Why Amsterdam Became Europe’s Number Registry Capital

Summary

• On 5 May 1992, the RIPE NCC reported an Amsterdam contact address, three permanent staff, a formal RARE framework and first-year funding from RARE national members and EARN. Its declared remit covered registry administration, databases, DNS coordination, statistics, software and document access.
• The 1991 establishment plan separated finance, personnel and site responsibilities. It proposed preliminary annual costs of ECU 240,000, a three-year funding objective of ECU 720,000, open staff and site solicitations, mixed boards and a manager-first appointment order.
• By 1996–97, an operation expected to exceed 30 permanent staff and carrying an ECU 2 million expense budget required a legal container capable of organising contributors, employees, management authority, liabilities, reserves and service agreements.
• The association established on 12 November 1997 joined an Amsterdam seat to contractual membership, member government, board representation, delegated management and formal record duties. It made an established location legally durable while leaving the original comparative site reasoning unresolved.

A technical setting before an institutional capital

Amsterdam’s place in this history began with a modest institutional problem. European Internet Protocol networks were multiplying across universities, research institutes, national organisations and specialist providers, yet their ownership remained dispersed. Number registration, routing information, domain coordination, contact data and technical publication crossed organisational boundaries. A permanent centre promised continuity between meetings without taking ownership of the networks it helped coordinate. The prospective office would sit among autonomous operators and maintain information needed for cooperation.

Its influence would initially arise from the usefulness of that work, not from corporate command over the underlying infrastructure.

The distinction matters for understanding what kind of “capital” Amsterdam eventually became. The RIPE NCC was conceived as a focal point for coordination in a technical environment containing many owners, sponsors and communities. Concentrating staff and records in one city could give the location institutional gravity even though cables, computers, operators and users remained distributed. The centre’s importance would depend on maintaining shared administrative surfaces and durable contacts across that distribution.

Amsterdam could therefore become central to European number-resource coordination without becoming the territorial centre of a single European network. Amsterdam already belonged to the networking landscape in which such an office could make sense. CWI’s institutional history dates its connection to NSFNET to 17 November 1988 and describes the Amsterdam institute as the first organisation outside the United States connected to that network. The underlying agreement is outside the available material, so the claim remains CWI’s account of its own achievement.

It nevertheless places consequential transatlantic networking activity in the city before the RIPE NCC opened and establishes a concrete setting in which international connectivity and technical expertise were already visible.

Physical proximity to research-network experience offered an intelligible advantage. Engineers, institutions, communications links and organisations accustomed to cooperative technical work could lower the friction involved in assembling a small professional centre. Such conditions could help with recruitment, access to knowledge and everyday contact among people working across institutional boundaries. Amsterdam’s networking ecology accordingly made it a credible place in which coordination capacity could be established.

Credibility, however, is a background condition; it is not a recovered ranking of prospective sites or a complete account of the decision that produced the later address.

European research networking was institutionally untidy. Olivier Martin’s historical account places EARN, EUnet, RARE, RIPE and the emerging Internet within overlapping communities whose technical concerns often crossed their formal borders. RARE offered a broad European framework, RIPE concentrated on practical coordination among IP networks, and EARN and EUnet carried different histories and constituencies. Personnel could cooperate across these settings while finance, ownership and organisational responsibility remained divided.

A coordination centre was useful partly because the surrounding field lacked a single institutional home for all the activities requiring sustained attention. Martin warns that lost archives required portions of his account to be reconstructed from memory. That makes his work most useful as context for the larger landscape. It helps explain why a permanent centre could arise from cooperation among several organisations while remaining distinct from each of them.

RIPE-057 would later describe RIPE as the IP activity of RARE and as a collaborative organisation open to European Internet service providers, while also stating that RIPE operated no network of its own. Coordination and ownership occupied different planes from the beginning.

Amsterdam’s bounded advantage lay in the combination of connectivity, institutions and people capable of sustaining technical cooperation. A city with that combination could host an office whose working field extended far beyond its premises. The office’s later importance came from what accumulated around it: employees, sponsorship, records, contracts, contributor relationships and eventually a separate legal identity. The urban setting enabled part of that development, but the institutions and people performed the work.

This is why Amsterdam’s rise is better understood as an institutional natural history than as the simple consequence of geography.

