[teampractices] [Wmfall] Please help define the term "core work"

Kevin Smith ksmith at wikimedia.org
Wed Jan 6 21:05:04 UTC 2016


I think you're right that the goal will be to spend as little as possible
on "core" work. But if someone asks why we're spending so little on core, I
don't see a problem explaining that we are spending more time on
"strategic" work instead. I had no input into the selection of the word
"core", but at least so far, I don't object to it.

Personally, I have no reason to expect there to be any kind of "50%"
quotas, at the org or team level. Each team has a different context, and
some might be doing 5% core work, and others might be doing 95% core work,
and both might be doing exactly what they should to support the mission and
vision. I can imagine a team seeing that it is spending a lot of time
putting out fires, and end up choosing a quarterly goal to reduce their
"core" effort on a certain system by half, or something like that. But it
would be a means to an end (of being more efficient), not chosen just to
make a number look pretty.

You mentioned a large gray area between core and strategic, and a desire
not to be forced to put work into one bucket or the other. I agree that
there will be ambiguities, but in those cases I don't have a problem
picking one of the two buckets. That is different from the large gray area
of work that is neither core nor strategic. If I were an executive or
donor, I would look very warily at "other" work, and wonder why we're
putting effort into it, if it is neither core nor strategic. For that work,
I think it would be dysfunctional to try to jam it into one of the two
buckets that it doesn't really belong in.




Kevin Smith
Agile Coach, Wikimedia Foundation


On Wed, Jan 6, 2016 at 11:21 AM, C. Scott Ananian <cananian at wikimedia.org>
wrote:

