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

XP in fixed-price bids.

Expand Messages
  • Alain Kreienbuhl
    Hello, I d like to have some feed back on fixed-price bid projects that are using XP. How does the planning game in such settings? Especially the initial
    Message 1 of 5 , May 29, 2000
    • 0 Attachment
      Hello,

      I'd like to have some feed back on fixed-price bid projects that are using XP. How
      does the planning game in such settings? Especially the initial planning needed for the
      offer.

      Thanks.

      Alain.


      Adam Spitz wrote:

      > Our first iteration is done, on our first experimental XP-ish project. I
      > thought it might be useful for the XP newbies and skeptics to hear how it
      > went, and useful for us to hear what you think of what we've been doing.
      >
      > We're not doing real XP. (We're all well aware of that, so, Mr. Jeffries,
      > please don't worry that we'll fail and blame XP.) It's a weird project. Our
      > program is already written; our job is to make it work with various
      > different brands of databases. To do that, we'll need to test our stuff,
      > and the company is willing to invest in making those tests automated. So
      > our job is to write some automated tests for a whole mess of ugly,
      > mostly-working C++ code.
      >
      > Pairing is great. All four of us like it. We're definitely going faster
      > this way.
      >
      > The planning game is going well. Initially, we had a little bit of trouble
      > figuring out who our Customer was. We thought it was our Marketing
      > department, until we realized that our job was mostly to write tests. So we
      > went to our QA department - they'd done lots of manual testing of our
      > product, so they knew what to test, and they knew which tests were most
      > important (i.e. most likely to break).
      >
      > At first, our QA person was reluctant to be the customer. He didn't want
      > anything to do with XP. After a few tries to get him on board, we stopped
      > using the XP vocabulary around him, and just asked him the questions we
      > wanted him to answer. He was glad to help. Yesterday we had our second
      > iteration planning meeting, and he seemed to enjoy it. I think he likes
      > bossing us around. :)
      >
      > Estimating is going well. It took a while to explain to everybody how to
      > make up the initial estimates (ideal days aren't as intuitive to everyone
      > as I thought they would be), but they turned out pretty well - we guessed
      > we could do 13 points, and ended up doing 16.
      >
      > Our cowboy is getting better. We had a little meeting where we argued about
      > coding style and design principles, and we explained to him that one of our
      > goals was to have our tests usable in the future, by people other than us.
      > He's smart enough to know that the code he writes isn't so readable. He's
      > still not happy about it, but sometimes he even writes comments and
      > extracts methods without being asked to.
      >
      > We're doing continuous integration reasonably well. The one problem we
      > notice is that we have a lot of configuration problems - things work fine
      > on the machine they're coded on, but break elsewhere because something
      > outside the code needs to be done. This is annoying, but it hasn't really
      > been slowing us down much, so we're not too worried. We're experimenting
      > with making some of the configuration automated, but we don't know how to
      > do all of it that way.
      >
      > We're bad at doing the simplest thing that could possibly work, especially
      > when it comes to debugging. We're good at *intending* to do the simple
      > thing, but sometimes we're wrong and the thing isn't so simple after all.
      > By the time we notice this, it's usually taken us a lot more time than we
      > wanted to spend. We're experimenting with using our watches to keep us on
      > track - set the countdown timer for two or five or ten minutes, and if it
      > beeps before we've fixed the problem, we know we should try to think of
      > something simpler.
      >
      > We don't know how to do acceptance tests for this project. Our product IS
      > the set of tests. How do you write tests to make sure that you've got
      > enough tests?
      >
      > Any advice you can give us would be appreciated. I'm sure that a lot of our
      > problems are things that you've encountered before.
      >
      > Adam Spitz
      >
      > To Post a message, send it to: extremeprogramming@...
      >
      > To Unsubscribe, send a blank message to: extremeprogramming-unsubscribe@...
      >
      > Ad-free courtesy of objectmentor.com
    • John Brewer
      ... The short answer is, Just say no to fixed price bids. Fixed price software contracts are a sucker bet. The only thing you can do is try to make sure
      Message 2 of 5 , May 29, 2000
      • 0 Attachment
        --- In extremeprogramming@egroups.com, Alain Kreienbuhl
        <alain.kreienbuhl@f...> wrote:
        > I'd like to have some feed back on fixed-price bid projects that are
        > using XP. How does the planning game in such settings? Especially the
        > initial planning needed for the offer.