The timing also invites caution about hindsight. Once a city has hosted an organisation for decades, its position can appear inevitable. In 1991, the proposed NCC was a small service whose organisers were still trying to assemble finance, personnel, premises and legal shelter. The significance of its location became clearer only as functions and relationships accumulated. Looking backward from the later association risks assigning the first office all the weight of the mature institution.

The contemporary documents instead reveal a gradual convergence of professional capacity, contractual structure and legal form around a place already in use.

Designing a permanent centre before choosing its home

On 10 June 1991, RIPE-039 set out procedures for establishing a permanent Network Coordination Centre. Its author, R. Blokzijl, organised the undertaking around finance, personnel and site. That division placed location beside money and professional capacity as a component requiring deliberate responsibility. RIPE was moving beyond agreement that an NCC would be useful and into the practical work of making one operable. The paper was an establishment design: it described actions still to be taken, financial goals still to be met and procedures intended to govern later choices. RIPE-039 referred to RIPE-019 for the need for the NCC and RIPE-035 for its activity plan. It also reported that the eighth RIPE meeting had asked RARE to provide the centre’s legal environment. RARE’s Council of Administration accepted that request at its meeting of 16–17 May 1991. The acceptance gave the planned centre an institutional shelter, but RARE lacked the resources to finance it centrally. Legal framework and operating money therefore had to be assembled through different relationships from the start.

That separation defined the prospective organisation more precisely than a single sponsorship label could. RARE could house the centre formally while organisations active within and outside RIPE participated in funding arrangements. RIPE supplied the technical community and activity plan, but it was not described as the sole financier or employer acting independently of RARE. The contemplated NCC occupied an intermediate position: more permanent than a voluntary working arrangement, less self-contained than the association created in 1997, and dependent on cooperation among institutions whose roles overlapped without becoming identical.

The financial design began with a first approximation. RIPE-039 proposed preliminary annual components of ECU 180,000 for staff, ECU 30,000 for equipment, ECU 15,000 plus overheads for office space and ECU 15,000 for travel. The stated total was ECU 240,000 per annum. The figures were presented as financial goals requiring further refinement before operation. They expressed the intended scale and composition of a permanent centre; they were not observed expenditure, settled invoices or a report of money spent during 1991.

Staff represented three quarters of the stated total, making professional capacity the dominant preliminary commitment. Equipment, premises and travel mattered because the office required tools, a physical base and contact across a distributed community. Yet the budget’s centre of gravity was human. The proposed NCC was not chiefly an expensive technical installation. It was an attempt to provide sustained attention to functions that meetings and volunteers could not carry reliably on their own. That priority also gave personnel decisions unusual influence over the character of the new institution.

The corresponding funding goal was ECU 720,000 for an initial three-year period. RIPE-039 proposed inviting organisations active within and outside RIPE to participate in the arrangements required to reach it. The target remained aspirational: it described support to be organised before operation, not finance already secured. Prospective funders were being asked to underwrite continuity, while professional selection and host evaluation were assigned to separate procedures. Money enabled the centre, but financial participation was not made synonymous with choosing every employee or determining the site.

This ordered separation distributed responsibility across three fields. Funders would help make the proposed service financially possible. A mixed personnel board would exercise professional appointment authority. A different mixed board, joined by the manager, would evaluate the place in which the centre could operate. The architecture guarded against treating any single contribution as a complete institutional mandate. Finance supplied resources; appointment supplied staff; site assessment supplied a working environment. Each depended on the others, but each called for a distinct kind of judgement.

Recruitment was designed as an open solicitation. Position announcements were to circulate through the RIPE and RARE Council of Administration distribution lists. A staff-selection board representing RIPE and the RARE Executive Committee would appoint the personnel, with RIPE expected to have a strong voice because close relations between the technical forum and the permanent centre were considered necessary. The open call widened the proposed search, while the board placed appointment responsibility in named institutional hands instead of leaving staffing to informal emergence.

The manager was to be appointed first. Once selected, that person would participate in choosing the remaining employees. This order gave the future professional leader a role in shaping the team and tied subsequent recruitment to operational responsibility. The manager was also expected to participate in the final site choice, linking personnel formation to the assessment of premises and communications. Leadership thus stood between the general establishment plan and its implementation: the manager would inherit neither a complete staff nor an already fixed institutional home under the procedure as written.

