Summary

  • RFC 1430 proposed a global Internet X.500 directory in February 1993. Its authors said the labor of mapping data into a coherent information framework would far outweigh the technical work of installing servers and user agents.
  • The proposed operating bargain divided responsibility: large organizations would run their own directory branches, smaller ones could rent space, and central operators would maintain root, naming, location knowledge and replication services. Funding and data quality sat inside the design, not at its edge.

The server could be ordered. The directory still needed somebody to decide where a person belonged, convert existing records, correct the next change of job and answer for the wrong entry. That was the awkward bill in the February 1993 plan for an Internet X.500 Directory Service. Its five authors stated the cost hierarchy plainly: mapping data into a coherent distributed framework was operational work, and it would far outweigh setting up servers and user agents.

The plan began with plausible uses. A white-pages lookup could find people. A directory could support X.509 authentication, with resource lookup, infrastructure information and mail applications to follow. Several cooperating pilots already existed. Yet an operational Internet-wide service required sufficient funding for central services. A pilot proved that components could talk; it did not hire the people who would keep a global address book current.

The proposed division of labor was concrete. A large institution would maintain its portion of the Directory Information Tree on its own Directory System Agent, or DSA. A smaller institution would rent DSA space from a regional network or another provider. That arrangement put the entry near its owner while making hosting available to organizations without their own machine and staff. The word “rent” exposed an economic decision beneath the technical one: inclusion would require a custodian and a continuing budget.

The branches also needed common plumbing. RFC 1430 called for a root DSA, name assignment, knowledge of which DSA held which portion of the tree, and replication of information high in the tree. Paradise was then managing a root DSA and was proposed as the initial basis for central operations. Participating countries would need corresponding services at national level. Each new country branch would enlist central help to install them. A directory could distribute data custody while still concentrating some coordination work.

Then came the records themselves. The memo warned that the whole infrastructure stood little chance without data management and procedures for quality. Owners would maintain their own data, so one uniform procedure would not fit every branch. The authors called for guidance, a minimum quality level, pilot projects connecting existing databases and reusable tools. They also wanted an executive summary persuasive enough to make computer-center managers invest manpower. The missing input was not another server specification. It was time on an institution's payroll.

People data sharpened the obligation. The authors asked national directory operators to examine local privacy and registration law and proposed a user bill of rights. Their plan named legal research as part of deployment, along with routing, schema and protocol work. A searchable entry about a person could be useful and still be costly to collect, update and publish responsibly.

RFC 1430 was a plan, not a balance sheet or a victory report. It did not count the staff needed for each branch or show that every country funded one. But it identified the weak point with unusual honesty: a global directory would fail in the space between a working server and a maintained record. The institution that wanted the lookup had to budget for the clerk as well as the machine.