Loading ...
Sorry, an error occurred while loading the content.

Re: [scrumdevelopment] Iteration Length

Expand Messages
  • Ron Jeffries
    ... In the olden days, we used to say that you should set your iteration size to the amount of money you could afford to totally waste. I have observed a big
    Message 1 of 27 , Feb 8, 2004
    • 0 Attachment
      On Sunday, February 8, 2004, at 12:16:28 AM, Ken Schwaber wrote:

      > What is the optimal length of an iteation, and what are the factors
      > that are relevant to the various lengths of iterations that are
      > deemed most appropriate?

      In the olden days, we used to say that you should set your iteration size
      to the amount of money you could afford to totally waste.

      I have observed a big difference in estimates at one month. A three week
      estimate seems to mean three weeks a lot more strongly than a one month
      estimate means one month. "That'll take three weeks." "Ohhhh about a
      month."

      I'm now recommending one-week cycles even if the team isn't doing any XP at
      all. I think it's really good to be able to turn around an updated system
      that fast -- it focuses attention on automating all the stuff that should
      be. On the other hand, the planning wagon comes around an awful lot. That
      might be a bad thing -- or it might be good.

      Ron Jeffries
      www.XProgramming.com
      This is how I develop software.
      Take the parts that make sense to you.
      Ignore the rest.
    • PaulOldfield1@compuserve.com
      (responding to Ken) ... There are surely a lot of answers to this. About the only general, all-purpose reply I can give is that if iterations are too long,
      Message 2 of 27 , Feb 8, 2004
      • 0 Attachment
        (responding to Ken)

        > What is the optimal length of an iteation, and what are
        > the factors that are relevant to the various lengths of
        > iterations that are deemed most appropriate?

        There are surely a lot of answers to this.

        About the only general, all-purpose reply I can give is
        that if iterations are too long, people spend too long
        'relaxing' at the start of an iteration, and either the work
        rate slows or there's too much to be covered in the
        panic period at the end of the iteration; and if iterations
        are too short, the panic period at the end of an iteration
        takes up too much of the overall time and people
        start to burn out.


        Paul Oldfield
        www.aptprocess.com
      • Ron Jeffries
        ... What are the issues around skipping or shortening the panic period? Ron Jeffries www.XProgramming.com The practices are not XP. They are a path to XP.
        Message 3 of 27 , Feb 8, 2004
        • 0 Attachment
          On Sunday, February 8, 2004, at 7:57:12 AM, PaulOldfield1@... wrote:

          > (responding to Ken)

          >> What is the optimal length of an iteation, and what are
          >> the factors that are relevant to the various lengths of
          >> iterations that are deemed most appropriate?

          > There are surely a lot of answers to this.

          > About the only general, all-purpose reply I can give is
          > that if iterations are too long, people spend too long
          > 'relaxing' at the start of an iteration, and either the work
          > rate slows or there's too much to be covered in the
          > panic period at the end of the iteration; and if iterations
          > are too short, the panic period at the end of an iteration
          > takes up too much of the overall time and people
          > start to burn out.

          What are the issues around skipping or shortening the panic period?

          Ron Jeffries
          www.XProgramming.com
          The practices are not XP. They are a path to XP.
        • PaulOldfield1@compuserve.com
          (responding to Ron) ... Anything that can shorten the relaxation period and the panic period has got to be good. (?) More up-front analysis of the
          Message 4 of 27 , Feb 8, 2004
          • 0 Attachment
            (responding to Ron)

            >> About the only general, all-purpose reply I can give is
            >> that if iterations are too long, people spend too long
            >> 'relaxing' at the start of an iteration, and either the work
            >> rate slows or there's too much to be covered in the
            >> panic period at the end of the iteration; and if iterations
            >> are too short, the panic period at the end of an iteration
            >> takes up too much of the overall time and people
            >> start to burn out.
            >
            > What are the issues around skipping or shortening the
            > panic period?

            Anything that can shorten the relaxation period and the
            panic period has got to be good. (?) More up-front
            analysis of the requirements seems to allow finer
            subdivision of requirements; finer subdivision of
            requirements seems to allow a regular pace throughout
            the iteration. I think the issues will be about how much
            up-front analysis is the right amount, and how much
            this varies between projects. We may end up with a
            trade-off between doing more up-front analysis and
            having a steady pace through the iteration.
            Feel free to differ.