> Kevin: your response is helpful, but it does confirm my impression that
> there currently is a large grey area between 'core' and 'strategic'.
> Trying to define a 'third term' for this grey area might be better than
> leaving it ambiguous, since otherwise there will be a great temptation to
> force things into one bucket or the other depending on "which is better for
> me at the moment".  For example, if we say "ideally our work will be 50%
> core and 50% strategic", I'm pretty sure that I can just massage things to
> match that artificial target by slushing stuff over into one bucket or the
> other.  Vague definitions corrupt the metrics and turn assessments into
> box-checking exercises.  It seems the solution isn't tweaking the
> definition (which is what you're asking us to do) because the basic
> concepts presuppose a sharp distinction between "feature development" and
> "bug fixing" which doesn't align with actual development processes.
>
> WRT to your first question: the common English word "core" has a
> connotation of importance and large size.  Yes, perhaps the "core" work of
> the Foundation is actually keeping the servers running.  But it certainly
> isn't the "core" of the work I do every day -- we try very hard to create
> services that *don't* require large ongoing maintenance burdens.  If the
> "core" of my responsibilities was "core" work ("just keeping the lights on
> and the servers from falling over!"), I would conclude that we were
> flailing and that our software was not very well constructed.
>
> Even among (say) the infrastructure group I suspect they would classify
> most of their work as "strategic" -- ongoing improvements to ensure that we
> can spend less and less time on the "core" of our mission.  My point is
> just that there is cognitive dissonance between the common English usage
> and the use of this word in this planning context.  We will inevitably
> asked why we try to spend only <small percentage> of our time on "core"
> tasks, if they are indeed "core" to our mission.
>
> Perhaps these concepts make more sense for the legal department than they
> do for engineering.
>   --scott
>
>
> On Wed, Jan 6, 2016 at 9:54 AM, Kevin Smith <ksmith at wikimedia.org> wrote:
>
>> Hi,
>>
>> I'm having a little trouble parsing your statement that "our 'core work'
>> is maintenance now". When you say "our", do you mean your team, or the WMF
>> as a whole? And when you say "now", do you mean "prior to this moment" or
>> "after these new definitions go into effect"?
>>
>> What follows are opinions off the top of my head. I could easily change
>> my mind.
>>
>> For me, it was helpful to consider Core/Strategic/Other as being
>> completely independent of the concept of "maintenance". With that in mind,
>> version 1 of a strategic feature would be strategic. Version 2 would
>> probably also be strategic, since if a version 2 weren't necessary for
>> strategic reasons, we probably shouldn't be working on it. Once the feature
>> is "done" to the point of meeting the strategic needs, then keeping it
>> running would probably be "Core" work.
>>
>> Maintaining an existing feature would usually fall into the "core"
>> bucket, although the way some people define "maintenance" it could actually
>> be strategic. Or, if the feature really isn't serving the movement, that
>> maintenance might be neither core nor strategic. There are a lot of gray
>> areas, and in some cases there are no objectively "right" answers. We're
>> hoping to build up the definitions and examples to the point where we can
>> generally agree on which bucket a particular piece of work would go in. But
>> there will always be some ambiguity at the edges.
>>
>> All of that is really a better discussion for the wiki page, so I
>> encourage you to take it up there. But I hope this helps.
>>
>>
>>
>> Kevin Smith
>> Agile Coach, Wikimedia Foundation
>>
>>
>> On Tue, Jan 5, 2016 at 10:32 PM, C. Scott Ananian <cananian at wikimedia.org
>> > wrote:
>>
>>> It definitely seems odd to me that our "core work" is maintenance now.
>>>
>>> It also seems at odds with software development processes where the
>>> conventional wisdom is that "version 3" of the Thing is the version that is
>>> first generally useful and usable.
>>>
>>> So building "version 1" is "strategic" and versions 3 and following are
>>> "core" (the Thing is presumably feature-complete then) but the long slog of
>>> bug fixing, refactoring, dealing with the inevitable encounters of design
>>> vs real world and rebuilding the parts which never worked... Is that
>>> maintenance ("core") or (we-didn't-realize-we-actually-needed-that-)
>>> feature development ("strategic")?
>>>
>>> Much of the development on Parsoid and VE over the past two years has
>>> been of this sort.  I don't think the proposed terminology is particularly
>>> helpful.  I suppose I'd label all that feature-fixing as "strategic" just
>>> because it makes me feel better about my work.
>>>   --scott
>>> On Jan 5, 2016 12:36 PM, "Oliver Keyes" <okeyes at wikimedia.org> wrote:
>>>
>>> So we created categories with values we don't know the meaning of?
>>>
>>> That seems sort of backwards ;p.
>>>
>>> On 5 January 2016 at 15:23, Kevin Smith <ksmith at wikimedia.org> wrote:
>>>
>>>> Summary:
>>>> Please visit this wiki page[2] to help define the term "core work".
>>>>
>>>> Long form:
>>>> The WMF annual planning process[1] is being revised this year in
>>>> several ways. One is to attempt to categorize work as "Core", "Strategic",
>>>> or "Other", which should help us communicate our work externally, and to
>>>> improve our internal budget and strategic planning processes.
>>>>
>>>> As Lila mentioned in a recent update, we need to define these terms.
>>>> The Team Practices Group was asked to help the stakeholders reach agreement
>>>> on these definitions, so we have created a wiki page[2] about "core". It
>>>> contains a draft (strawdog) definition, along with rationale and examples.
>>>> Through the discussion there, the definition will be refined so it can be
>>>> used throughout the foundation. Note that this will not be limited to
>>>> product development teams--it will be org-wide.
>>>>
>>>> Early planning for FY 2015-2016 will start next week, and that will
>>>> benefit from having a working definition. So if you have any strong
>>>> feelings about this, please voice them as soon as possible. However, I
>>>> envision this as being a living definition which can continue to evolve
>>>> over time. So even if you can't participate this week, please visit the
>>>> page when you can, and share your ideas then.
>>>>
>>>> This work relates to various discussions and pilots last year related
>>>> to "maintenance fraction", and "interrupt" or "unplanned" work. However,
>>>> this conversation has a tighter focus, along with (hopefully) a clear
>>>> rationale for the categorization. It also relates to the
>>>> "New/Reactive/Maintenance" categorization that have been used recently in
>>>> the Quarterly Reviews, but is distinct from that as well.
>>>>
>>>> Aside from annual planning, we believe that categorizing this work will
>>>> help teams themselves. For example, it might help them realize when they
>>>> are getting pulled into doing work that is neither core nor strategic, or
>>>> it might help identify opportunities to invest in improvements that would
>>>> greatly reduce the ongoing effort required to keep existing features
>>>> working smoothly.
>>>>
>>>> Please continue this discussion on wiki, or email me (or just the Team
>>>> Practices list) if you have process questions or concerns.
>>>>
>>>> [1]
>>>> https://wikimediafoundation.org/wiki/2015-2016_Annual_Plan/Questions_and_Answers
>>>> [2]
>>>> https://www.mediawiki.org/wiki/Team_Practices_Group/Tracking_core_and_strategic_work
>>>>
>>>> Kevin Smith
>>>> Agile Coach, Wikimedia Foundation
>>>>
>>>>
>>>> _______________________________________________
>>>> Wmfall mailing list
>>>> Wmfall at lists.wikimedia.org
>>>> https://lists.wikimedia.org/mailman/listinfo/wmfall
>>>>
>>>>
>>>
>>>
>>> --
>>> Oliver Keyes
>>> Count Logula
>>> Wikimedia Foundation
>>>
>>> _______________________________________________
>>> Wmfall mailing list
>>> Wmfall at lists.wikimedia.org
>>> https://lists.wikimedia.org/mailman/listinfo/wmfall
>>>
>>>
>>
>
>
> --
> (http://cscott.net)
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.wikimedia.org/pipermail/teampractices/attachments/20160106/79f009ca/attachment.html>


More information about the teampractices mailing list