Home » Archived » EPF » Length of an iteration
| Length of an iteration [message #45711] |
Mon, 12 November 2007 05:51  |
Eclipse User |
|
|
|
Hi Per, others,
in your book you latest book
http://www.amazon.com/Agility-Discipline-Made-Easy-Addison-W esley/dp/0321321308
(Classical RUP says the same I guess) you say the Elaboration usually has
longer iterations vs the Construction having shorter iterations (you have
probably focused on a level of parallelism enabled by stable architecture
later in the Construction, haven't you?) - however this doesn't correspond
to my experience: I see this level of parallelism as one of many, actually;
to have shorter iterations in the Elaboration actually enables faster
feedback we are on the wrong track (quite likely in the phase) - I'm strong
advocate for this even in challenging environment like distribute team and
challenging architecture - it is always matter of imagination/intuition
(challenge it!) to find out how to make the iteration as efficient as
possible, isn't it? however if you simply accept "let's have longer
iterations" then you will lose the opportunity;-) On the other hand in the
Construction we focus on complete increment (beta version of a bunch of
functionality) and this also consumes certain amount of time. Actually what
I have observed is sort of similar length of iterations (3 +- one week)
distributed among phases (specific length depends on several factors).
Can you please correct me if I'm wrong/share your view?
Thanks a lot,
Roman
|
|
|
| Re: Length of an iteration [message #46395 is a reply to message #45711] |
Mon, 10 December 2007 16:50  |
| Eclipse User |
|
|
|
Hi Roman,
good points. I am in general recommending to have the same iteration length
throughout, if no strong reasons for changing. This gives the project team a
firm heartbeat.
In table 2.3 on page 50, we describe what drives shorter iterations, which
includes higher risk, architectural uncertainty, new & poorly understood
technology, and so on. This is typically the case for Elaboration, so yes, I
agree with your reasoning, if not the same length, you should have shorter
iterations in Inception / Elaboration.
Where do we contradict table 2.3 and say that you should have shorter
iterations in Construction?
Note that this is un-intuitive. The human nature leads us to do the
opposite. "This is risky, I need longer iterations to get it right: which is
obviously flawed thinking...
On a separate but related note. Many teams feel less stressed in Inception /
Elaboration, since you are further from final release. You actually want to
feel more sense of urgency. It is often easier and cheaper to 'shave off' a
week of schedule time early in the project than late in the project, where
you often have more people and have little room to accelerate things by
procrastinating less... Short iterations redues procrastination...
Cheers
/Per
"Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
news:fh9b80$so0$1@build.eclipse.org...
> Hi Per, others,
>
> in your book you latest book
> http://www.amazon.com/Agility-Discipline-Made-Easy-Addison-W esley/dp/0321321308
> (Classical RUP says the same I guess) you say the Elaboration usually has
> longer iterations vs the Construction having shorter iterations (you have
> probably focused on a level of parallelism enabled by stable architecture
> later in the Construction, haven't you?) - however this doesn't correspond
> to my experience: I see this level of parallelism as one of many,
> actually; to have shorter iterations in the Elaboration actually enables
> faster feedback we are on the wrong track (quite likely in the phase) -
> I'm strong advocate for this even in challenging environment like
> distribute team and challenging architecture - it is always matter of
> imagination/intuition (challenge it!) to find out how to make the
> iteration as efficient as possible, isn't it? however if you simply accept
> "let's have longer iterations" then you will lose the opportunity;-) On
> the other hand in the Construction we focus on complete increment (beta
> version of a bunch of functionality) and this also consumes certain amount
> of time. Actually what I have observed is sort of similar length of
> iterations (3 +- one week) distributed among phases (specific length
> depends on several factors).
>
> Can you please correct me if I'm wrong/share your view?
>
> Thanks a lot,
>
> Roman
>
>
|
|
|
| Re: Length of an iteration [message #46485 is a reply to message #46395] |
Wed, 12 December 2007 06:55 |
| Eclipse User |
|
|
|
Exactly! Thanks Per to confirm my thoughts and conclusions I have made on
the past experience; it happens the conclusions might be wrong due to
limited view and experience.
