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

Re: [agile-usability] Article on Design and Agile on A list Apart

Expand Messages
  • Adam Sroka
    Fundamentally, Agile is about delivering business value. The inherent assumption is that the most important thing a software team can do is deliver working
    Message 1 of 16 , Dec 2, 2008
    • 0 Attachment
      Fundamentally, Agile is about delivering business value. The inherent
      assumption is that the most important thing a software team can do is
      deliver working software. Those of us who believe this have taken up
      with the Agile movement and not looked back.

      However, from a macro-economic viewpoint the value in software
      development isn't always delivering working software. Despite all the
      dire predictions about the economy, the US government is still
      spending at unprecedented levels. Hundreds of billions of dollars go
      into government sponsored software projects every year. Yet, more than
      half of those will never deliver working software to an actual user.

      So, what is the value of those projects? Why wouldn't they just become
      Agile and deliver something useful. The reason is that the real value
      of those projects is in their ability to infuse large sums of money
      into a large bureaucracy to fuel it's continued existence. Even if
      they produce nothing they still keep people in jobs and money flowing
      through the economy. If every project had to produce value we would
      soon find that there wasn't enough value to be had to justify all of
      these jobs.

      So, when will Agile arrive? I believe that it already has. So why
      aren't we all doing it? Because the economy overvalues software
      development. Eventually there will come a day when the market will
      value software development services according to the value that they
      produce. Once that is true we will have to produce something valuable
      or find a new line of work. When will this happen? I don't know, but
      it happens in every new market and our day will come.

      On Tue, Dec 2, 2008 at 9:29 AM, James Page <jamespage@...> wrote:
      > Would be interesting here peoples views on the article.
      > http://www.alistapart.com/articles/gettingrealaboutagiledesign
      >
      > James
      >
    • William Pietri
      ... I couldn t disagree more strongly. As long as governments don t just use the savings from Agile methods to cut spending, then a switch to delivering actual
      Message 2 of 16 , Dec 2, 2008
      • 0 Attachment
        Adam Sroka wrote:
        > However, from a macro-economic viewpoint the value in software
        > development isn't always delivering working software. Despite all the
        > dire predictions about the economy, the US government is still
        > spending at unprecedented levels. Hundreds of billions of dollars go
        > into government sponsored software projects every year. Yet, more than
        > half of those will never deliver working software to an actual user.
        >
        > So, what is the value of those projects? Why wouldn't they just become
        > Agile and deliver something useful. The reason is that the real value
        > of those projects is in their ability to infuse large sums of money
        > into a large bureaucracy to fuel it's continued existence. Even if
        > they produce nothing they still keep people in jobs and money flowing
        > through the economy. If every project had to produce value we would
        > soon find that there wasn't enough value to be had to justify all of
        > these jobs.

        I couldn't disagree more strongly. As long as governments don't just use
        the savings from Agile methods to cut spending, then a switch to
        delivering actual value would be macroeconomically positive, not negative.

        Part of your analysis seems reasonable to me: sometimes the value of a
        project to a bureaucracy is in perpetuating the bureaucracy and
        advancing the career of its leaders by seeming big and important.

        Macroeconomically, though, there's no reason to waste money. Your take
        is a common one, but among economists it's known as the "broken window
        fallacy":

        http://en.wikipedia.org/wiki/Parable_of_the_broken_window

        It's true that during a recession the government shouldn't cut spending,
        as a counter-cyclical fiscal policy can lessen the length and severity
        of a recession. But the only reason to do a low-value project is if
        government is already spending all it can on higher-value projects. I
        guess that's theoretically possible, but off the top of my head I could
        name project after project that would be more beneficial to the public
        than the typical large-project development failure.

        Your analysis also neglects the human cost. If we employ a lot of people
        in jobs that are essentially valueless, then we've in effect trained an
        entire time-wasting army. So not only do we lose the resources spent on
        the individual project, but we severely impair the future productivity
        of the workers involved.


        Despite having advocated XP since 2000, I have never seen a company that
        used the efficiency gains to cut staff. Instead, they find other things
        for people to do, and better ways to spend their excess money. There's
        no reason governments couldn't do the same thing.