            Paul Oldfield
            www.aptprocess.com
          • Mike Cohn
            I ve come to prefer two-week iterations. With a team that is very experienced and has worked together for a long time I ve had good success with four weeks. In
            Message 5 of 27 , Feb 8, 2004
            • 0 Attachment
              I've come to prefer two-week iterations. With a team that is very
              experienced and has worked together for a long time I've had good success
              with four weeks. In general, though, two seems better. It gets a bit frantic
              with two week iterations to do all the planning and prep work. For example,
              we just finished a sprint on last Thursday. Friday was devoted to the sprint
              review and sprint planning. On Tuesday and Wednesday of this week (or just
              over a week from the next sprint start the following Friday) there will be
              planning meetings where the product owner and stakeholders argue about what
              to prioritize highest. The team isn't necessarily in these (some often are
              for short parts) but they take a lot of my time as ScrumMaster. Often out of
              these meetings we come up with new priorities or partial priorities from old
              ones--these require the team to estimate new work which we do some morning
              after the daily scrum.

              --Mike

              -----Original Message-----
              From: Ken Schwaber [mailto:ken.schwaber@...]
              Sent: Saturday, February 07, 2004 10:16 PM
              To: scrumdevelopment@yahoogroups.com
              Subject: [scrumdevelopment] Iteration Length

              What is the optimal length of an iteation, and what are the factors
              that are relevant to the various lengths of iterations that are
              deemed most appropriate?
              Ken




              To Post a message, send it to: scrumdevelopment@...
              To Unsubscribe, send a blank message to:
              scrumdevelopment-unsubscribe@...
              Yahoo! Groups Links
            • Joshua Kerievsky
              I went from doing 3 week iteraions to lots of 2 week iterations to lots of 1 week iterations. Now we routinely do either 1 or 2 week iterations. We prefer to
              Message 6 of 27 , Feb 8, 2004
              • 0 Attachment
                I went from doing 3 week iteraions to lots of 2 week iterations to lots
                of 1 week iterations. Now we routinely do either 1 or 2 week
                iterations. We prefer to do 1-week iterations. However, we find that 2
                week iterations work better for our C++ clients, since they tend to have
                a lot of legacy code, which slows everything down. --jk
              • Matt Bennett
                ... Ken- I have been thinking more about this question of iteration length lately. I had a good bit of success 2 years ago as a developer on a couple XP
                Message 7 of 27 , Feb 8, 2004
                • 0 Attachment
                  > Message: 5
                  > Date: Sun, 08 Feb 2004 05:16:28 -0000
                  > From: "Ken Schwaber" <ken.schwaber@...>
                  > Subject: Iteration Length
                  >
                  > What is the optimal length of an iteation, and what are the factors
                  > that are relevant to the various lengths of iterations that are
                  > deemed most appropriate?
                  > Ken
                  >

                  Ken-

                  I have been thinking more about this question of iteration length lately. I
                  had a good bit of success 2 years ago as a developer on a couple XP projects
                  with 2 or 3 week iterations. Now I am managing a team using mostly SCRUM
                  elements. The team as a whole is on one-month cycles. This team is made up
                  of 18 individuals including 3 business analysts, 1 architect, 1 ui designer,
                  8 developers, and 5 testers. I have separated this group into 2 teams: the
                  analysis team and the implementation team. The analysis team is always
                  researching functional and technical specs for enhancements with both
                  internal and external customers. These enhancements are coded and tested
                  within the next month or so by the other team. The developers and testers
                  work closely together to (drum-roll please) code and test all functionality
                  that their team committed to for the month by the end of the month.

                  So back to your question on iteration length, one month just seems too long
                  for development. It is easy for the team to lack urgency early in the
                  month, lose focus in the middle of the month, and finally have to pour all
                  that urgency and focus into the last week or so. In my past experience,
                  shorter iterations of 2 weeks seem to keep the pace moving. As a developer
                  you knew that what you committed to needed to be completed soon, before you
                  are working on something else. However, one month iterations are great for
                  project reporting. I am starting to move towards a model of 2 week
                  iterations for coding and testing within a monthly reporting of the planning
                  and progress of the team.

                  So my thoughts are that some of the factors of iteration length are:
                  -Maturity of the items in the backlog: Is enough known about the items to
                  start coding them almost immediately?
                  -Size of the items: I have long argued that in new development that any
                  piece of work can be broken down into smaller bits of work that can fit a
                  small iteration cycle and still deliver complete functionality with each
                  iteration. However, my thoughts on this have changed a bit since working on
                  significant enhancements to an enterprise product. Many items require
                  several developers for several weeks approaching an enhancement holistically
                  to ensure that complete functionality can be delivered without adversely
                  impacting the existing system.
                  -Who/What are the iterations for? Management, the developers, the analysts,
                  external customers? This probably seems like something that should strictly
                  be driven by the development cycles, but sometimes the access to the
                  customers is not frequent enough for shorter iterations. Of course that
                  just points at a larger problem of customer involvement, but this cannot be
                  ignored as a valid factor.
                  -What is the intent of the iteration- to deliver new, tested functionality
                  into a system? It seems that including the iteration planning meeting, the
                  coding, the testing, the deployment, and the review meeting even for small
                  items requires no less than 2 weeks from end-to-end.