Site selection followed its own proposed sequence. Detailed requirements were to be prepared, an invitation for site proposals was to circulate through the RIPE and RARE Council of Administration distribution lists, and a mixed board would make the final choice. That board was to include representatives of RIPE, the RARE Research and Engineering Council and the appointed NCC manager. The design placed technical-community representatives, a RARE research-and-engineering body and professional leadership around the same evaluation without merging their roles.

The personnel and site boards were intentionally different. Staff appointments belonged to RIPE and the RARE Executive Committee; location responsibility belonged to RIPE, the RARE Research and Engineering Council and the manager. Executive representation helped select the people who would operate the centre. Research-and-engineering representation joined the manager in assessing the environment required for their work. The order also meant that detailed site requirements could be considered with input from the person responsible for running the organisation, giving practical needs a route into host evaluation.

A site under this design was not a ceremonial capital. It was a working package of premises, communications and conditions suitable for the centre’s staff and equipment. Preparing requirements before inviting proposals would have allowed the organisers to convert an abstract preference into assessable needs. Circulating the invitation through RIPE and RARE channels would have exposed the opportunity to their institutional communities. The proposed board would then have carried responsibility for reconciling those needs with whatever offers arrived.

Every step remained prospective in RIPE-039, but together they show a carefully segmented conception of institutional choice.

The architecture is especially revealing when read against the centre’s limited scale. ECU 240,000 per annum and a three-year objective of ECU 720,000 described a small professional service being assembled for continuity, not the headquarters of a mature corporation. Early decisions could nevertheless become unusually durable because the first team embodied much of the organisation’s knowledge. A manager involved in recruiting colleagues and assessing the site would connect leadership, workforce and place during the same formative period. Modest initial scale magnified the influence of those choices rather than making them inconsequential.

RIPE-039 therefore records more than a budget and an invitation. It assigns prospective responsibility among supporters, professional selectors and host evaluators while ordering the choices needed to create a permanent centre. Finance had to be assembled, a manager had to be chosen, further staff had to follow, requirements had to be written, proposals had to be invited and a site had to be assessed. The paper captures an organisation under design, with every amount preliminary or aspirational and every appointment or selection procedure still proposed.

The Amsterdam office enters the contemporary record

Less than eleven months later, the historical view changes abruptly. RIPE-057, published on 5 May 1992 by D. Karrenberg, gave the RIPE NCC’s contact address as Kruislaan 409, NL-1098 SJ Amsterdam. It supplied telephone, fax and electronic-mail contacts and reported three permanent staff members. RIPE-161 would later state that NCC operations began in April 1992. The address and staffing statement provide a firm contemporary anchor: a permanent coordination centre was operating from Amsterdam by the spring of 1992. The office sat within RARE’s formal framework. RIPE-057 identified RARE national members and EARN as the funders of the first operating year. Framework and finance were connected but remained institutionally distinct. RARE supplied the formal setting; its national members and EARN provided the reported first-year support; RIPE supplied the collaborative technical field the NCC had been established to assist. The arrangement gave the three-person centre both organisational shelter and a financial base without collapsing RARE, RIPE, EARN or the participating organisations into one body.

RIPE-057 placed that small office against a much larger coordination environment. It reported more than 60 organisations participating in RIPE’s work, more than 170,000 computers throughout Europe reachable through networks coordinated by RIPE, and a worldwide estimate above 800,000 reachable systems. These were three distinct institutional measures: participating organisations, European reachable computers and worldwide reachable computers. The figures did not describe employees, association members or completed service transactions.

Their connection lies in scale: a three-person centre was being presented as permanent capacity within an expanding, distributed technical field. The contemporary notice described RIPE as a collaborative organisation open to European Internet service providers and said it had functioned since 1989. It also characterised RIPE as the IP activity of RARE and stated that RIPE operated no network of its own. That combination clarifies why an NCC could become important without owning the systems in question.

The centre supported coordination among separately operated networks, making records, contacts and administrative functions available within a cooperative setting whose reach exceeded any one sponsor.

