|| ||David Ascher <david.ascher-AT-gmail.com>|
|| ||Thunderbird 3 Planning|
|| ||Mon, 28 Jan 2008 14:53:49 -0800|
It's time to define the Thunderbird 3 plan. I've spent a fair bit of
time learning about the state of affairs and talking to many people, and
I feel I've accumulated enough information to start this process.
Note: I'm cross-posting this to the planning, calendar and thunderbird
newsgroups, but expect discussion on the thunderbird newsgroup and have
set followup-to accordingly. There will be a summary post in the
planning newsgroup if the final plan differs significantly from the one
The long-term roadmap of Thunderbird is still in flux, but there are
four high-level points which drive my thinking about Thunderbird 3:
1. Thunderbird's impact is proportional to its user count. Thus driving
adoption is my primary concern. Our current user base is very
significant (many millions of mostly quite satisfied users), but the
number of possible users of Thunderbird is orders of magnitude greater
than our current reach.
2. The reasons why people don't choose to use Thunderbird are varied,
but two primary reasons appear to be: the lack of a built-in calendar
integration (compared to Outlook for example), or a search experience
that doesn't match that offered by competitors (gmail and Mail.app for
3. In addition, Thunderbird's codebase has a fair bit of technical debt
due to insufficient resourcing over the years, which has led to a
codebase which has too many scary bits, not enough test coverage, and
isn't yet able to leverage the ongoing platform improvements. In
addition, while communications clients are by nature great targets for
extension authors, the current codebase isn't extension-friendly enough,
making it too hard to build installation-specific features or experiment
with new feature ideas.
4. A fair number of Thunderbird changes have already landed on trunk,
including some important bug fixes, by a variety of contributors.
There's appropriate pressure to ship an update to Thunderbird 2 to take
advantage of those and of the platform improvements.
With all that as background, I propose:
* Goal: to have at public milestone build of Thunderbird 3 in 2008.
Thunderbird 3's overall aim is to significantly grow its user base
worldwide, as well as build a strong foundation for later Thunderbird
* Release-defining features:
- an integrated calendaring feature, based on Lightning
- a better search experience, especially for message content searches
- a better overall user experience
* Less user-visible but important goals include:
- Significant headway on getting rid of Mork and RDF
- A concerted effort to improving the extensions ecosystem for
Thunderbird, including refactorings, FUEL, developer documentation, and
- Better test coverage and performance metrics in place to support
There will be of course lots of other bug fixes and enhancements
(patches welcome ;-))
* Schedule: Figuring out the schedule at this stage is hard, as it will
depend on who shows up with energy and talent. I would like to set some
placeholder milestones for discussion, however:
- alpha builds in Q1
- beta builds without calendaring starting in Q2
- beta builds with calendaring starting in Q3
- widely useful builds by Q4 (although whether they're branded
"release" will depend on quality, as always).
We're revise the schedule as we gain knowledge.
* Thunderbird 3 work will happen on trunk, with branching strategy to be
figured out closer to the endgame (and reviewed next when 1.9 is cut),
* The Mailnews/Thunderbird folks and the Calendar folks will have to
figure out how to best allocate dev and testing effort on the
calendaring features, how we support Sunbird, etc.
Given the scope of the work, the aggressive schedule, and the amount of
new feature develoment, integration and stabilization work involved,
help of all kinds is more than welcome! Thanks in advance for any input
you may have, either on process or on deliverables.
The central wiki page for Thunderbird 3 is
http://wiki.mozilla.org/Thunderbird:Thunderbird3. IRC discussion will
take place in #maildev. The newsgroup/mailing list of record for Tb3 is
I look forward to the discussion!
-- David Ascher
to post comments)