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

Re: [XP] XP Survival

Expand Messages
  • Brad Appleton
    ... Interesting question. I actually had a conversation about this with David Anderson (author of Agile Management , also see http://www.agilemanagement.net).
    Message 1 of 80 , Dec 1, 2004
    • 0 Attachment
      On Tue, Nov 30, 2004 at 07:26:06AM -0800, Stede Troisi wrote:
      > With the advent of Domain Specific Modeling, MDA,
      > Software factories and the wonderful tools included in
      > Visual Studio 2005 I was wondering if XP can or even
      > needs to survive in the next 50 years?

      Interesting question. I actually had a conversation about this
      with David Anderson (author of "Agile Management", also see
      http://www.agilemanagement.net). David is an FDD practitioner.

      In my opinion, one of the most fundamental differences
      between FDD and XP is what each perceives to be the
      central artifact and mechanism for conveying knowledge
      of the system.
      - with XP, it is the source code.
      - with FDD, it is the domain model

      Another interesting thing about FDD is several of its
      "color modeling" rules/patterns are really just refactoring
      rules being applied to the model rather than to code
      (IMHO).

      I believe that MDA/MDD wants not just the central
      mechanism for communicating system organization and
      design to be the model, but the ultimate goal is that the
      model will be the ONLY means necessary to do it.

      I do believe MDA/MDD, in conjunction with IDEs like
      Eclipse, will indeed reshape the face of software
      development within as little as 10 years. I think it's
      already started. A few select IDE's did it before, but
      with standards like Eclipse and UML it has already become
      more prevalent.

      I think the modeling environment will merge with the coding
      environment such that most developers will rarely have
      to bother to think in terms of files and directories when
      organizing and navigating their code. The UML model and a
      logical "package" hierarchy will be the primary means of
      navigating to a particular snippet of code.

      But before we get to the point where 100% code generation
      is the norm, we'll end-up using the model for the
      high-level stuff, and much of the interface, and the we'll
      simply double-click on a class or a method to enter the
      code for a given method. Eventually the difference between
      the programming language used, and the constraint/action
      specification language used for MDA will get smaller and
      smaller. (It's already happening with some tool vendors,
      where they are using a java-like syntax since the action
      specification semantics in UML don't specify enough to do
      it all on its own).

      This combined with increasing integrations with tools like
      JUnit and Hibernate will likely result in an MDA-way of
      doing TDD with the constraint language and some other code.

      The model will also give multiple ways of viewing
      the code structure (not simply a single hierarchy of
      files+directories). One of those views might be from a CM
      build/deployment perspective. And the modeling tool will
      automatically know a lot of the physical dependencies
      since it will take care of creating & generating the
      files+directory structure.

      And the more those things get integrated together as
      part of an overall model, the more the underlying IDE
      will be able to automatically do lots of traceability
      that otherwise would be painfully tedious and manual,
      and which may also be increasingly likely to be mandated
      by the customer due to things like Sarbanes/Oxley.

      I think XP will likely learn to adapt pretty well to this given
      that:
      - the creators of XP had much of their experience using
      Smalltalk and its IDE, which already did a lot of this
      stuff anyway
      - XP itself embraces adapting itself via reflection & retrospection

      --
      Brad Appleton <brad@...> www.bradapp.net
      Software CM Patterns (www.scmpatterns.com)
      Effective Teamwork, Practical Integration
      "And miles to go before I sleep." -- Robert Frost
    • Jim Standley
      Gartner Group evaluates our Model Execution System and declares a 12 to 47 times improvement over other development approaches. So do they recommend I cut my
      Message 80 of 80 , Dec 5, 2004
      • 0 Attachment
        "Gartner Group evaluates our Model Execution System and declares a 12 to
        47 times improvement over other development approaches." So do they
        recommend I cut my development budget by 92% or 98%?

        Israel Antezana wrote:
        > Steven,
        >
        > The ONME software is not a demo tool its a commercial tool ready to be used. But there is a demo about how to use this tool on its website (www.care-t.com)
        >
        > It was the first time i worked with this tool, and when worked with it I noticed it really boosted my productivity, actually this company has passed a Gartner group evaluation in which Gartner declares a 12 to 47 times improvement when using this MDA tool.
        >
        > I could use the tool in an iterative way, working in an increment on the model and generating the code in each increment with good results. Although I did not work in a test driven way, I am interested in exploring alternatives, if necessary, about using a TDD aproach in this context.
        >
        > Best regards,
        >
        > Israel Antezana
        >
        >
        >
        > Steven Gordon <sagordon@...> wrote:
        >
        > Isreal,
        >
        > Since you have actually been using this demo tool, please tell us
        > about your actuall experiences using this tool.
        >
        > Did this tool affect how you went about doing your project? Did
        > the tool make it difficult for you to implement your research
        > project one requirement at a time (or are you implementing your
        > research project in a waterfall fashion)? Did this tool help,
        > hinder, or have no effect on how you defined tests for your
        > software?
        >
        > Steven A. Gordon, Ph.D.
        > Manager, Software Factory
        > Arizona State University
        > PO Box 875506
        > Tempe, AZ 85287-9509
        > http://sf.asu.edu
        > (480)-727-6271
        >
      Your message has been successfully submitted and would be delivered to recipients shortly.