Three permanent employees marked a decisive institutional change. RIPE-057 described the NCC as supporting RIPE activities that volunteers in participating organisations could not effectively perform. Permanent staffing meant that responsibility could be attached to an office instead of being recreated at meetings or left entirely to voluntary effort. The document gives no schedule of individual work or account of particular requests. Its contribution is more fundamental: it identifies a professional team, a physical address and a declared set of responsibilities at a specific date. The administrative portion of the declared remit was broad.

RIPE-057 described a network-management database covering IP networks, DNS domains, routing policies and contacts; delegated Internet-registry activity and distribution of IP network numbers; coordinated statistics gathering; DNS coordination; and planned graphical maps of IP networks. These functions gathered several kinds of information required for cooperation across independently managed systems. Their appearance in one institutional description shows the surface RIPE associated with the NCC, while the document’s mixture of offerings and plans preserves the provisional status of some elements.

Publication and access formed another part of the same surface. The NCC described a network-operations software repository, the RIPE document store and a menu-driven interactive information service through which technical documents, Internet RFCs, software and institutional information could be reached. Circulation was therefore placed beside registry and database work in the centre’s announced role. RIPE-057 reported services offered or planned and named their access channels; it did not measure demand, user conduct, uptake, accuracy, availability, continuity or outcomes. The technical channels convey the period character of the office.

The database was described as accessible through an RFC 954 whois server and through the interactive information service. The document store offered RIPE documents and RFCs through anonymous FTP, gopher and WAIS. The menu-driven service accepted telnet and two X.25 routes and could be used to browse the document store, read documents and send them by electronic mail. World Wide Web and OSI-FTAM access appeared as future additions.

Read together, the service description and scale figures show why permanent coordination could appear necessary. More than 60 participating organisations represented a collaborative environment. More than 170,000 reachable European computers represented an estimate of technical reach across the region. The worldwide figure above 800,000 supplied a broader network context. None was a measure of the NCC’s workload. They instead framed the environment in which a small staff was assigned responsibilities spanning registration, databases, domains, statistics, software and documentary access.

The juxtaposition is striking: three permanent staff stood behind a publicly declared remit extending across several administrative and publication functions. That contrast should not be turned into an imagined day inside Kruislaan 409. The notice records institutional scale at two different levels—the size of the team and the breadth of the environment around it—without showing how employees divided their time or how outside organisations engaged with individual services. Its value lies precisely in the disciplined connection among address, sponsorship, staffing and declared purpose.

By May 1992, Amsterdam held a dated office, a professional team, a formal framework, reported first-year funding and an identifiable coordination surface. These facts mark the passage from institutional proposal to documented operation. They also show that Amsterdam’s practical importance preceded independent incorporation. The city first entered the record as the location of staff and contact channels. Its later role as the legal seat of an association would be a separate development layered onto that operating presence.

How an address gathers institutional weight

An operating address begins as geography. It acquires institutional weight as people, responsibilities and relationships gather around it. For the RIPE NCC, personnel were the first source of that weight. RIPE-039 had assigned ECU 180,000 of a preliminary ECU 240,000 annual estimate to staff, and RIPE-057 soon reported three permanent employees. A centre designed primarily around professional capacity would carry much of its knowledge through people familiar with its technical remit, participating institutions and documentary responsibilities. Records supplied a second source.

A network-management database, routing policies, domain information, contacts, technical documents and statistical material all possess value beyond a single exchange. Their significance lies partly in continuity of custody and interpretation. A stable office gives recurring responsibility a durable location, allowing the same institutional centre to remain associated with the materials even as individual entries, documents and contacts change. The sources identify the records and declared functions; they leave the detailed working practices inside the office unobserved.

The centre also occupied a widening web of institutional contacts. RIPE connected technical participants; RARE supplied a formal setting; EARN joined RARE national members in first-year funding. Participating organisations formed the collaborative environment around the NCC. These relationships made the Amsterdam office a focal point among institutions that retained their own identities and responsibilities. The centre’s value arose in part from crossing those boundaries without absorbing the networks, sponsors or organisations into a single owner. Once an operation exists, continuation and relocation present different practical tasks.

