[Chandler-dev] Decisions about web work

[Katie Capps Parlante]

We met last week to discuss all of the non-desktop coding projects. We 
made some decisions about priorities and process.

Our two primary goals remain the same:
(1) Fix problems for people who get blocked when they first try and use 
Chandler
(2) Add functionality that makes Chandler more viral

* We need to add a bit of work on the current web UI to our short term 
list. We hand-waved this in our discussions during the on-site week, but 
this has stood out as we've done more detailed planning. We need this 
work to achieve both of our big picture goals. Mimi has identified a set 
of known bugs that get in the away of account creation and initial use. 
Also, we don't want to require desktop usage for the new web widgets to 
be interesting to new users -- the current web UI will play a role in 
basic workflows for these users.

* We think a 1.0 version for the desktop is not far away. We don't want 
to release a desktop 1.0 until we've hit a usability milestone on the 
web UI -- we have a depenency there. (We have a similar dependency on 
improvements to the website -- we'll start planning that once we have 
the engineering work queues in place).

* We have a project wide milestone of getting a 1.0 desktop release out 
and having a few web widgets polished and working well -- probably late 
spring/early summer.

* After scoping out work on the Thunderbird plugin, we decided that 
Brian Kirsch's time would be better spent helping on fixing problems 
with the current web UI and contributing to the new widgets. Fixing the 
current web UI is a higher priority than doing the Thunderbird plugin, 
and we think we can get some quicker wins from the web widgets. It also 
seems right to reduce the number of projects we're working on initially 
-- the new plan feels more focused.

* We decided that we'd use a single work queue to manage all of the 
non-desktop related projects (which we are calling the web work queue). 
We want a single prioritized list for assigning tasks between Travis, 
Jeffrey, Brian, Randy and Jared. We believe this will help us allocate 
work efficiently. We're using the wiki so that we can prioritize item by 
item. Sheila and Mimi will do a draft from the product perspective -- 
engineering will follow up with a second iteration.

* Most of the work on the current web UI has a dependency on the upgrade 
to Dojo 1.0, which Travis is working on -- Jeffrey and Brian will work 
on web widgets projects in the meantime. Travis will propose a few tasks 
that can be done in parallel.

* Given that infrastructure work is required in the service, and given 
that Jared will be out for a bit, we'll prioritize some of the email 
in/out features later in the queue.

Cheers,
Katie