                  Since this is getting rather long-winded, I'll wait to see how others reply
                  or add to this before I add more to this.

                  Deceivingly simple question Ken asked isn't it? :)

                  -Matt Bennett
                • Phlip
                  ... Not sure how many plurals under we here. There are those who think that a young team should never vary iteration length from iteration to iteration, and
                  Message 8 of 27 , Feb 8, 2004
                  • 0 Attachment
                    Joshua Kerievsky wrote:

                    > I went from doing 3 week iteraions to lots of 2 week
                    > iterations to lots
                    > of 1 week iterations. Now we routinely do either 1
                    > or 2 week
                    > iterations. We prefer to do 1-week iterations.
                    > However, we find that 2
                    > week iterations work better for our C++ clients,
                    > since they tend to have
                    > a lot of legacy code, which slows everything down.
                    > --jk

                    Not sure how many plurals under "we" here. There are
                    those who think that a young team should never vary
                    iteration length from iteration to iteration, and that
                    a mature team shouldn't have a reason to. Do you mean
                    that one team looks at its stack of stories and sets
                    iteration length accordingly?

                    Call this the "rhythm method"...

                    =====
                    Phlip
                    http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                    __________________________________
                    Do you Yahoo!?
                    Yahoo! Finance: Get your refund fast by filing online.
                    http://taxes.yahoo.com/filing.html
                  • acockburn@aol.com
                    I read this as we Industrial Logic, with various clients .... some clients get 1-week iterations, some get 2-week iterations. The clients getting 2-week
                    Message 9 of 27 , Feb 8, 2004
                    • 0 Attachment
                      I read this as "we" Industrial Logic, with various clients .... some clients get 1-week iterations, some get 2-week iterations. The clients getting 2-week iterations tend to be those using C++.  I don't see an indication in Joshua's note that iteration length varies within a project, although with a mature team, I don't see that they couldn't arbitrarily change the iteration length to suit the circumstances discovered on planning day.
                       
                      In a message dated 2/8/2004 9:40:10 PM Mountain Standard Time, phlipcpp@... writes:
                      Joshua Kerievsky wrote:

                      > I went from doing 3 week iteraions to lots of 2 week
                      > iterations to lots
                      > of 1 week iterations.  Now we routinely do either 1
                      > or 2 week
                      > iterations.  We prefer to do 1-week iterations.
                      > However, we find that 2
                      > week iterations work better for our C++ clients,
                      > since they tend to have
                      > a lot of legacy code, which slows everything down.
                      > --jk

                      Not sure how many plurals under "we" here. There are
                      those who think that a young team should never vary
                      iteration length from iteration to iteration, and that
                      a mature team shouldn't have a reason to. Do you mean
                      that one team looks at its stack of stories and sets
                      iteration length accordingly?

                      Call this the "rhythm method"...

                      =====
                      Phlip
                      ==============================================
                      Alistair Cockburn
                      President, Humans and Technology

                      http://alistair.cockburn.us alistair.cockburn@...
                      1814 E. Fort Douglas Circle, Salt Lake City, UT 84103
                      Phone: 801.582-3162            Fax: 775.416.6457

                      Author of
                      "Surviving Object-Oriented Projects" (1998)
                      "Writing Effective Use Cases" (Jolt Productivity Award 2001)
                      "Agile Software Development" (Jolt Productivity Award 2002)

                      "La perfection est atteinte non quand il ne reste rien a ajouter,
                      mais quand il ne reste rien a enlever." (Saint-Exupery)
                      ==============================================

                    • Phlip
                      ... I wasn t accusing. I pointed out his exact text was ambiguous, by adding the rhythm method context to the thread. Among its quantifyable benefits are a
                      Message 10 of 27 , Feb 9, 2004
                      • 0 Attachment
                        acockburn@... wrote:

                        > I read this as "we" Industrial Logic, with various
                        > clients .... some clients
                        > get 1-week iterations, some get 2-week iterations.
                        > The clients getting 2-week
                        > iterations tend to be those using C++. I don't see
                        > an indication in Joshua's
                        > note that iteration length varies within a project,
                        > although with a mature
                        > team, I don't see that they couldn't arbitrarily
                        > change the iteration length to
                        > suit the circumstances discovered on planning day.

                        I wasn't accusing.