        William
      • Adam Sroka
        ... I think you misunderstood my intentions. I was attempting to describe the current situation and why I think Agile has not been more widely adopted. I was
        Message 3 of 16 , Dec 2, 2008
        • 0 Attachment
          On Tue, Dec 2, 2008 at 12:54 PM, William Pietri <william@...> wrote:
          > Adam Sroka wrote:
          >> However, from a macro-economic viewpoint the value in software
          >> development isn't always delivering working software. Despite all the
          >> dire predictions about the economy, the US government is still
          >> spending at unprecedented levels. Hundreds of billions of dollars go
          >> into government sponsored software projects every year. Yet, more than
          >> half of those will never deliver working software to an actual user.
          >>
          >> So, what is the value of those projects? Why wouldn't they just become
          >> Agile and deliver something useful. The reason is that the real value
          >> of those projects is in their ability to infuse large sums of money
          >> into a large bureaucracy to fuel it's continued existence. Even if
          >> they produce nothing they still keep people in jobs and money flowing
          >> through the economy. If every project had to produce value we would
          >> soon find that there wasn't enough value to be had to justify all of
          >> these jobs.
          >
          > I couldn't disagree more strongly. As long as governments don't just use
          > the savings from Agile methods to cut spending, then a switch to
          > delivering actual value would be macroeconomically positive, not negative.
          >
          > Part of your analysis seems reasonable to me: sometimes the value of a
          > project to a bureaucracy is in perpetuating the bureaucracy and
          > advancing the career of its leaders by seeming big and important.
          >
          > Macroeconomically, though, there's no reason to waste money. Your take
          > is a common one, but among economists it's known as the "broken window
          > fallacy":
          >
          > http://en.wikipedia.org/wiki/Parable_of_the_broken_window
          >
          > It's true that during a recession the government shouldn't cut spending,
          > as a counter-cyclical fiscal policy can lessen the length and severity
          > of a recession. But the only reason to do a low-value project is if
          > government is already spending all it can on higher-value projects. I
          > guess that's theoretically possible, but off the top of my head I could
          > name project after project that would be more beneficial to the public
          > than the typical large-project development failure.
          >
          > Your analysis also neglects the human cost. If we employ a lot of people
          > in jobs that are essentially valueless, then we've in effect trained an
          > entire time-wasting army. So not only do we lose the resources spent on
          > the individual project, but we severely impair the future productivity
          > of the workers involved.
          >
          > Despite having advocated XP since 2000, I have never seen a company that
          > used the efficiency gains to cut staff. Instead, they find other things
          > for people to do, and better ways to spend their excess money. There's
          > no reason governments couldn't do the same thing.
          >
          > William
          >

          I think you misunderstood my intentions. I was attempting to describe
          the current situation and why I think Agile has not been more widely
          adopted. I was also trying to discount the premise of TFA that current
          economic conditions compel us to be more Agile. I was not arguing that
          the current situation was reasonable or valid.

          It is possible that government projects could adopt Agile and produce
          more value. It is also possible that government projects that don't
          produce value could cease and that those people could find productive
          jobs elsewhere. I was merely making the case that there is currently
          no economic force compelling them to do either. (I also made the point
          that software development is overvalued and that when its valuation
          becomes more realistic other factors may be more compelling.)

          BTW, there are a lot of government projects that do produce value and
          many that have adopted Agile. However, these are not the majority by
          any stretch. Also, the same situation can and does exist outside of
          the government. I think that government tends to be a refuge for these
          projects, however, because their oversight is not motivated by
          profitability.
        • Ron Jeffries
          Hello, Adam. On Tuesday, December 2, 2008, at 3:37:25 PM, you ... Huh? Value isn t a finite quantity that projects draw down. Projects [can] /create/ value.
          Message 4 of 16 , Dec 2, 2008
          • 0 Attachment
            Hello, Adam. On Tuesday, December 2, 2008, at 3:37:25 PM, you
            wrote:

            > So, what is the value of those projects? Why wouldn't they just become
            > Agile and deliver something useful. The reason is that the real value
            > of those projects is in their ability to infuse large sums of money
            > into a large bureaucracy to fuel it's continued existence. Even if
            > they produce nothing they still keep people in jobs and money flowing
            > through the economy. If every project had to produce value we would
            > soon find that there wasn't enough value to be had to justify all of
            > these jobs.

            Huh? Value isn't a finite quantity that projects draw down. Projects
            [can] /create/ value.

