ldap
  • We should consider setting up standard test environments with various anti-virus vendors on Amazon VMs, and run them manually as part of release testing.


  • Fastmail is promoting JMAP which is an open-source replacement for ActiveSync
  • They should have open client and server libraries available in 2-3 months
  • Thunderbird wants to make it easier to implement new specs, and when that is available we would welcome an implementation of JMAP
  • There is a step beyond that, where Mozilla takes a solid position to encourage a particular protocol. Neither JMAP nor Thunderbird is ready to take that step, but we recognize there in the future we would be willing to discuss a more proactive approach.