Skip to content

dr ses section 3

Dan Connolly edited this page Jul 19, 2026 · 2 revisions

Dr. SES — Section 3: Toward Distributed Electronic Rights

Source: Section 3 (pp. 10–12) of Distributed Electronic Rights in JavaScript, Miller, Van Cutsem, and Tulloh (ESOP 2013).

Parent page: Dr. SES

Object References as Rights

Dr. SES shows how JavaScript can become a distributed, secure, and resilient system. Its central observation is that an object reference represents a right: authority to perform a defined set of operations on a particular resource.

The paper connects this computational model to society's use of rights. Rights let people coordinate independent plans, prevent conflicting uses of resources, and cooperate despite mutual suspicion or competing goals. This social analogy suggests a way to secure distributed computation.

The Open-Access Problem

Every computational system must control access to shared resources. Unrestricted global mutable state resembles a tragedy of the commons: anyone may change it, so nobody can safely depend on it. Both malicious actions and ordinary bugs can create use conflicts.

The paper identifies two broad strategies for managing this problem:

Social strategy Computational analogue Method
Governance Access-control lists (ACLs) Define membership, exclude unauthorized users, and regulate each authorized user's actions
Property rights Object capabilities (OCaps) Partition authority among agents and govern how specific rights are acquired, transferred, and protected

ACL systems rely heavily on perimeter defenses and detailed rules for insiders. Governance can manage shared resources successfully, but increasing numbers of users, uses, and access changes put growing pressure on both the perimeter and the ACLs.

The property-rights strategy decentralizes control. It partitions a shared domain into areas controlled by particular agents, establishes general rules for acquiring and transferring rights, and lets each agent act independently within its rights. Contracts move rights between protected domains.

OCaps as Electronic Rights

In an OCap system, a local unforgeable object reference or remote unguessable reference is an electronic right (eright): the right to invoke the public interface of the designated object.

An OCap reference combines designation with authority. It is possessory:

  • Possessing the reference is sufficient to use it.
  • Its holder decides whether and how to exercise it.
  • Its holder may transfer it to another agent.

The paper compares OCap mechanics with private law:

Private law OCap system
Property determines how rights are initially acquired Object creation assigns initial authority
Contract governs transfers of rights Message passing transfers references
Tort protects rights against interference Encapsulation protects object state and authority

Where the Analogy Differs

Object references are normally easy to share: giving someone a reference usually copies it without depriving the giver. Rights over physical objects are normally exclusive because those objects are rivalrous; transferring a car title, for example, transfers exclusive access. Physical systems need extra legal structures to enable sharing, while computational systems face the reverse problem: sharing is easy and exclusivity is difficult.

Object references can nevertheless serve as building blocks for higher-level erights with different properties.

Money as a Higher-Level Eright

The paper motivates the next section by contrasting ordinary object references with money:

Object references Money Significance
Shareable Exclusive Payment requires the recipient to know the payer no longer controls the transferred funds
Specific Fungible Currency holders care about quantity and denomination, not a particular unit
Opaque Measurable A recipient must be able to verify the amount and currency received
Exercisable Symbolic A reference has value through operations it permits; money has value through exchange rather than direct use

Contracts manipulate rights. Each participant supplies rights to a contract; the contract's logic and the participants' choices determine which derived rights each participant receives. Because money forms one side of many exchanges, the paper next implements money as a smart contract before considering direct trade.

References Used by Section 3

Full citations appear in the paper's bibliography as references 14–23.

  • Jensen and Meckling on knowledge and organizational structure
  • Steiner on rights
  • Smith on exclusion and governance
  • Ostrom on governing commons
  • Hayek on abstract rules
  • Hardy on the confused deputy
  • Mossoff on property
  • Epstein on property, contract, and tort
  • Fox on property rights in money
  • Barnett on contract

Clone this wiki locally