            Any project whose sole value is that it is causing people to be paid
            would be of more value if it also built something useful.

            Ron Jeffries
            www.XProgramming.com
            www.xprogramming.com/blog
            The fact that we know more today, and are more capable today,
            is good news about today, not bad news about yesterday.
          • Adam Sroka
            On Tue, Dec 2, 2008 at 1:35 PM, Ron Jeffries ... Yes, but what compels them to do so? ... Can it be that some projects create /sufficient/ value to justify
            Message 5 of 16 , Dec 2, 2008
            • 0 Attachment
              On Tue, Dec 2, 2008 at 1:35 PM, Ron Jeffries
              <ronjeffries@...> wrote:
              > Hello, Adam. On Tuesday, December 2, 2008, at 3:37:25 PM, you
              > wrote:
              >
              >> So, what is the value of those projects? Why wouldn't they just become
              >> Agile and deliver something useful. The reason is that the real value
              >> of those projects is in their ability to infuse large sums of money
              >> into a large bureaucracy to fuel it's continued existence. Even if
              >> they produce nothing they still keep people in jobs and money flowing
              >> through the economy. If every project had to produce value we would
              >> soon find that there wasn't enough value to be had to justify all of
              >> these jobs.
              >
              > Huh? Value isn't a finite quantity that projects draw down. Projects
              > [can] /create/ value.
              >

              Yes, but what compels them to do so?

              > Any project whose sole value is that it is causing people to be paid
              > would be of more value if it also built something useful.
              >

              Can it be that some projects create /sufficient/ value to justify
              their own existence but not as much value as is /possible/? And, what
              if a project provides sufficient value (to someone) without providing
              value in the form of working software?
            • Adam Sroka
              P.S. More to the point: much of the value in working software is that is satisfies some business need. What happens if my business has more budget for software
              Message 6 of 16 , Dec 2, 2008
              • 0 Attachment
                P.S. More to the point: much of the value in working software is that
                is satisfies some business need. What happens if my business has more
                budget for software than it has need?

                On Tue, Dec 2, 2008 at 1:53 PM, Adam Sroka <adam.sroka@...> wrote:
                > On Tue, Dec 2, 2008 at 1:35 PM, Ron Jeffries
                > <ronjeffries@...> wrote:
                >> Hello, Adam. On Tuesday, December 2, 2008, at 3:37:25 PM, you
                >> wrote:
                >>
                >>> So, what is the value of those projects? Why wouldn't they just become
                >>> Agile and deliver something useful. The reason is that the real value
                >>> of those projects is in their ability to infuse large sums of money
                >>> into a large bureaucracy to fuel it's continued existence. Even if
                >>> they produce nothing they still keep people in jobs and money flowing
                >>> through the economy. If every project had to produce value we would
                >>> soon find that there wasn't enough value to be had to justify all of
                >>> these jobs.
                >>
                >> Huh? Value isn't a finite quantity that projects draw down. Projects
                >> [can] /create/ value.
                >>
                >
                > Yes, but what compels them to do so?
                >
                >> Any project whose sole value is that it is causing people to be paid
                >> would be of more value if it also built something useful.
                >>
                >
                > Can it be that some projects create /sufficient/ value to justify
                > their own existence but not as much value as is /possible/? And, what
                > if a project provides sufficient value (to someone) without providing
                > value in the form of working software?
                >
              • tmfspeck
                ... Then your business needs to learn to better budget ;)
                Message 7 of 16 , Dec 2, 2008
                • 0 Attachment
                  --- In agile-usability@yahoogroups.com, "Adam Sroka" <adam.sroka@...>
                  wrote:
                  >
                  > P.S. More to the point: much of the value in working software is that
                  > is satisfies some business need. What happens if my business has more
                  > budget for software than it has need?
                  >

                  Then your business needs to learn to better budget ;)
                • Ron Jeffries
                  Hello, Adam. On Tuesday, December 2, 2008, at 4:57:10 PM, you ... Never seen it happen but if that were true you could spend less or contract to do software
                  Message 8 of 16 , Dec 2, 2008
                  • 0 Attachment
                    Hello, Adam. On Tuesday, December 2, 2008, at 4:57:10 PM, you
                    wrote:

                    > P.S. More to the point: much of the value in working software is that
                    > is satisfies some business need. What happens if my business has more
                    > budget for software than it has need?