                        I pointed out his exact text was ambiguous, by adding
                        the "rhythm method" context to the thread. Among its
                        quantifyable benefits are a stable Yesterday's Weather
                        metric - or so I'm told. Accuse whoever told me that.

                        > "La perfection est atteinte non quand il ne reste
                        > rien a ajouter,
                        > mais quand il ne reste rien a enlever."
                        > (Saint-Exupery)

                        "le meilleur est l'ennemi du bien" ([apocryphally] Voltaire)

                        =====
                        Phlip
                        http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                        __________________________________
                        Do you Yahoo!?
                        Yahoo! Finance: Get your refund fast by filing online.
                        http://taxes.yahoo.com/filing.html
                      • Karl Scotland
                        ... A great ScrumMaster once told me, It depends on common sense :) Personally, I m doing one week iterations wih my team at the moment. I wouldn t
                        Message 11 of 27 , Feb 9, 2004
                        • 0 Attachment
                          > -----Original Message-----
                          > From: Ken Schwaber [mailto:ken.schwaber@...]
                          >
                          > What is the optimal length of an iteation, and what are the factors
                          > that are relevant to the various lengths of iterations that are
                          > deemed most appropriate?

                          A great ScrumMaster once told me, "It depends on common sense" :)

                          Personally, I'm doing one week iterations wih my team at the moment. I
                          wouldn't necessarily work the same way with every team though. Others
                          seem to be saying the same thing.

                          Factirs which would affect my decision would include...
                          1) Business chaos - a business which is really ducking and diving would
                          benefit from shorter iterations.
                          2) Project length - the shorter the overall projects, the shorter the
                          iterations.
                          3) Technical ability - shorter iterations benefit less mature teams
                          because they promote more focus.

                          If I was working with a great team, on a longterm project, for a stable
                          business, I'd probably consider monthly iterations.

                          I've submitted a paper on doing weekly iterations with Scrum and XP for
                          the gathering. Hope it gets accepted :)

                          Karl

                          BBCi at http://www.bbc.co.uk/

                          This e-mail (and any attachments) is confidential and may contain
                          personal views which are not the views of the BBC unless specifically
                          stated.
                          If you have received it in error, please delete it from your system.
                          Do not use, copy or disclose the information in any way nor act in
                          reliance on it and notify the sender immediately. Please note that the
                          BBC monitors e-mails sent or received.
                          Further communication will signify your consent to this.
                        • Joshua Kerievsky
                          ... We is Industrial Logic. We don t generally vary iteration length. However, we have changed iteration length mid-project, when we found that the length
                          Message 12 of 27 , Feb 9, 2004
                          • 0 Attachment
                            Phlip wrote:

                            >Not sure how many plurals under "we" here. There are
                            >those who think that a young team should never vary
                            >iteration length from iteration to iteration, and that
                            >a mature team shouldn't have a reason to. Do you mean
                            >that one team looks at its stack of stories and sets
                            >iteration length accordingly?
                            >
                            >
                            We is Industrial Logic. We don't generally vary iteration length.
                            However, we have changed iteration length mid-project, when we found
                            that the length wasn't working well. We also once extended an iteration
                            by a week, as it was the last iteration before going into production and
                            it just felt right to the whole team to take an extra week for our work.

                            I once considered 1-week iterations to be a bad idea. "We'll spend too
                            much time planning and not have enough time programming." When I got
                            over my fear and actually tried it, I found that I loved 1-week
                            iterations. Customers learn to plan more efficiently by doing
                            continuous iteration planning preparation, customers gets to make
                            finer-grained changes to the plan and programmers have less time to do
                            wasteful work.

                            In our workshops, we do 3-hour iterations. That's like working under a
                            high-powered microscope -- every flaw gets revealed, which tends to
                            leads folks to important insights.

                            best regards,
                            jk

                            --
                            I n d u s t r i a l L o g i c , I n c .
                            Joshua Kerievsky
                            Founder, Extreme Programmer & Coach
                            http://industriallogic.com
                            http://industrialxp.org
                            866-540-8336 (toll free)
                            510-540-8336 (phone)
                            Berkeley, California
                          • Patrick Parato
                            We are currently using one week interations with success. We have a couple of reasons for doing just this. The development team and the business are both
                            Message 13 of 27 , Feb 9, 2004
                            • 0 Attachment
                              We are currently using one week interations with success. We have a couple of reasons for doing just this. The development team and the business are both relatively new to Agile and we are trying to expose everybody to different types of situations quickly. We hope people will learn about the whole life cycle of an Agile project faster. By making the iterations more frequent, we get more practice with the overall Agile experience. We may be trading off some efficiency for some beneift in education, but everybody understands that the iteration length can change when it is considered necessary to do so.
                               
