Platform Q2 Goals

[Damon Sicore]

All,

We're finalizing the Q2 Platform goals. Please take a look and
provide feedback.

What I'd ask people to consider:

* The goals are structured so that the top level bullets are the
actual goals with a strategy and purpose. The supporting bullets are
specific items that we can complete that help accomplish the goal. If
something doesn't make sense, please provide your thoughts.
* Identify things that you are working on and make sure you feel both
the tasks and goal is reasonable for the quarter.
* It's really important to look at other team's goals and do a quick
check to see if there are dependencies on what you are working on.
* Also, make sure other teams are not missing something important. If
they are, please bring it up.

NOTE: Please don't edit the wiki. Please send your changes directly
to me.

Here they are:

https://wiki.mozilla.org/Platform/2009-Q2-Goals

Thanks for any help and/or comments. Very much appreciated.

Sincerely,

Damon

[Clint Talbert]

On 4/3/09 11:48 AM, Damon Sicore wrote:
> All,
>
> We're finalizing the Q2 Platform goals. Please take a look and provide
> feedback.
>
> What I'd ask people to consider:
>
> * The goals are structured so that the top level bullets are the actual
> goals with a strategy and purpose. The supporting bullets are specific
> items that we can complete that help accomplish the goal. If something
> doesn't make sense, please provide your thoughts.

I really like the way you structured these. I will probably copy that
on my own list. My comments are all in the realm of team
interdependence, between test development and platform.

One thing I was surprised to see is the amount of new feature work to be
"implemented". In the current draft of the test dev goals, I'm
advocating that we focus more on implementing tests to burn down the
massive test-needed backlog and plug places where we are lacking decent
test coverage rather than working on tests for new 1.9.2 features.

On the whole, this is probably the right balance if you consider that
most of these implementation tasks will be prioritized lower than 3.5
blockers and will trail along in parallel to the 3.5 ship for most of
the next few months. So, we won't see the fruit of these labors until
late in the quarter and by then it'd be too late for us to do much
functional test dev on them.

One thing I am advocating that test dev do (in Q2) is to figure out
which features we will target for in depth functional testing and
publicize that. That way, we can be sure to be in the loop on those
features and can help them along with interim tests or by taking peeks
at patches or what-have-you. But our backlog work will take slightly
higher priority for Q2.

For Q3, I imagine we'll reverse these priorities and prioritize
functional testing for new 1.9.2 features over backlog (but both will
continue to be important). We're just trying to get the biggest bang
for the project with our tiny number of folks.

Thoughts on that balance?

Clint

[Vladimir Vukicevic]

On 4/3/09 3:32 PM, Clint Talbert wrote:
> On the whole, this is probably the right balance if you consider that
> most of these implementation tasks will be prioritized lower than 3.5
> blockers and will trail along in parallel to the 3.5 ship for most of
> the next few months. So, we won't see the fruit of these labors until
> late in the quarter and by then it'd be too late for us to do much
> functional test dev on them.

I don't think you should assume this -- there is still work to be done
on 3.5, but that's not across the entire platform team. Pieces of new
features will be showing up throughout the quarter.

- Vlad

[Boris Zbarsky]

Clint Talbert wrote:
> On the whole, this is probably the right balance if you consider that
> most of these implementation tasks will be prioritized lower than 3.5
> blockers and will trail along in parallel to the 3.5 ship for most of
> the next few months.

I'd hope not! In fact, there's been pretty significant post-1.9.1 work
happening for months now (that's why we branched all the way back in
November), and it should be ramping up as 1.9.1 locks down. I fully
expect most people to NOT be working on blockers between the first RC
and final ship, for example: we just shouldn't have that many blockers
by then.

Also note that "features" from a testing perspective includes core arch
changes that are not user-visible features. Some of these are covered
by our existing regression tests, but we have code areas with poor
coverage too....

-Boris