Thanks a lot,
Roman
"Per Kroll" <pkroll@us.ibm.com> wrote in message
news:fjkcag$nbk$1@build.eclipse.org...
> Hi Roman,
>
> good points. I am in general recommending to have the same iteration
> length throughout, if no strong reasons for changing. This gives the
> project team a firm heartbeat.
>
> In table 2.3 on page 50, we describe what drives shorter iterations, which
> includes higher risk, architectural uncertainty, new & poorly understood
> technology, and so on. This is typically the case for Elaboration, so yes,
> I agree with your reasoning, if not the same length, you should have
> shorter iterations in Inception / Elaboration.
>
> Where do we contradict table 2.3 and say that you should have shorter
> iterations in Construction?
>
> Note that this is un-intuitive. The human nature leads us to do the
> opposite. "This is risky, I need longer iterations to get it right: which
> is obviously flawed thinking...
>
> On a separate but related note. Many teams feel less stressed in Inception
> / Elaboration, since you are further from final release. You actually want
> to feel more sense of urgency. It is often easier and cheaper to 'shave
> off' a week of schedule time early in the project than late in the
> project, where you often have more people and have little room to
> accelerate things by procrastinating less... Short iterations redues
> procrastination...
>
> Cheers
>
> /Per
>
>
> "Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
> news:fh9b80$so0$1@build.eclipse.org...
>> Hi Per, others,
>>
>> in your book you latest book
>> http://www.amazon.com/Agility-Discipline-Made-Easy-Addison-W esley/dp/0321321308
>> (Classical RUP says the same I guess) you say the Elaboration usually has
>> longer iterations vs the Construction having shorter iterations (you have
>> probably focused on a level of parallelism enabled by stable architecture
>> later in the Construction, haven't you?) - however this doesn't
>> correspond to my experience: I see this level of parallelism as one of
>> many, actually; to have shorter iterations in the Elaboration actually
>> enables faster feedback we are on the wrong track (quite likely in the
>> phase) - I'm strong advocate for this even in challenging environment
>> like distribute team and challenging architecture - it is always matter
>> of imagination/intuition (challenge it!) to find out how to make the
>> iteration as efficient as possible, isn't it? however if you simply
>> accept "let's have longer iterations" then you will lose the
>> opportunity;-) On the other hand in the Construction we focus on complete
>> increment (beta version of a bunch of functionality) and this also
>> consumes certain amount of time. Actually what I have observed is sort of
>> similar length of iterations (3 +- one week) distributed among phases
>> (specific length depends on several factors).
>>
>> Can you please correct me if I'm wrong/share your view?
>>
>> Thanks a lot,
>>
>> Roman
>>
>>
>
>
|
|
|
| Re: Length of an iteration [message #585869 is a reply to message #45711] |
Mon, 10 December 2007 16:50 |
| Eclipse User |
|
|
|
Hi Roman,
good points. I am in general recommending to have the same iteration length
throughout, if no strong reasons for changing. This gives the project team a
firm heartbeat.
In table 2.3 on page 50, we describe what drives shorter iterations, which
includes higher risk, architectural uncertainty, new & poorly understood
technology, and so on. This is typically the case for Elaboration, so yes, I
agree with your reasoning, if not the same length, you should have shorter
iterations in Inception / Elaboration.
Where do we contradict table 2.3 and say that you should have shorter
iterations in Construction?
Note that this is un-intuitive. The human nature leads us to do the
opposite. "This is risky, I need longer iterations to get it right: which is
obviously flawed thinking...
On a separate but related note. Many teams feel less stressed in Inception /
Elaboration, since you are further from final release. You actually want to
feel more sense of urgency. It is often easier and cheaper to 'shave off' a
week of schedule time early in the project than late in the project, where
you often have more people and have little room to accelerate things by
procrastinating less... Short iterations redues procrastination...
Cheers
/Per
"Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
news:fh9b80$so0$1@build.eclipse.org...
