[unrev-II] Intrinsic and Extrinsic Ratings wrt IBIS

From: Eric Armstrong (eric.armstrong@sun.com)
Date: Wed Oct 31 2001 - 13:22:02 PST

  • Next message: Eric Armstrong: "Re: [unrev-II] Lucid Thinking about Traction Work Product"

    Good questions.

    In the textual representation I keep for design documents,
    there is a more less equal "weight" of the symbols:
       +: This is a thought that favors the idea
       -: This is a thought that discourages it.

    However:
      a) It is good to be able to record thoughts without
          determining in advance if they are for or against,
          and have the ability to decide that later.

      b) You are absolutely right that the content of a single
           item can outweigh a whole flock of others. There
           is no good way to cature that idea that I have seen
           so far.

    In part, I think that is a "background" thing, that depends
    in large part on your evaluation heuristics -- what you
    consider important. So this structure:
        i: An idea
           +: Produces an efficient implementation
           -: Doesn't have a very good interface.

    Produces a huge negative that totally outweighs the positive
    if you are an interface specialist or if you are designing a
    product in which the interface is crucial. On other hand,
    the positive totally blows away the negative if you are a
    performance specialist or building a batch-mode system
    that will be configured once and run a million times
    thereafter.

    Thinking about it, those observations capture the notion of
    :"intrinsic" vs. "extrinsic" (project-relative) rankings. A bubble
    sort has a + in that it is fast to code, and a negative in that it is
    inefficient for large amounts of unsorted data. Those are "intrinsic"
    evalations -- truisms that stand unchanged, regardless of
    circumstances.

    When making the decision as to which sort to implement
    for a particlar project, however, those rankings need to
    "feed into" the evaluation criteria, using either a verbal or
    automated equivalent of:
       1) This is a quick and dirty, one-time program, so
           a bubble sort is great. (Big weighting on the plus.)
    or
       2) This is something that will run for years on very
           large sets of data, so a bubble sort is really not a
           good idea (Big weighting on the negative).

    The intrinsic ratings represent "knowledge" -- things to
    know about bubble sort, in addition to the mechanics of
    constructing one.

    The extrinsic weightings, represent the product of intrinsic
    ratings relative to the project's design criteria -- the principles
    used to make selections among competing design alternatives.

    I suspect that many an IBIS-style conversation could be
    improved by clearly distinguishing the two kinds of
    information.

    ------------------------ Yahoo! Groups Sponsor ---------------------~-->
    Get your FREE VeriSign guide to security solutions for your web site: encrypting transactions, securing intranets, and more!
    http://us.click.yahoo.com/UnN2wB/m5_CAA/yigFAA/IHFolB/TM
    ---------------------------------------------------------------------~->

    Community email addresses:
      Post message: unrev-II@onelist.com
      Subscribe: unrev-II-subscribe@onelist.com
      Unsubscribe: unrev-II-unsubscribe@onelist.com
      List owner: unrev-II-owner@onelist.com

    Shortcut URL to this page:
      http://www.onelist.com/community/unrev-II

    Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/



    This archive was generated by hypermail 2.0.0 : Wed Oct 31 2001 - 13:09:29 PST