Quiz — Chapter 42: The Future of COBOL
Multiple Choice
1. What is the estimated number of lines of COBOL in active production?
a) 22 million b) 220 million c) 220 billion d) 2.2 trillion
2. Which AI tool is specifically designed by IBM for COBOL modernization on z/OS?
a) GitHub Copilot b) ChatGPT c) watsonx Code Assistant for Z d) Amazon CodeWhisperer
3. In automated COBOL-to-Java conversion, which type of code has the highest conversion accuracy?
a) CICS online programs b) Complex business logic with nested IFs c) Simple I/O and MOVE statements d) VSAM file processing with REDEFINES
4. What is the primary risk of large-scale COBOL replacement projects?
a) COBOL developers refuse to participate b) The new language cannot handle the volume c) Incomplete business rule migration and project failure (40-60% failure rate) d) Cloud providers do not support financial applications
5. In a hybrid architecture, what role does the mainframe typically play?
a) User interface and mobile apps b) Machine learning and analytics c) Transaction processing and system of record d) Content delivery and caching
6. What is the most effective model for developing new COBOL talent?
a) Self-study with online videos only b) 4-year university degree in COBOL c) Apprenticeship combining formal training with mentored on-the-job experience d) Automated AI replacement of all COBOL developers
7. GlobalBank's 5-year modernization roadmap costs $4.2 million. The alternative full replacement was estimated at:
a) $8.5 million b) $85 million c) $850 million d) $8.5 billion
8. Which COBOL standard introduced JSON and XML support?
a) COBOL-85 b) COBOL 2002 c) COBOL 2014 d) These are vendor extensions (Enterprise COBOL 6.1+), not in the standard
9. What is Change Data Capture (CDC) used for in hybrid architectures?
a) Capturing changes to source code in Git b) Replicating mainframe data changes to cloud databases in near-real-time c) Converting COBOL data types to Java d) Monitoring production changes
10. According to the chapter, what is the single most critical variable in COBOL's future?
a) The performance of the z16 processor b) IBM's quarterly earnings c) The people — developers, architects, and organizations d) The price of cloud computing
True or False
11. AI can currently convert complex COBOL programs to Java with 100% accuracy. ____
12. COBOL developer salaries typically exceed those of Java developers with equivalent experience. ____
13. The COBOL language standard has not been updated since 1985. ____
14. Most organizations that maintain COBOL systems have decided to replace them entirely. ____
15. A COBOL developer's domain knowledge (understanding the business) is often more valuable than their language skills. ____
Short Answer
16. Explain why the talent gap is described as "not primarily a technical problem" but a "knowledge transfer problem."
17. In the debate about whether COBOL will still be running in 2050, present the strongest argument for each side.
18. Why do hybrid architectures (mainframe + cloud) make more economic sense than full replacement for most organizations?
19. How might AI tools change the COBOL talent equation? Consider both positive and negative effects.
20. Derek Washington says he wants to become "the person who speaks both languages — COBOL and modern." Why is this position strategically valuable?
Answer Key
1. c) 220 billion 2. c) watsonx Code Assistant for Z 3. c) Simple I/O and MOVE statements 4. c) Incomplete business rule migration and project failure 5. c) Transaction processing and system of record 6. c) Apprenticeship combining formal training with mentored on-the-job experience 7. c) $850 million 8. d) These are vendor extensions in Enterprise COBOL 6.1+, not yet in the ISO standard 9. b) Replicating mainframe data changes to cloud databases in near-real-time 10. c) The people
11. False — complex COBOL conversion accuracy ranges from 40-70% and requires significant human review. 12. True 13. False — COBOL 2002 and COBOL 2014 are subsequent standards with significant new features. 14. False — most organizations choose modernization (APIs, CI/CD, hybrid) over full replacement. 15. True
16. Learning the COBOL language takes months; it is well-documented and has clear syntax. The hard part is learning what the programs do, why they were designed that way, what edge cases they handle, and what business rules they encode. This domain knowledge accumulated over decades and exists primarily in the memories of experienced developers who are retiring. Transferring this knowledge requires time, structured mentoring, and organizational commitment.
17. For (still running): Economic argument — replacement is too expensive ($500M-$2B for large systems) with 40-60% failure risk. No compelling successor language. AI tools will reduce maintenance cost. Every decade's prediction of COBOL's death has been wrong. Against (replaced): 24 years is enough for AI conversion tools to mature, for the last experienced developers to retire, and for the economic equation to shift. Cloud platforms offer better cost-elasticity. At some point, the maintenance cost exceeds replacement cost.
18. Hybrid architectures cost a fraction of full replacement ($4.2M vs $850M in GlobalBank's case), deliver incremental value rather than big-bang risk, preserve proven business logic, allow each platform to do what it does best (mainframe for transactions, cloud for UX/analytics), and can be implemented gradually without disrupting operations.
19. Positive: AI makes code understanding faster (automated documentation), reduces the barrier for new developers (AI explains what code does), and may eventually enable reliable automated conversion. Negative: Organizations may over-rely on AI tools and underinvest in human expertise, AI-generated code may introduce subtle bugs in financial systems, and AI tools may create a false sense of understanding that does not include deep domain knowledge.
20. People who understand both COBOL and modern technology are extremely rare. Most COBOL developers do not know Kubernetes or CI/CD; most cloud developers cannot read COBOL. The person who bridges both worlds can lead modernization projects, design hybrid architectures, and translate between legacy and modern teams. This makes them indispensable during the decades-long modernization transition that most organizations are undergoing.