> Hi Per, others,
>
> in your book you latest book
> http://www.amazon.com/Agility-Discipline-Made-Easy-Addison-W esley/dp/0321321308
> (Classical RUP says the same I guess) you say the Elaboration usually has
> longer iterations vs the Construction having shorter iterations (you have
> probably focused on a level of parallelism enabled by stable architecture
> later in the Construction, haven't you?) - however this doesn't correspond
> to my experience: I see this level of parallelism as one of many,
> actually; to have shorter iterations in the Elaboration actually enables
> faster feedback we are on the wrong track (quite likely in the phase) -
> I'm strong advocate for this even in challenging environment like
> distribute team and challenging architecture - it is always matter of
> imagination/intuition (challenge it!) to find out how to make the
> iteration as efficient as possible, isn't it? however if you simply accept
> "let's have longer iterations" then you will lose the opportunity;-) On
> the other hand in the Construction we focus on complete increment (beta
> version of a bunch of functionality) and this also consumes certain amount
> of time. Actually what I have observed is sort of similar length of
> iterations (3 +- one week) distributed among phases (specific length
> depends on several factors).
>
> Can you please correct me if I'm wrong/share your view?
>
> Thanks a lot,
>
> Roman
>
>
|
|
|
| Re: Length of an iteration [message #585906 is a reply to message #46395] |
Wed, 12 December 2007 06:55 |
| Eclipse User |
|
|
|
Exactly! Thanks Per to confirm my thoughts and conclusions I have made on
the past experience; it happens the conclusions might be wrong due to
limited view and experience.
Thanks a lot,
Roman
"Per Kroll" <pkroll@us.ibm.com> wrote in message
news:fjkcag$nbk$1@build.eclipse.org...
> Hi Roman,
>
> good points. I am in general recommending to have the same iteration
> length throughout, if no strong reasons for changing. This gives the
> project team a firm heartbeat.
>
> In table 2.3 on page 50, we describe what drives shorter iterations, which
> includes higher risk, architectural uncertainty, new & poorly understood
> technology, and so on. This is typically the case for Elaboration, so yes,
> I agree with your reasoning, if not the same length, you should have
> shorter iterations in Inception / Elaboration.
>
> Where do we contradict table 2.3 and say that you should have shorter
> iterations in Construction?
>
> Note that this is un-intuitive. The human nature leads us to do the
> opposite. "This is risky, I need longer iterations to get it right: which
> is obviously flawed thinking...
>
> On a separate but related note. Many teams feel less stressed in Inception
> / Elaboration, since you are further from final release. You actually want
> to feel more sense of urgency. It is often easier and cheaper to 'shave
> off' a week of schedule time early in the project than late in the
> project, where you often have more people and have little room to
> accelerate things by procrastinating less... Short iterations redues
> procrastination...
>
> Cheers
>
> /Per
>
>
> "Roman Smirak" <roman.smirak@tietoenator.com> wrote in message
> news:fh9b80$so0$1@build.eclipse.org...
>> Hi Per, others,
>>
>> in your book you latest book
>> http://www.amazon.com/Agility-Discipline-Made-Easy-Addison-W esley/dp/0321321308
>> (Classical RUP says the same I guess) you say the Elaboration usually has
>> longer iterations vs the Construction having shorter iterations (you have
>> probably focused on a level of parallelism enabled by stable architecture
>> later in the Construction, haven't you?) - however this doesn't
>> correspond to my experience: I see this level of parallelism as one of
>> many, actually; to have shorter iterations in the Elaboration actually
>> enables faster feedback we are on the wrong track (quite likely in the
>> phase) - I'm strong advocate for this even in challenging environment
>> like distribute team and challenging architecture - it is always matter
>> of imagination/intuition (challenge it!) to find out how to make the
>> iteration as efficient as possible, isn't it? however if you simply
>> accept "let's have longer iterations" then you will lose the
>> opportunity;-) On the other hand in the Construction we focus on complete
>> increment (beta version of a bunch of functionality) and this also
>> consumes certain amount of time. Actually what I have observed is sort of
>> similar length of iterations (3 +- one week) distributed among phases
>> (specific length depends on several factors).
>>
>> Can you please correct me if I'm wrong/share your view?
>>
>> Thanks a lot,
>>
>> Roman
>>
>>
>
>
|
|
|
Goto Forum:
Current Time: Thu May 08 08:03:38 EDT 2025
Powered by FUDForum. Page generated in 0.07792 seconds |
Home