                              The only drawback we have seen is that the development team fizzles out close to the end of the iteration. The tasks will be all complete but people don't want to admit it because the one week iterations make them feel constantly under pressure. The developers don't really like the feeling that they can't vary thier effort depending on mood, health, or other external stimuli. Some developers believe that one month long iteration allows them more freedom to vary thier pace, and that the one week iteration is too much pressure. Very similar to micro management.
                               
                              Regards,
                               
                              Patrick Parato
                              CertaPay
                              Toronto, Canada
                            • Phlip
                              ... Weeks are biological, and part of our circadian rhythms. (Of course months are too, which we derive weeks from.) I suspect even rats kept in isolation from
                              Message 14 of 27 , Feb 9, 2004
                              • 0 Attachment
                                Joshua Kerievsky wrote:

                                > I once considered 1-week iterations to be a bad
                                > idea.

                                Weeks are biological, and part of our circadian
                                rhythms. (Of course months are too, which we derive
                                weeks from.)

                                I suspect even rats kept in isolation from cues to the
                                date vary their activity patterns in synch with human
                                weekends.


                                =====
                                Phlip
                                http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                                __________________________________
                                Do you Yahoo!?
                                Yahoo! Finance: Get your refund fast by filing online.
                                http://taxes.yahoo.com/filing.html
                              • Michael Campbell
                                ... Never knew that. Why is it then that only 1 month, and then only 75% of the time even HAS an integral # of them? __________________________________ Do you
                                Message 15 of 27 , Feb 9, 2004
                                • 0 Attachment
                                  Phlip wrote:

                                  > Weeks are biological, and part of our circadian
                                  > rhythms. (Of course months are too, which we derive
                                  > weeks from.)

                                  Never knew that. Why is it then that only 1 month, and then only
                                  75% of the time even HAS an integral # of them?


                                  __________________________________
                                  Do you Yahoo!?
                                  Yahoo! Finance: Get your refund fast by filing online.
                                  http://taxes.yahoo.com/filing.html
                                • Joshua Kerievsky
                                  ... By planning according to a team s velocity (which is calculated at every iteration end), XPers ensure that programmers have enough time to implement
                                  Message 16 of 27 , Feb 9, 2004
                                  • 0 Attachment
                                    Patrick Parato wrote:

                                    > We are currently using one week interations with success. We have a
                                    > couple of reasons for doing just this. The development team and the
                                    > business are both relatively new to Agile and we are trying to expose
                                    > everybody to different types of situations quickly. We hope people
                                    > will learn about the whole life cycle of an Agile project faster. By
                                    > making the iterations more frequent, we get more practice with the
                                    > overall Agile experience. We may be trading off some efficiency for
                                    > some beneift in education, but everybody understands that the
                                    > iteration length can change when it is considered necessary to do so.
                                    >
                                    > The only drawback we have seen is that the development team fizzles
                                    > out close to the end of the iteration. The tasks will be all complete
                                    > but people don't want to admit it because the one week iterations make
                                    > them feel constantly under pressure. The developers don't really like
                                    > the feeling that they can't vary thier effort depending on mood,
                                    > health, or other external stimuli. Some developers believe that one
                                    > month long iteration allows them more freedom to vary thier pace, and
                                    > that the one week iteration is too much pressure. Very similar to
                                    > micro management.

                                    By planning according to a team's velocity (which is calculated at every
                                    iteration end), XPers ensure that programmers have enough time to
                                    implement iteration stories using good development practices without
                                    feeling under pressure. Feeling under pressure results from committing
                                    to doing too much during a given period of time. So 1-week iterations
                                    don't cause pressure. Taking on more work than you can do causes
                                    pressure. Yet many programmers haven't yet had that insight.

                                    best regards
                                    jk
                                  • Phlip
                                    ... Uh, because periods of the rotation of the Earth, its orbit, and the Moon s orbit all don t divide evenly? ===== Phlip
                                    Message 17 of 27 , Feb 9, 2004
                                    • 0 Attachment
                                      Michael Campbell wrote:

                                      > Never knew that. Why is it then that only 1 month,
                                      > and then only
                                      > 75% of the time even HAS an integral # of them?

                                      Uh, because periods of the rotation of the Earth, its
                                      orbit, and the Moon's orbit all don't divide evenly?


