Position Summary: Dapper Drake Release Management

On 00:35, Tue 18 Oct 05, Mark Shuttleworth wrote:
> Now, the big question that this thread hasn't addressed has been the
> amount of work we try to put into new features. There were a LOT fo
> goals for Breezy, many delivered, many partially completed, some
> deferred entirely. It is going to be important that we prioritise and
> set tightly defined and managed goals for Dapper, so that we balance the
> enrgy that goes into new features with the energy we put into polish.
> It's not just stabilisation that will make Dapper rock - it's lots of
> polish in easy-to-find but time consuming-to-execute places that will
> make people think "wow, this raises the bar in the Just Works (tm) game".

We should probably try to follow the absolute "must have's".
That is , research and get a good understanding about what
are the the most important features we lack for raising the
bar in the "Just Works" game. Prioritization should also
consider an estimation  of the milage to reaching any
specific goal, in view of time limits, available manpower
and obtrusivness nature a goal/feature might present.

Ideally (IMHO) , the resulting goals / feature list would
consist of <=50% of the must have's that are not too trivial
to accomplish but represent milestones for "Just Work(s)", and 
the remaining portion being those "nice to have(s)" that can
be more easily implemented.

Limiting ourselves to a fixed number of either of the two
will force us to probably drop some of each, but will surely
enable us to do it in a better way and produce more stable
results.

My 2NIS on that ;)

Sivan