2000 to 2009: The Rise of Social Networks and Economic Crises - The Technical and Social History of Software Engineering - page 306

Function points for normalization combined with defect removal efficiency are
the best for quality metrics. Work hours per function point and function points
per staff month are the best productivity metrics. Benchmark data expressed using
IFPUG function points outnumber other function point variations by about 10 to
1.
The same 1,000 function point application would look like this:
• Source code for 1,000 function points: 40,000 logical code statements
• Programming language: C # and MySQL
• Reuse percentage: 0% to 40%
• Methodology: Agile with Scrum
• Productivity: 9.50 function points per staff month
• Defect potentials: 3.50 defects per function point
• Defect removal efficiency (DRE): 92%
• Delivered defects: 0.28 defects per function point
• Ratio of development personnel to maintenance:
• Development: 45%
• Maintenance: 55%
The following are the background data for 2005:
• Average language level: 8.00
• Number of programming languages: 2,500
• Logical statements per function point: 40
• Average application size: 550 function points
• Average application size: 22,000 logical code statements
Summary
By the end of this decade, social networks such as Facebook and Twitter had be-
come a new way of social interaction, with millions of subscribers and sometimes
many hours per day devoted to cybercommunications; often, more time was spent
online than in face-to-face communications.
Open-source applications such as Firefox and Linux began to have corporate
significance and were added to thousands of corporate portfolios.