                                      =====
                                      Phlip
                                      http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                                      __________________________________
                                      Do you Yahoo!?
                                      Yahoo! Finance: Get your refund fast by filing online.
                                      http://taxes.yahoo.com/filing.html
                                    • Steven Gordon
                                      That would argue that days and months (and years) are natural, but not weeks. ... From: Phlip [mailto:phlipcpp@yahoo.com] Sent: Monday, February 09, 2004 9:52
                                      Message 18 of 27 , Feb 9, 2004
                                      • 0 Attachment
                                        That would argue that days and months (and years) are natural, but not weeks.
                                        -----Original Message-----
                                        From: Phlip [mailto:phlipcpp@...]
                                        Sent: Monday, February 09, 2004 9:52 AM
                                        To: scrumdevelopment@yahoogroups.com
                                        Subject: Re: [scrumdevelopment] Iteration Length

                                        Michael Campbell wrote:

                                        > Never knew that.  Why is it then that only 1 month,
                                        > and then only
                                        > 75% of the time even HAS an integral # of them?

                                        Uh, because periods of the rotation of the Earth, its
                                        orbit, and the Moon's orbit all don't divide evenly?


                                        =====
                                        Phlip
                                            http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                                        __________________________________
                                        Do you Yahoo!?
                                        Yahoo! Finance: Get your refund fast by filing online.
                                        http://taxes.yahoo.com/filing.html


                                        To Post a message, send it to:   scrumdevelopment@...
                                        To Unsubscribe, send a blank message to: scrumdevelopment-unsubscribe@...



                                      • J. B. Rainsberger
                                        ... I have enjoyed two-weeks iterations to start a project, then one-week iterations, as needed, towards the release date. The idea is that we need larger
                                        Message 19 of 27 , Feb 9, 2004
                                        • 0 Attachment
                                          Ken Schwaber wrote:

                                          > What is the optimal length of an iteation, and what are the factors
                                          > that are relevant to the various lengths of iterations that are
                                          > deemed most appropriate?

                                          I have enjoyed two-weeks iterations to start a project, then one-week
                                          iterations, as needed, towards the release date.

                                          The idea is that we need larger chunks of time to hit our stride and to
                                          allow for some experimental error. Without that experimental time, we
                                          tend not to learn as much, which inhibits our ability to prepare for the
                                          next game.

                                          I find that one-week iterations are best for focusing on doing what we
                                          know how to do. At this granularity, we can steer very well. I find that
                                          two-week iterations are best when we need to learn things, because then
                                          we feel less pressure when we might accidentally burn a day or two going
                                          down a rathole.
                                          --
                                          J. B. Rainsberger,
                                          Diaspar Software Services
                                          http://www.diasparsoftware.com :: +1 416 791-8603
                                          Let's write software that people understand
                                        • agiletoken
                                          I totally agree. I have implemented an agile process with four (4) one (1) week iterations enveloped within a 30 day Scrum Sprint. The one week XP interations
                                          Message 20 of 27 , Feb 11, 2004
                                          • 0 Attachment
                                            I totally agree. I have implemented an agile process with four (4)
                                            one (1) week iterations enveloped within a 30 day Scrum Sprint. The
                                            one week XP interations enforce good engineering practices and the 30
                                            day sprint ensures business to technical community alignment. The
                                            team meets Monday morning and establishes goals for the week - taken
                                            from sprint backlog.

                                            The daily standup reflects on the progress for the. The team has an
                                            option to change the goals for the week on Wednesday, as it is a mid-
                                            point and the team generally has a very good idea on whether the
                                            goals will be met on Friday. On Friday, the team confirms the
                                            deliverables. From the engineering perspective, the once the team
                                            commits to the Friday goals, no one leaves until the team has
                                            delivered the goals and an integration build completed and all
                                            testcases executed.

                                            Tareq



                                            --- In scrumdevelopment@yahoogroups.com, Joshua Kerievsky <jlk@i...>
                                            wrote:
                                            > Patrick Parato wrote:
                                            >
                                            > > We are currently using one week interations with success. We have
                                            a
                                            > > couple of reasons for doing just this. The development team and
                                            the
                                            > > business are both relatively new to Agile and we are trying to
                                            expose
                                            > > everybody to different types of situations quickly. We hope
                                            people
                                            > > will learn about the whole life cycle of an Agile project faster.
                                            By
                                            > > making the iterations more frequent, we get more practice with
                                            the
                                            > > overall Agile experience. We may be trading off some efficiency
                                            for
                                            > > some beneift in education, but everybody understands that the
                                            > > iteration length can change when it is considered necessary to do
                                            so.
                                            > >
                                            > > The only drawback we have seen is that the development team
                                            fizzles
                                            > > out close to the end of the iteration. The tasks will be all
                                            complete
                                            > > but people don't want to admit it because the one week iterations
                                            make
                                            > > them feel constantly under pressure. The developers don't really
                                            like
                                            > > the feeling that they can't vary thier effort depending on mood,
                                            > > health, or other external stimuli. Some developers believe that
                                            one
                                            > > month long iteration allows them more freedom to vary thier pace,
                                            and
                                            > > that the one week iteration is too much pressure. Very similar to
                                            > > micro management.
                                            >
                                            > By planning according to a team's velocity (which is calculated at
                                            every
                                            > iteration end), XPers ensure that programmers have enough time to
                                            > implement iteration stories using good development practices
                                            without
                                            > feeling under pressure. Feeling under pressure results from
                                            committing
                                            > to doing too much during a given period of time. So 1-week
                                            iterations
                                            > don't cause pressure. Taking on more work than you can do causes
                                            > pressure. Yet many programmers haven't yet had that insight.
                                            >
                                            > best regards
                                            > jk
                                          • Ken Schwaber
                                            Does the length of the iteration have any impact on the granularity of the input (Stories, Product Backlog). What is the type of demonstration that is desired
                                            Message 21 of 27 , Feb 11, 2004
                                            • 0 Attachment
                                              Does the length of the iteration have any impact on the granularity of the
                                              input (Stories, Product Backlog). What is the type of demonstration that is
                                              desired at the end of the iteration?
                                              Ken