                    Never seen it happen but if that were true you could spend less or
                    contract to do software for someone else.

                    Ron Jeffries
                    www.XProgramming.com
                    www.xprogramming.com/blog
                    Accroche toi a ton reve. --ELO
                  • Adam Sroka
                    On Tue, Dec 2, 2008 at 2:36 PM, Ron Jeffries ... Where I have seen it, in the context of government, it looks like this: The process of Undulation is very
                    Message 9 of 16 , Dec 2, 2008
                    • 0 Attachment
                      On Tue, Dec 2, 2008 at 2:36 PM, Ron Jeffries
                      <ronjeffries@...> wrote:
                      > Hello, Adam. On Tuesday, December 2, 2008, at 4:57:10 PM, you
                      > wrote:
                      >
                      >> P.S. More to the point: much of the value in working software is that
                      >> is satisfies some business need. What happens if my business has more
                      >> budget for software than it has need?
                      >
                      > Never seen it happen but if that were true you could spend less or
                      > contract to do software for someone else.
                      >

                      Where I have seen it, in the context of government, it looks like
                      this: The process of Undulation is very expensive. It costs the
                      taxpayer millions of dollars every quarter. We would like to develop
                      the Super-new Obfuscated Undulator (SOU) so that we can increase the
                      speed and reduce the amount of error in typical Undulations thus
                      reducing costs. Congress has allotted umpteen million to this project
                      and we have contracted Gratuitous Spending Corporation (GSC) to come
                      up with a design. GSC has told us that they will spend the next two
                      years designing the project after which we will return to Congress to
                      ask for more money.

                      A couple of things about this hypothetical project:

                      1) The agency is quite capable of continuing to do "Undulation" the
                      way they have always done it. It is expensive, but they have the money
                      and will continue to get the money as long as they ask for it.

                      2) The project is justified by the perception that it could reduce the
                      cost of existing processes. Lower costs are good. Thus, it fills a
                      real need and it is easy to justify the additional expense in terms of
                      future value.

                      3) The process is a black box. GSC walks away with the money and has
                      very little to account for in terms of what happens to it. At most,
                      they need to have a plan outlining how the money will be spent and
                      they need to deliver some documents showing that someone spent some
                      time thinking about a solution.

                      4) The agency's continued funding depends on it's continued spending.
                      It too is not very accountable for what happens to the money so long
                      as it can outline a plan and deliver some documents saying what
                      happened.

                      5) All that anyone really cares is that the "Undulation" gets done
                      regardless of how it gets done or how much money is wasted in the
                      process. Thus, there is no force /compelling/ success. Regardless of
                      the success or failure of the software project business will continue
                      as usual. In fact, in the long run occasional failures produce greater
                      revenues for the agency. So long as "Undulation" continues this is
                      good for business.

                      I won't get into specifics, but this is something I have seen first
                      hand at more than one government office. It is also part of the reason
                      that I used to live in DC and now live in Los Angeles :-)
                    • Ron Jeffries
                      Hello, Adam. On Tuesday, December 2, 2008, at 4:53:02 PM, you ... Of course. All projects create less value than is possible. Many create sufficient value.
                      Message 10 of 16 , Dec 3, 2008
                      • 0 Attachment
                        Hello, Adam. On Tuesday, December 2, 2008, at 4:53:02 PM, you
                        wrote:

                        > Can it be that some projects create /sufficient/ value to justify
                        > their own existence but not as much value as is /possible/?

                        Of course. All projects create less value than is possible. Many
                        create sufficient value.

                        > And, what
                        > if a project provides sufficient value (to someone) without providing
                        > value in the form of working software?

                        Then it wouldn't make much sense to have a software project would
                        it?

