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

Re: [XP] Iteration Length

Expand Messages
  • Ron Jeffries
    ... I would expect that the number of questions in mid-flight would be reduced as the team learns more about the domain ... and as they all learn more about
    Message 1 of 49 , Jun 3, 2008
      Hello, Kim. On Tuesday, June 3, 2008, at 5:14:28 AM, you wrote:

      > On Tue, Jun 3, 2008 at 12:24 AM, Chet Hendrickson
      > <lists@...> wrote:

      >> Is one visit per iteration enough? Or, is that just the best we can get away with?

      > I wonder if the necessity of having the customer available 100%
      > declines as the rest of the team learns more about the customer's
      > expectations and about the domain in general?

      I would expect that the number of questions in mid-flight would be
      reduced as the team learns more about the domain ... and as they all
      learn more about how to communicate needs and wants.

      However, when the customer is away and a question comes up, what do
      we want to have happen? The alternatives seem to be:

      1. Wait, getting nothing done;
      2. Work on something of lower priority;
      3. Guess.

      None of these strike me as ideal ...

      > Could an initial effort where the customer is available full-time, say
      > 5-10 iterations, be enough, and that they just check in every once in
      > a while after that to answer questions and help out with priorities
      > (say twice per iteration). In my mind, they would need to be present
      > for the whole of iteration planning, anyway.

      Less might be enough depending what we mean by enough, I suppose.
      Insofar as we want to eliminate waste, it seems to me that having
      the customer less present works to increase waste, not decrease it.

      Ron Jeffries
      www.XProgramming.com
      Adapt, improvise, overcome.
      --Gunnery Sergeant Tom Highway (Heartbreak Ridge)
    • Kim Gräsman
      Hi Chris, ... I ve had some time to mull this over, and this is an excellent point. Within a single release plan, I think it may be doable, but then the
      Message 49 of 49 , Jun 15, 2008
        Hi Chris,

        On Thu, Jun 5, 2008 at 14:44, Chris Hulan <2chulan@...> wrote:
        >
        > From: Kim Gräsman <kim.grasman@...>
        > > ...
        > > from. The way I see it is, if everybody shares a common understanding
        > > of what is to be done and how, the *need* for collaboration is no
        > > longer as urgent. That's a big IF, right there, but I don't see it as
        > > ...
        >
        > I think the one problem here is the assumption that "what is to be done" is not going to change.
        > As soon as you give the customer "what they want"TM, they realize that what they really wanted was
        > something different and/or something more.
        >
        > You need the customer collaboration because the target is always moving.

        I've had some time to mull this over, and this is an excellent point.
        Within a single release plan, I think it may be doable, but then the
        customer would probably have to be re-established in the team.

        Also, my point that a single customer rep could run multiple projects
        in parallel seems less and less interesting, because multitasking
        generally does not make perfect.

        Many people with the organizational pull and knowledge to become
        Customers are fairly busy, so I suppose one could make the case that
        they could spend an initial time cultivating a team and then spend
        more time on their other responsibilities, but having them
        ever-present is surely better.

        Thanks for bouncing ideas,
        - Kim
      Your message has been successfully submitted and would be delivered to recipients shortly.