                                              -----Original Message-----
                                              From: Joshua Kerievsky [mailto:jlk@...]
                                              Sent: Sunday, February 08, 2004 9:02 PM
                                              To: scrumdevelopment@yahoogroups.com
                                              Subject: Re: [scrumdevelopment] Iteration Length


                                              I went from doing 3 week iteraions to lots of 2 week iterations to lots
                                              of 1 week iterations. Now we routinely do either 1 or 2 week
                                              iterations. We prefer to do 1-week iterations. However, we find that 2
                                              week iterations work better for our C++ clients, since they tend to have
                                              a lot of legacy code, which slows everything down. --jk






                                              To Post a message, send it to: scrumdevelopment@...
                                              To Unsubscribe, send a blank message to:
                                              scrumdevelopment-unsubscribe@...
                                              Yahoo! Groups Links
                                            • Joshua Kerievsky
                                              ... The sizing of iteration stories isn t impacted by iteration length, at least the way we do it. At the end of each iteration, a customer demonstrates the
                                              Message 22 of 27 , Feb 12, 2004
                                              • 0 Attachment
                                                Ken Schwaber wrote:

                                                >Does the length of the iteration have any impact on the granularity of the
                                                >input (Stories, Product Backlog). What is the type of demonstration that is
                                                >desired at the end of the iteration?
                                                >
                                                The sizing of iteration stories isn't impacted by iteration length, at least the way we do it. At the end of each iteration, a customer demonstrates the new features we've added to the system. Of late, the customer also demonstrates some of the FIT-based storytests we've produced. Upper management attends these iteration demonstrations. We find that they like to see both the system and a few of its new storytests.

                                                best regards,
                                                jk

                                                --
                                                I n d u s t r i a l L o g i c , I n c .
                                                Joshua Kerievsky
                                                Founder, Extreme Programmer & Coach
                                                http://industriallogic.com
                                                http://industrialxp.org
                                                866-540-8336 (toll free)
                                                510-540-8336 (phone)
                                                Berkeley, California
                                              • Devon Miller
                                                Actually, weeks are natural too. They correspond to the phases of the moon New moon, waxing quarter, full moon, waning quarter.
                                                Message 23 of 27 , Feb 14, 2004
                                                • 0 Attachment
                                                  Actually, weeks are natural too. They correspond to the phases of the moon
                                                  New moon, waxing quarter, full moon, waning quarter.

                                                  Steven Gordon wrote:

                                                  > That would argue that days and months (and years) are natural, but not
                                                  > weeks.
                                                  >
                                                  > -----Original Message-----
                                                  > *From:* Phlip [mailto:phlipcpp@...]
                                                  > *Sent:* Monday, February 09, 2004 9:52 AM
                                                  > *To:* scrumdevelopment@yahoogroups.com
                                                  > *Subject:* Re: [scrumdevelopment] Iteration Length
                                                  >
                                                  > Michael Campbell wrote:
                                                  >
                                                  > > Never knew that. Why is it then that only 1 month,
                                                  > > and then only
                                                  > > 75% of the time even HAS an integral # of them?
                                                  >
                                                  > Uh, because periods of the rotation of the Earth, its
                                                  > orbit, and the Moon's orbit all don't divide evenly?
                                                  >
                                                  >
                                                  > =====
                                                  > Phlip
                                                  > http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces
                                                  >
                                                • Phlip
                                                  ... Ken pointed out he forbade the Industrial Logic metaphor of nebulous units of time because nobody needed we got big NUTs getting back to any number of
                                                  Message 24 of 27 , Feb 14, 2004
                                                  • 0 Attachment
                                                    Devon Miller wrote:

                                                    > Actually, weeks are natural too. They correspond to
                                                    > the phases of the moon
                                                    > New moon, waxing quarter, full moon, waning quarter.