                        Ron Jeffries
                        www.XProgramming.com
                        www.xprogramming.com/blog
                        The practices are not the knowing: they are a path to the knowing.
                      • Adam Sroka
                        On Wed, Dec 3, 2008 at 8:25 PM, Ron Jeffries ... Except that it is the Emperor s New Clothing. They have to believe we re selling something or we don t get
                        Message 11 of 16 , Dec 3, 2008
                        • 0 Attachment
                          On Wed, Dec 3, 2008 at 8:25 PM, Ron Jeffries
                          <ronjeffries@...> wrote:
                          > Hello, Adam. On Tuesday, December 2, 2008, at 4:53:02 PM, you
                          > wrote:
                          >
                          >> Can it be that some projects create /sufficient/ value to justify
                          >> their own existence but not as much value as is /possible/?
                          >
                          > Of course. All projects create less value than is possible. Many
                          > create sufficient value.
                          >
                          >> And, what
                          >> if a project provides sufficient value (to someone) without providing
                          >> value in the form of working software?
                          >
                          > Then it wouldn't make much sense to have a software project would
                          > it?
                          >

                          Except that it is the Emperor's New Clothing. They have to believe
                          we're selling something or we don't get paid. Software is the perfect
                          thing to be selling, because no one is quite sure what it is (The
                          magical stuff that the internet is made from.) And, everyone already
                          expects us to fail (After all it's /really hard/.)

                          I'm being cynical, and you don't have to join me. On the other hand,
                          cynics are people who have seen how screwed up the world can be.
                        • Ron Jeffries
                          Hello, Adam. On Thursday, December 4, 2008, at 12:02:31 AM, you ... Well, to operate that way, wouldn t one have to be living a lie? ... Everyone with their
                          Message 12 of 16 , Dec 4, 2008
                          • 0 Attachment
                            Hello, Adam. On Thursday, December 4, 2008, at 12:02:31 AM, you
                            wrote:

                            > Except that it is the Emperor's New Clothing. They have to believe
                            > we're selling something or we don't get paid. Software is the perfect
                            > thing to be selling, because no one is quite sure what it is (The
                            > magical stuff that the internet is made from.) And, everyone already
                            > expects us to fail (After all it's /really hard/.)

                            Well, to operate that way, wouldn't one have to be living a lie?

                            > I'm being cynical, and you don't have to join me. On the other hand,
                            > cynics are people who have seen how screwed up the world can be.

                            Everyone with their eyes open sees how screwed up the world can be.
                            The question is what do you do next.

                            Ron Jeffries
                            www.XProgramming.com
                            www.xprogramming.com/blog
                            Find the simple path to what works and follow it,
                            always looking for a simpler path. -- Patrick D. Smith
                          • Adam Sroka
                            On Thu, Dec 4, 2008 at 6:15 AM, Ron Jeffries ... More or less. It is a mob mentality. The more people you have working on something the less stake they have
                            Message 13 of 16 , Dec 4, 2008
                            • 0 Attachment
                              On Thu, Dec 4, 2008 at 6:15 AM, Ron Jeffries
                              <ronjeffries@...> wrote:
                              > Hello, Adam. On Thursday, December 4, 2008, at 12:02:31 AM, you
                              > wrote:
                              >
                              >> Except that it is the Emperor's New Clothing. They have to believe
                              >> we're selling something or we don't get paid. Software is the perfect
                              >> thing to be selling, because no one is quite sure what it is (The
                              >> magical stuff that the internet is made from.) And, everyone already
                              >> expects us to fail (After all it's /really hard/.)
                              >
                              > Well, to operate that way, wouldn't one have to be living a lie?
                              >

                              More or less.

                              It is a mob mentality. The more people you have working on something
                              the less stake they have individually. Add deep hierarchies and long,
                              slow feedback loops and you have an environment where no one has any
                              personal responsibility.

                              I think a lot of software people just want to build things. If the
                              "architect" comes and says, "I need a component that takes these
                              inputs and produces these outputs, and I need it in six weeks," they
                              are content to do that with little stake in how that relates to the
                              business and the success of the overall project.

                              I think the blame for mismanagement generally lies with such
                              "architects" and whoever is pulling their strings. Sometimes it is
                              malicious. Sometimes it is incompetence. Sometimes you have True
                              Believers who are sure that what they are doing is the Right Way to
                              develop software even though they've never personally seen it succeed.

                              >> I'm being cynical, and you don't have to join me. On the other hand,
                              >> cynics are people who have seen how screwed up the world can be.
                              >
                              > Everyone with their eyes open sees how screwed up the world can be.
                              > The question is what do you do next.
                              >

                              Personally, I make as much noise as I can and then try to get the next
                              gig lined up. At some point I'd like to settle down and take some
                              responsibility, but that hasn't been in the cards yet.
                            Your message has been successfully submitted and would be delivered to recipients shortly.