No Silver Bullet Refired

1
No Silver Bullet Refired

• Chapter Summary
• Prepared by Ocdt Bowie

2
Overall

• Many comments and criticisms have been published
  about NSB
• Nine years later lets see whos right and whos
  wrong

3
Misunderstood Writing

• Accident- not misfortunate but rather incidental
• Essential- the nature of software engineering
  that can never be removed
• Accidentals do not account for more than 9/10 of
  the effort, therefore the essentials must be
  attacked (as well)

4
Misunderstandings (2)

• Essential ! Hopeless
• Essentials can be improved but not removed
• Complexity
• From implementation, not conformity
• Can possibly be simplified into axioms like
  science

5
Misunderstandings (3)

• Skepticism ! Pessimism
• No Silver Bullet seems to be ready to emerge
  within ten years
• Pipe-dreaming inhibits forward progression

6
Gloominess

• Separation of essence and accident
• One is literally the antonym of the other
• Silver bullets addressed in isolation
• They are proposed in isolation so they should be
  critiqued in the same manner
• Only ten year prediction
• Does not want to over estimate his prediction
  skills

7
Harels Thought Experiment

• False assumptions
• Accidentals were much greater in the 1950s and
  have been resolved
• Program systems were larger than Harel gives them
  credit

8
Harels Ideas

• The Vanilla Framework will address modeling of
  the essence, proper crafting and debugging of
  concepts
• Software can be represented topographically but
  cannot be bounded by three-space

9
Jones Point

• Focus on quality, not productivity

10
How are we doing?

• Have not seen an order-of-magnitude improvement
• Buying of-the-shelf seems to have been effective
• Reverse engineer simpler products

11
Object-Oriented Techniques

• It has been effective and shows promise
• Slow to develop
• Costly to train employees and slow to bring profit

12
Reuse

• Reuse saves time and effort
• Must be done systematically to avoid redundancy
• Many forms of reuse and many different outcomes

13
Vocabulary

• The need for a larger vocabulary has become
  greater with more complicated languages
• This makes reuse difficult
• Language must be studied to increase learnability
  and standards

14
Position Unchanged

• There is no silver bullet for the next ten years
  that will magically solve our problems