                                                    Ken pointed out he forbade the Industrial Logic
                                                    metaphor of "nebulous units of time" because nobody
                                                    needed "we got big NUTs" getting back to any number of
                                                    Human Resource Departments.

                                                    (I wax nostalgic for Omnigon's way cool HR department.
                                                    Whenever she wandered into the programming pit we'd
                                                    drop what we were doing and indulge in the most
                                                    advanced and witty HR harrassment. She know'd she'd
                                                    had to start this with her standard lecture at hiring
                                                    time; we all had a blast. But some HR is hired because
                                                    they help the brass think they'l reduce their odds of
                                                    getting sued.)

                                                    Workplace diversity requires us to allow folks who
                                                    don't "believe in" bicameralism, or this or that
                                                    diety, or lunar phases, to opt-out without limiting
                                                    their own career.

                                                    A lunar week is a nebulous unit of time - 7 days ~11
                                                    hours - and matching iterations to it would cause
                                                    friction on two fronts - the business calendar and the
                                                    diversity aspect.

                                                    But evoking biological circadian rhythms as the
                                                    natural basis for our calendar, and iteration lengths,
                                                    might be pragmatic.


                                                    =====
                                                    Phlip
                                                    http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                                                    __________________________________
                                                    Do you Yahoo!?
                                                    Yahoo! Finance: Get your refund fast by filing online.
                                                    http://taxes.yahoo.com/filing.html
                                                  • Doug Swartz
                                                    ... Some of us in Nebraska, various Canadian provinces, and 50 some democratic nations around the world, take issue with your blatant anti-unicameral bias. Our
                                                    Message 25 of 27 , Feb 14, 2004
                                                    • 0 Attachment
                                                      Saturday, February 14, 2004, 4:52:41 AM, Phlip wrote:

                                                      > Workplace diversity requires us to allow folks who
                                                      > don't "believe in" bicameralism, ....

                                                      Some of us in Nebraska, various Canadian provinces, and 50
                                                      some democratic nations around the world, take issue with your
                                                      blatant anti-unicameral bias. Our "uni's" are just as capable
                                                      of passing stupid legislation at the speed of a glacier, as
                                                      your "bi's"! But, we only have to pay one legislator, instead
                                                      of two.

                                                      The moderators of this esteemed list should take note of your
                                                      predilection to promote mono-culture and consider banning your
                                                      deviant opinions from the list.




                                                      --

                                                      ;-)
                                                      Doug Swartz
                                                      daswartz@...
                                                    • Phlip
                                                      ... And that was how, in early 2004, the beginnings of a schism appeared in the previously solid mailing list, ScrumDevelopment...
                                                      Message 26 of 27 , Feb 14, 2004
                                                      • 0 Attachment
                                                        Doug Swartz wrote:

                                                        > Phlip wrote:
                                                        >
                                                        > > Workplace diversity requires us to allow folks who
                                                        > > don't "believe in" bicameralism, ....
                                                        >
                                                        > Some of us in Nebraska, various Canadian provinces,
                                                        > and 50
                                                        > some democratic nations around the world, take issue
                                                        > with your
                                                        > blatant anti-unicameral bias. Our "uni's" are just
                                                        > as capable
                                                        > of passing stupid legislation at the speed of a
                                                        > glacier, as
                                                        > your "bi's"! But, we only have to pay one
                                                        > legislator, instead
                                                        > of two.

                                                        <ponderously>
                                                        And that was how, in early 2004, the
                                                        beginnings of a schism appeared in the
                                                        previously solid mailing list,
                                                        ScrumDevelopment...
                                                        </ponderously>

                                                        > The moderators of this esteemed list should take
                                                        > note of your
                                                        > predilection to promote mono-culture and consider
                                                        > banning your
                                                        > deviant opinions from the list.

                                                        Hey - my coffee is shade-grown, pal!

                                                        > ;-)
                                                        > Doug Swartz
                                                        > daswartz@...

                                                        ( http://eatthestate.org/05-22/FairTradeCoffee.htm for
                                                        those who didn't catch that last one... )

                                                        =====
                                                        Phlip
                                                        http://www.xpsd.org/cgi-bin/wiki?TestFirstUserInterfaces

                                                        __________________________________
                                                        Do you Yahoo!?
                                                        Yahoo! Finance: Get your refund fast by filing online.
                                                        http://taxes.yahoo.com/filing.html
                                                      Your message has been successfully submitted and would be delivered to recipients shortly.