Eclipse Community Forums: EPF » Plan Iteration at the beginning

Help

Home

Home » Archived » EPF » Plan Iteration at the beginning

Show: Today's Messages :: Show Polls :: Message Navigator

Plan Iteration at the beginning [message #34073]

Fri, 29 June 2007 08:40

Roman Smirak

Messages: 136
Registered: July 2009

Senior Member

Hi,

as far as I understood, OpenUP plans forthcoming iteration at the
beginning of the iteration (as well as Scrum?); RUP plans the next iteration
in the end of current iteration (as well as XP?).

Did I understand correctly that the beginning of an iteration is now
preferred time to plan the iteration? Why?

I understand Scrum case where you have fixed time frame 4 weeks for every
iteration however in case the iteration planning also sets an iteration
deadline it sounds quite strange to start an unclear iteration. Moreover it
is good practice to have also planning meetings time-boxed (we get this in
case of planning in the end of iteration inherently from the concept
iterations are time-boxed), on the other hand if you start an iteration with
the unclear deadline you can easily give in the "enough time syndrome":-)

Can you please clarify this?

Roman

Report message to a moderator

Re: Plan Iteration at the beginning [message #34277 is a reply to message #34073]

Wed, 04 July 2007 00:27

Ricardo Balduino

Messages: 191
Registered: July 2009

Senior Member

Roman,

IMHO the distinction is not much on whether you plan the iteration today in
the afternoon or tomorrow morning (assuming the current iteration finishes
today and a new one starts tomorrow).
The distinction in OpenUP is that project manager does not spend time
creating an iteration plan for next iteration during the current iteration.
Any new requirements and change requests that are raised during the current
iteration are captured in the work items list (the Scrum product backlog),
and are prioritized (by stakeholders and team) with everything else in the
bucket, then agreed to be allocated to next iteration. That doesn't affect
the time-box, or imply that an iteration starts without any objectives and
work being assigned.

It's a matter of how practical it is to get stakeholders and team together
for this meeting that should take a few hours, where you assess the
iteration results by showing working software to stakeholders, then based on
feedback and prioritized work items, plan the next iteration. These meetings
(iteration assessment and planning) should not take more than a couple hours
or so each. They can happen on the same day (let's say the day the current
iteration finishes). Worst case, you can have the assessment meeting one day
and the planning meeting on the next day.

We use to do this way in the OpenUP team (well, we follow OpenUP to develop
OpenUP :-). The last business day of each month is usually when the
iteration assessment and planning for next iteration happens, all at once,
on a conference call with OpenUP developers. It's been working quite well
for our team.

I hope that it clarifies.
Feel free to ask additional questions.

Ricardo Balduino
IBM | EPF Committer

"Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
news:f62gio$c04$1@build.eclipse.org...
> Hi,
>
> as far as I understood, OpenUP plans forthcoming iteration at the
> beginning of the iteration (as well as Scrum?); RUP plans the next
> iteration in the end of current iteration (as well as XP?).
>
> Did I understand correctly that the beginning of an iteration is now
> preferred time to plan the iteration? Why?
>
> I understand Scrum case where you have fixed time frame 4 weeks for every
> iteration however in case the iteration planning also sets an iteration
> deadline it sounds quite strange to start an unclear iteration. Moreover
> it is good practice to have also planning meetings time-boxed (we get this
> in case of planning in the end of iteration inherently from the concept
> iterations are time-boxed), on the other hand if you start an iteration
> with the unclear deadline you can easily give in the "enough time
> syndrome":-)
>
> Can you please clarify this?
>
> Roman
>

Report message to a moderator

Re: Plan Iteration at the beginning [message #36142 is a reply to message #34277]

Fri, 27 July 2007 13:02

Charles Edwards

Messages: 28
Registered: July 2009

Junior Member

Hi,

I agree with Roman (mostly).

I personally prefer to make sure that we spend the time both evaluating how
we could improve things (retrospective) as well as what we are going to do
next before the end of the current iteration. We found that if you don't
give a deadline for getting the next iterations plan well defined, then it
tends to drift on into the iteration and there is no pressure to finalise
it. We have never managed to finalise all the planning in one day. Maybe the
team needs more practice at this... or maybe we are trying to go into too
much detail.

Having said that I also agree with Ricardo on the point of having a demo of
the software that was completed at the end of the last iteration, so that
you plan the current iteration based on Stakeholder as well as team
feedback.