Continuing at an established address allows the same staff location, custody of records, communications details, contacts and sponsor relationships to remain aligned. Relocation would require those elements to be moved, reassigned, communicated or renegotiated as appropriate. Even where the functions themselves are portable, the transition has to be managed. Institutional continuity can therefore acquire practical value after operation begins, simply because an existing configuration already coordinates several dependencies at once.

The same logic applies to declared services and agreements. When registry administration, databases, DNS coordination, statistics, software and document access are publicly attached to a permanent centre, later planning begins from an organisation with an identifiable remit. When sponsorship and contributor relationships acquire formal terms, those relationships become part of the inherited arrangement. Preserving the centre in place can preserve the alignment among office, people, records, counterparties and institutional expectations while governance changes around it. This account of persistence remains qualitative.

The sources price neither relocation nor any savings produced by staying. They supply no comparison of transition costs, staffing disruption or alternative premises. The analytical point is narrower: continuity can become valuable because changing location entails work across several institutional attachments, while remaining in place carries those attachments forward. That explains how an operating address may become sticky after 1992 without explaining the earlier judgement that first placed the office in Amsterdam.

Time reinforced the effect. RARE’s successor, TERENA, later provided the NCC service to contributors under formal service agreements. Sponsorship evolved, but the operation remained connected to the same institutional lineage. Service agreements gave a contractual perimeter to relationships that had begun within a broader field of technical cooperation. By 1997, the separation debate concerned named contributors, staff rights, reserves, management authority and TERENA’s residual liabilities. Those questions belonged to an established organisation, not to the preliminary centre imagined six years earlier.

Contracts and permanent staff alter the cost of organisational change even where no monetary estimate exists. Contracts identify parties and obligations. Employees possess rights and expectations tied to their employer and working environment. Records carry custodial responsibilities. Public contacts create a stable reference point. Each attachment makes continuity a practical option with known institutional coordinates. A successor organisation that preserved them could focus on changing governance and legal identity; one that also changed place would add another transition to an already demanding reorganisation.

Amsterdam’s relevance consequently deepened through accumulation. The office supplied a place; permanent staff supplied continuing professional capacity; records and contacts supplied institutional memory; sponsorship supplied a framework; service agreements supplied a contractual boundary. None depended on the city alone, yet each became attached to an operation already located there. Path dependence in this setting means that early placement changed the context of later decisions, making the inherited configuration one of the facts any successor design had to address.

The city’s networking ecology and the NCC’s accumulated institutional position should remain analytically separate. Connectivity, research institutions and technical expertise explain why Amsterdam was a plausible setting for the centre. The office reported in 1992 explains how Amsterdam later became an inherited organisational fact. The former concerns conditions around the city; the latter concerns staff, records, contacts and agreements gathered after operations began. Combining them too quickly would allow the durability of the later organisation to stand in for the unknown comparative reasoning of the earlier site choice.

By the mid-1990s, that inherited fact was no longer attached to an experimental office of three people. Contributor finance, formal agreements and anticipated staff growth created a problem of ownership and accountability. Continuity had carried the NCC far enough that its position inside another association required reconsideration. The next institutional question was not simply where the centre should work, but who should hold authority over a larger service, how contributors should participate and which legal body should carry its employees, liabilities and reserves.

Scale forces a new institutional container

The review began in September 1996, when contributors and TERENA agreed to investigate separating the NCC from TERENA. Paul Ridley, Karel Vietsch and Wim Vink formed the three-person committee charged with developing a structure. Their work aimed at a new organisation beginning on 1 January 1998. The proposed timetable required legal, financial and managerial questions to be settled around an operation that already had staff, service agreements, contributors and an established city.

RIPE-161, dated 8 August 1997 and authored by Ridley and Vietsch, described the scale confronting the committee. The NCC’s operating-expense budget for 1997 was ECU 2 million, and permanent staff were expected to exceed 30 by year end. The expense figure was a budget and the headcount a forecast. Their status matters, but so does their institutional meaning: the designers were preparing governance for an organisation far larger than the preliminary centre contemplated in 1991.

The contrast traces expansion without turning unlike figures into a direct performance measure. RIPE-039 had estimated ECU 240,000 per annum, including ECU 180,000 for staff, and had sought ECU 720,000 for three years. Six years later, the separation inquiry addressed an ECU 2 million expense budget and an expected workforce above 30. The later figures increased the number and complexity of interests that the new legal structure would need to organise. Governance had to catch up with operational scale.

