• Visitors can check out the Forum FAQ by clicking this link. You have to register before you can post: click the REGISTER link above to proceed. To start viewing messages, select the forum that you want to visit from the selection below. View our Forum Privacy Policy.
  • Want to receive the latest contracting news and advice straight to your inbox? Sign up to the ContractorUK newsletter here. Every sign up will also be entered into a draw to WIN £100 Amazon vouchers!

ThoughtWorks

Collapse
X
  •  
  • Filter
  • Time
  • Show
Clear All
new posts

    #51
    Originally posted by Antman View Post
    The way I see it is that although agile is the buzzword-du-jour and that's fine and dandy, how can any board sign-off on something that they don't know how much it'll cost and when it'll be delivered.

    Agile software development methodology is fine but how can you report on progress to shareholders? So you're always going to have a point as you go up the org. structure that the way the project gets done isn't important anymore but the dates and cost are. Hence frozen middle ground.
    Don't really agree with this.

    At the beginning of any project the cost and the time to be delivered are usually a guestimate (at best) based on the number of people available to work on the project, their skills, and prior delivery rate (assuming that's measured accurately and it's not a newly formed team).

    The work required for a project (product) to be delivered won't change whether you use Waterfall or Agile. The approach to delivering will change though. Having said that (and given the initial values for cost and delivery time are guestimates), in Agile you have the option to stop the project at any time given the progress, the benefits realised with incremental deliveries, etc.

    CFDs are quite good to report progress on an Agile project. You can also collate data and do it across projects under a Programme.

    How does that work in the Waterfall front?
    Are you not tied to get either all or nothing by a big-bang delivery?
    What metrics would you use for reporting progress?

    Comment


      #52
      Originally posted by ChrisFromGreece View Post
      Don't really agree with this.

      At the beginning of any project the cost and the time to be delivered are usually a guestimate (at best) based on the number of people available to work on the project, their skills, and prior delivery rate (assuming that's measured accurately and it's not a newly formed team).

      The work required for a project (product) to be delivered won't change whether you use Waterfall or Agile. The approach to delivering will change though. Having said that (and given the initial values for cost and delivery time are guestimates), in Agile you have the option to stop the project at any time given the progress, the benefits realised with incremental deliveries, etc.

      CFDs are quite good to report progress on an Agile project. You can also collate data and do it across projects under a Programme.

      How does that work in the Waterfall front?
      Are you not tied to get either all or nothing by a big-bang delivery?
      What metrics would you use for reporting progress?

      Progress would be compared to a baselined plan, including any scope creep registered.

      There are various GO/NO-GO decision points in waterfall to check that the benefits case will be realised. If you're partly-delivered in agile, what happens if you run out of money? Are you left with a part-completed product that is already partly live but doesn't deliver much compared to what it should have done?
      The greatest trick the devil ever pulled was convincing the world that he didn't exist

      Comment


        #53
        Originally posted by LondonManc View Post
        Progress would be compared to a baselined plan, including any scope creep registered.

        There are various GO/NO-GO decision points in waterfall to check that the benefits case will be realised. If you're partly-delivered in agile, what happens if you run out of money? Are you left with a part-completed product that is already partly live but doesn't deliver much compared to what it should have done?
        Running out of money seems a bit weird when you could have decision checkpoints every two weeks, which means you can cancel the project way before that point.

        Also given that the highest value/most risky items are delivered first, what remains to be delivered when money runs out should be the nice-to-haves. Unless the expected delivery date was way off and not all critical bits were delivered.

        Having said that, I would expect similarly money to run out on a Waterfall project and not having anything in production.

        Comment


          #54
          Can remember a large very critical migration I worked on last year, client had even brought in another 3rd party to run the migration, everything planned to last detail.

          On night of migration, main guy from another 3rd party ignores this plan and goes off and does his own thing, messing all the timings and deliverables up. I didn't sleep much that weekend.

          Agile/Kimball/Waterfall/ you can use whatever but when someone just decides to do their own thing...

          qh
          He had a negative bluety on a quackhandle and was quadraspazzed on a lifeglug.

          I look forward to your all knowing and likely sarcastic and unhelpful reply.

          Comment


            #55
            Originally posted by quackhandle View Post
            Can remember a large very critical migration I worked on last year, client had even brought in another 3rd party to run the migration, everything planned to last detail.

            On night of migration, main guy from another 3rd party ignores this plan and goes off and does his own thing, messing all the timings and deliverables up. I didn't sleep much that weekend.

            Agile/Kimball/Waterfall/ you can use whatever but when someone just decides to do their own thing...

            qh
            Everyone loves a maverick
            The Chunt of Chunts.

            Comment


              #56
              Originally posted by ChrisFromGreece View Post
              Running out of money seems a bit weird when you could have decision checkpoints every two weeks, which means you can cancel the project way before that point.

              Also given that the highest value/most risky items are delivered first, what remains to be delivered when money runs out should be the nice-to-haves. Unless the expected delivery date was way off and not all critical bits were delivered.

              Having said that, I would expect similarly money to run out on a Waterfall project and not having anything in production.
              You're clearly in a sheltered environment if you've never seen budget suddenly pulled.
              The greatest trick the devil ever pulled was convincing the world that he didn't exist

              Comment


                #57
                Originally posted by LondonManc View Post
                You're clearly in a sheltered environment if you've never seen budget suddenly pulled.
                Didn't say it doesn't happen... just trying to highlight the differences (in terms of items/features delivered by that point) between the two different methodologies.

                Comment


                  #58
                  Originally posted by ChrisFromGreece View Post
                  Didn't say it doesn't happen... just trying to highlight the differences (in terms of items/features delivered by that point) between the two different methodologies.
                  How does budget work with Agile?

                  With waterfall, you have a PID that includes a budget ask and a cost benefits case that needs to be signed off before you get very far.
                  The greatest trick the devil ever pulled was convincing the world that he didn't exist

                  Comment


                    #59
                    Originally posted by LondonManc View Post
                    How does budget work with Agile?

                    With waterfall, you have a PID that includes a budget ask and a cost benefits case that needs to be signed off before you get very far.
                    You'd think...

                    (Requirements out of that might be quite nice as well.)
                    "I can put any old tat in my sig, put quotes around it and attribute to someone of whom I've heard, to make it sound true."
                    - Voltaire/Benjamin Franklin/Anne Frank...

                    Comment


                      #60
                      Originally posted by cojak View Post
                      You'd think...

                      (Requirements out of that might be quite nice as well.)
                      A PID should be a reason to do (or not do) a project. What are we getting for the money and when, basically.
                      The greatest trick the devil ever pulled was convincing the world that he didn't exist

                      Comment

                      Working...
                      X