ToppyMicroServices

RFC 9518 Quiz (EN)

Centralization, Decentralization, and Internet Standards

ENJA
0 / 0

References (URLs)

Q1: Does this redundant directory remove the centralization concern?

Multiple Choice

One company operates an A2A directory on ten regions with independent failover. Only that company can admit agents, inspect all queries, and change ranking policy.

**Explanation:** A: RFC 9518 Section 2 separates technical distribution from control. Multiple components can still leave one entity with exclusive power over a function. B: Section 2 defines centralization by exclusive observation, capture, control, or rent extraction. The scenario retains those powers despite regional redundancy. C: Different databases can reduce monoculture risk, but they do not by themselves distribute admission or governance authority. D: Sections 2.1 and 2.2 reject a blanket equation between centralization and harm. The function and trade-offs still need contextual analysis. Decision rule: Name the function and the power being concentrated; server count alone cannot answer the control question.

Q2: How should a central rendezvous service be evaluated?

Multiple Choice

Independent agents need a globally consistent place to find short-lived endpoints. A central service provides that rendezvous, publishes an audit log, and supports export to compatible alternatives.

**Explanation:** A: RFC 9518 Section 2.2 recognizes that centralized functions can be technically helpful or necessary. The document does not impose a universal prohibition. B: A useful global function can justify concentration, but Section 2.2 still points to accountability and the ability to switch as important constraints. C: Section 3.1.2 explains that distributed consensus can decentralize operation while governance, implementation, or other functions remain centralized. D: Sections 2.2 and 4.4 support this function-specific review. The conclusion depends on necessity, checks on power, available alternatives, and substitution cost. Decision rule: Centralization can be beneficial; the review asks whether its benefit, accountability, and switchability justify the particular concentration of power.

Q3: What is the strongest justified finding about this federated Agent Card design?

Multiple Choice

The protocol allows any provider to host Agent Cards. In practice, the default client trusts one provider's index, identity rules, and proprietary reputation extension; exporting reputation is unsupported.

**Explanation:** A: RFC 9518 Sections 3.1.1 and 4.4 say federation creates conditions for decentralization but does not guarantee the outcome. Defaults and portability affect practical substitution. B: Section 3.1.1 explicitly warns that non-technical pressure can concentrate a federated function. Protocol capability is not deployment evidence. C: RFC 9518 is an analysis and design-guidance document, not a rule that automatically bans proprietary extensions. Section 4.7 instead asks designers to examine their effects. D: Section 2 analyzes control over functions, and Section 4.4 includes substitution cost. Discovery and reputation can matter even with many hosting endpoints. Decision rule: Evaluate decentralization function by function and include defaults, missing portability, and the real cost of changing providers.

Q4: Which revision best constrains this optional A2A broker?

Multiple Choice

A user explicitly selects a broker to hide its network address and queue tasks while offline. The broker also receives complete task bodies and may reuse them for unrelated analytics.

**Explanation:** A: RFC 9518 Section 4.5 notes that third parties can provide useful specialization, simplicity, or privacy. The document does not ban the intermediary role. B: Positive user action justifies interposition for the chosen function, not unlimited secondary use. Delegated power still needs boundaries. C: Section 4.5 gives both criteria: a primary party positively chooses interposition, and observation or control is limited to what the intended function needs. D: Automatic opaque interposition fails the positive-action criterion and can increase sticky dependence and unaccountable observation. Decision rule: Delegation is not inherently harmful; evaluate informed choice, minimum necessary power, and the ability to revoke or replace the intermediary.

Q5: What does distributed consensus establish in this A2A network?

Multiple Choice

Task receipts are committed by a diverse validator set. One vendor controls the protocol repository, release signing, client implementation, and emergency upgrade process.

**Explanation:** A: RFC 9518 Section 3.1.2 warns that distributed operation does not automatically decentralize governance, shared implementations, or documentation. B: Section 3.1.2 supports this bounded conclusion. The evidence concerns one function, while other functions retain a distinct concentration of power. C: Central governance is a risk but does not by itself prove control over every validator decision. The analysis should not infer more than the observed mechanism. D: RFC 9518 treats protocols and applications as compositions of many functions. Governance and implementation are explicit parts of its analysis. Decision rule: A decentralization mechanism supplies evidence about the function it changes; it does not transfer that conclusion to every adjacent function.

Q6: Which conclusion follows from this composed A2A architecture?

Multiple Choice

A central directory gives fast abuse response and globally consistent names. Agent messaging is interoperable across two independent implementations, directory records are exportable, but all default clients still select the same directory.

**Explanation:** A: Section 2.2 recognizes specialized security and global functions as possible benefits, but those benefits do not erase concentrated control or failure impact. B: RFC 9518 does not require one decentralization technology. Sections 2.2 and 4.2 call for contextual, function-specific evaluation. C: Implementation diversity helps messaging, but Section 4.4 asks whether users can actually substitute deployments. It does not establish discovery diversity by implication. D: Sections 2.1, 2.2, 4.2, and 4.4 support this balanced review. The Status of This Memo and Section 1 identify RFC 9518 as Informational and as the author's analysis, not a universal compliance test. Decision rule: Use RFC 9518 as a review lens: state what improves, where power remains, what can be switched, and what the evidence cannot decide by itself.