        The short answer is, "Just say no to fixed price bids." Fixed price
        software contracts are a sucker bet. The only thing you can do is try to
        make sure the other party is the sucker.

        Either the developer doesn't spec the contract clearly enough, and ends up
        doing additional work for free, or the developer nails down all the specs
        ahead of time, which means the customer gets just what he initially asked
        for, which is almost certainly not what he actually ends up needing.

        For a longer answer, as well as some suggested alternatives, see Kent
        Beck's essay "Optional Scope Contracts" at:

        http://www.egroups.com/files/extremeprogramming/Optional+scope+contracts.pdf


        John Brewer
        Jera Design

        JeraWorks, the cross-platform outliner.
        Now available for beta test at: http://www.jera.com/jeraworks/
      • Ian Hobson
        In article , Alain Kreienbuhl writes ... I expect the gurus (I m not one) will slam fixed price
        Message 3 of 5 , May 31, 2000
        • 0 Attachment
          In article <39327D95.2BCCB538@...>, Alain Kreienbuhl
          <alain.kreienbuhl@...> writes
          >Hello,
          >
          >I'd like to have some feed back on fixed-price bid projects that are using XP.
          >How
          >does the planning game in such settings? Especially the initial planning needed
          >for the
          >offer.
          >
          I expect the gurus (I'm not one) will slam fixed price for the whole
          thing probably because someone always gets screwed. BUT...

          I think things might work if you charged a fixed amount per "ideal day"-
          and made each iteration fixed price.

          Has anyone got any experience of this way of working?

          Regards

          Ian

          The Times Law Reports, April 7th 1999, when reporting "Express & Echo
          Publications v Tanton", Court of Appeal 1999, on page 279, "[W]here a
          person... was not required to provide those services personally it... [is]
          wholly inconsistent with the contract being a contract of service... [therefore]
          the only conclusion... was that the contract was one for services."
        • Ron Jeffries
          ... As far as I know, no one has experience yet with this. Others may chime in and prove me ignorant, which in this case would be fine. In any case, the above
          Message 4 of 5 , Jun 1 3:44 AM
          • 0 Attachment
            At 11:18 PM 5/31/2000 +0100, Ian Hobson wrote:
            >I think things might work if you charged a fixed amount per "ideal day"-
            >and made each iteration fixed price.
            >
            >Has anyone got any experience of this way of working?

            As far as I know, no one has experience yet with this. Others may chime in
            and prove me ignorant, which in this case would be fine.

            In any case, the above is essentially what Kent recommends. A fixed price
            for a fixed short time (a few iterations), if you're satisfied, renew.

            Ron Jeffries
            www.XProgramming.com
          • Simon Smith
            ... Check out Ken Auer s company, RoleModel Software: http://www.rolemodelsoft.com/DoingBusiness.htm Look at Deliverables-Based pricing; this looks to be fixed
            Message 5 of 5 , Jun 1 4:13 AM
            • 0 Attachment
              > <alain.kreienbuhl@...> writes
              > >Hello,
              > >
              > >I'd like to have some feed back on fixed-price bid projects
              > that are using XP.
              > >How
              > >does the planning game in such settings? Especially the initial
              > planning needed
              > >for the
              > >offer.
              > >
              > I expect the gurus (I'm not one) will slam fixed price for the whole
              > thing probably because someone always gets screwed. BUT...
              >
              > I think things might work if you charged a fixed amount per "ideal day"-
              > and made each iteration fixed price.
              >
              > Has anyone got any experience of this way of working?
              >
              > Regards
              >
              > Ian

              Check out Ken Auer's company, RoleModel Software:
              http://www.rolemodelsoft.com/DoingBusiness.htm

              Look at Deliverables-Based pricing; this looks to be fixed price per
              iteration.


              ---Simon

              mailto:Simon@...
              www.quintuslink.com
            Your message has been successfully submitted and would be delivered to recipients shortly.