regards Charles

"Ricardo Balduino" <balduino@us.ibm.com> wrote in message
news:f6ephk$rru$1@build.eclipse.org...
> Roman,
>
> IMHO the distinction is not much on whether you plan the iteration today
> in the afternoon or tomorrow morning (assuming the current iteration
> finishes today and a new one starts tomorrow).
> The distinction in OpenUP is that project manager does not spend time
> creating an iteration plan for next iteration during the current
> iteration. Any new requirements and change requests that are raised during
> the current iteration are captured in the work items list (the Scrum
> product backlog), and are prioritized (by stakeholders and team) with
> everything else in the bucket, then agreed to be allocated to next
> iteration. That doesn't affect the time-box, or imply that an iteration
> starts without any objectives and work being assigned.
>
> It's a matter of how practical it is to get stakeholders and team together
> for this meeting that should take a few hours, where you assess the
> iteration results by showing working software to stakeholders, then based
> on feedback and prioritized work items, plan the next iteration. These
> meetings (iteration assessment and planning) should not take more than a
> couple hours or so each. They can happen on the same day (let's say the
> day the current iteration finishes). Worst case, you can have the
> assessment meeting one day and the planning meeting on the next day.
>
> We use to do this way in the OpenUP team (well, we follow OpenUP to
> develop OpenUP :-). The last business day of each month is usually when
> the iteration assessment and planning for next iteration happens, all at
> once, on a conference call with OpenUP developers. It's been working quite
> well for our team.
>
> I hope that it clarifies.
> Feel free to ask additional questions.
>
> Ricardo Balduino
> IBM | EPF Committer
>
>
> "Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
> news:f62gio$c04$1@build.eclipse.org...
>> Hi,
>>
>> as far as I understood, OpenUP plans forthcoming iteration at the
>> beginning of the iteration (as well as Scrum?); RUP plans the next
>> iteration in the end of current iteration (as well as XP?).
>>
>> Did I understand correctly that the beginning of an iteration is now
>> preferred time to plan the iteration? Why?
>>
>> I understand Scrum case where you have fixed time frame 4 weeks for every
>> iteration however in case the iteration planning also sets an iteration
>> deadline it sounds quite strange to start an unclear iteration. Moreover
>> it is good practice to have also planning meetings time-boxed (we get
>> this in case of planning in the end of iteration inherently from the
>> concept iterations are time-boxed), on the other hand if you start an
>> iteration with the unclear deadline you can easily give in the "enough
>> time syndrome":-)
>>
>> Can you please clarify this?
>>
>> Roman
>>
>
>

Report message to a moderator

Re: Plan Iteration at the beginning [message #36176 is a reply to message #36142]

Tue, 31 July 2007 08:32

Roman Smirak

Messages: 136
Registered: July 2009

Senior Member

Hi,

> Having said that I also agree with Ricardo on the point of having a demo
> of the software that was completed at the end of the last iteration, so
> that you plan the current iteration based on Stakeholder as well as team
> feedback.

That's fundamental definition of an iteration and we do it same way as you
mentioned - it is probably hidden or missing in my text.

Thanks,

Roman

"Charles Edwards" <charles.edwards@processwave.com> wrote in message
news:f8cqds$h03$1@build.eclipse.org...
> Hi,
>
> I agree with Roman (mostly).
>
> I personally prefer to make sure that we spend the time both evaluating
> how we could improve things (retrospective) as well as what we are going
> to do next before the end of the current iteration. We found that if you
> don't give a deadline for getting the next iterations plan well defined,
> then it tends to drift on into the iteration and there is no pressure to
> finalise it. We have never managed to finalise all the planning in one
> day. Maybe the team needs more practice at this... or maybe we are trying
> to go into too much detail.
>
> Having said that I also agree with Ricardo on the point of having a demo
> of the software that was completed at the end of the last iteration, so
> that you plan the current iteration based on Stakeholder as well as team
> feedback.
>
> regards Charles
>
> "Ricardo Balduino" <balduino@us.ibm.com> wrote in message
> news:f6ephk$rru$1@build.eclipse.org...
>> Roman,
>>
>> IMHO the distinction is not much on whether you plan the iteration today
>> in the afternoon or tomorrow morning (assuming the current iteration
>> finishes today and a new one starts tomorrow).
>> The distinction in OpenUP is that project manager does not spend time
>> creating an iteration plan for next iteration during the current
>> iteration. Any new requirements and change requests that are raised
>> during the current iteration are captured in the work items list (the
>> Scrum product backlog), and are prioritized (by stakeholders and team)
>> with everything else in the bucket, then agreed to be allocated to next
>> iteration. That doesn't affect the time-box, or imply that an iteration
>> starts without any objectives and work being assigned.
>>
>> It's a matter of how practical it is to get stakeholders and team
>> together for this meeting that should take a few hours, where you assess
>> the iteration results by showing working software to stakeholders, then
>> based on feedback and prioritized work items, plan the next iteration.
>> These meetings (iteration assessment and planning) should not take more
>> than a couple hours or so each. They can happen on the same day (let's
>> say the day the current iteration finishes). Worst case, you can have the
>> assessment meeting one day and the planning meeting on the next day.
>>
>> We use to do this way in the OpenUP team (well, we follow OpenUP to
>> develop OpenUP :-). The last business day of each month is usually when
>> the iteration assessment and planning for next iteration happens, all at
>> once, on a conference call with OpenUP developers. It's been working
>> quite well for our team.
>>
>> I hope that it clarifies.
>> Feel free to ask additional questions.
>>
>> Ricardo Balduino
>> IBM | EPF Committer
>>
>>
>> "Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
>> news:f62gio$c04$1@build.eclipse.org...
>>> Hi,
>>>
>>> as far as I understood, OpenUP plans forthcoming iteration at the
>>> beginning of the iteration (as well as Scrum?); RUP plans the next
>>> iteration in the end of current iteration (as well as XP?).
>>>
>>> Did I understand correctly that the beginning of an iteration is now
>>> preferred time to plan the iteration? Why?
>>>
>>> I understand Scrum case where you have fixed time frame 4 weeks for
>>> every iteration however in case the iteration planning also sets an
>>> iteration deadline it sounds quite strange to start an unclear
>>> iteration. Moreover it is good practice to have also planning meetings
>>> time-boxed (we get this in case of planning in the end of iteration
>>> inherently from the concept iterations are time-boxed), on the other
>>> hand if you start an iteration with the unclear deadline you can easily
>>> give in the "enough time syndrome":-)
>>>
>>> Can you please clarify this?
>>>
>>> Roman
>>>
>>
>>
>
>

Report message to a moderator