Skip to content

Latest commit

 

History

History
36 lines (32 loc) · 1.51 KB

MythicalMan-Month.md

File metadata and controls

36 lines (32 loc) · 1.51 KB
Raw

Mythical Man-Month

Chapter 2 : The Mythical Man-Month

  • The author had been successfully using the following rule of thumb for scheduling a software task:
    • 1/3 planning
    • 1/6 coding
    • 1/4 component test and early system test
    • 1/4 system test, all components in hand
  • Oversimplifying outrageously, we state Brooks's Law:
    • Adding manpower to a late software project makes it later.

Chatper 3 : The Surgical Team

  • Productivity variations (best vs worst)

    • 10:1 on productivity
    • 5:1 on program speed and space
    • The data showed no correlation whatsoever between experience and performance (It seems quite shocking to me and I agree with that this would not be universally true)

  • Keep as few minds as possible

    • 200-man project with 25 managers $\rightarrow$ Fire 175 troops and put the managers back to programming?
    • The program is possible to be remained obsolete since few minds take more time even if they are sharp and the communication cost decreased dramatically

  • Mills's proposal: Reconciling between two needs: keeping few minds vs timely requirements

    • Each segment of a large job is tackled by a team
    • Each team to be organized like a surgical team
    • The team consists of many people who work in their professionals
      • But the system is the product of one mind (The surgeon) or at most two.
graph BT;
    A[Surgeon: cheif programmer] --> B[Copilot];
    B --> A;
    C[Administrator] --> A;
    D[Editor] --> A;
    E[Toolsmith] --> A;
    F[Tester] --> A;
    G[...] --> A;
Loading