Separation concerned more than corporate tidiness. TERENA provided the NCC service to contributors under formal service agreements. Transferring that service into an independent organisation implicated the contributors paying for it, TERENA as the association carrying liabilities, employees whose contractual and tax positions could be affected, and managers expected to preserve professional administration through the change. Each group approached the transition from a different position even when their interests in continuity overlapped.

Contributor criteria included continuity of service, neutrality, not-for-profit form, meaningful influence over governance, openness to potential Internet service providers, suitable tax treatment, resistance to unwelcome control and transfer of the surplus held by TERENA. These concerns joined operating continuity to institutional ownership. Contributors wanted the service preserved while gaining a clearer place inside the organisation responsible for it. The new structure had to make that influence real without converting every professional judgement into a general membership decision.

TERENA’s concerns centred on residual liabilities and an initial role in the successor. Staff interests included contractual and tax rights and sufficient reserves. Those priorities show why a separation could not be treated as a simple renaming. Liabilities had to remain attributable, employees required protection, accumulated funds needed a destination and transitional authority had to be defined. The committee was designing a successor capable of inheriting an operation without leaving its obligations behind.

Management authority required careful placement. An independent body needed member and board control, yet the service also required a professional domain insulated from intervention in individual cases. RIPE-161 contemplated a general assembly and executive board as principal bodies, with a treasury committee and management team in supporting roles. The proposal placed the greater weight of power in the general assembly while preserving delegation for the work performed under service agreements.

The committee received contributor input by email and at the Dublin RIPE meeting. Ridley and Vietsch described an overwhelming lack of criticism and inferred consensus from it. Their account records how the authors interpreted the feedback available to the process. Amendments followed the consultation, showing that contributor comment influenced the revised design. The statement remains an institutional assessment of response, not an independently measured vote among every organisation touched by registry coordination.

After considering possible structures, the committee obtained advice from Coopers & Lybrand on legal, financial and tax questions. A Dutch vereniging was preferred for the combined advantages the designers attributed to tax suitability, democratic character and resistance to unwelcome takeovers. These reasons addressed the needs of the proposed successor: a not-for-profit association could place contributors inside the legal body, distribute authority through member government and reduce vulnerability to control inconsistent with the organisation’s purposes.

The association form offered an answer to several stakeholder questions at once. A general assembly could exercise central powers. An executive board could represent and supervise the corporate body. A treasury committee could scrutinise finance. Management could administer service responsibilities under delegated authority. Membership could connect the organisations funding and receiving services to the institution that carried the employees, reserves and obligations. The design sought continuity of operation through a redistribution of formal power.

RIPE-161 remained prospective. It proposed membership tests, governing bodies, voting procedures, management protections and transitional arrangements for the intended successor. A current service agreement, payment of fees and conduct consistent with the association’s interests would qualify an applicant for membership, subject to the proposed rules. Employees were excluded from general-assembly membership, and candidate members initially lacked votes. These provisions defined a contributor association rather than an undifferentiated technical community.

The proposed executive board comprised three to five natural persons elected by the general assembly for staggered three-year terms. Employees were barred from serving, while the general manager would attend board meetings without a vote. Two board members acting together could represent the organisation. The board could appoint the general manager and delegate authority, but the proposal withheld power to influence management’s operational decisions concerning individual Internet service providers.

Discussion and voting were also specified. An online forum and four-week discussion period were intended to support consensus where possible, followed by authenticated network voting when consensus remained unclear. Votes were weighted one, two or three according to registry size. Transitional provisions contemplated selection of the first board by the contributor committee, a TERENA-named board member for two years and a first-year veto associated with TERENA’s remaining financial liabilities.

The proposal reveals an organisation trying to change legal container without abandoning the capacities built under RARE and TERENA. Contributors sought influence; TERENA sought protection against residual exposure; staff sought continuity of rights; management required an operating domain; the successor needed reserves and a workable tax position. Amsterdam entered this inquiry as the city of the established operation. Dutch law supplied the legal form through which its contributor relationships and internal authority could be reorganised.

From operating service to Dutch association

Formal incorporation followed before the proposed 1 January 1998 start. RIPE-176 records that the Dutch articles were deposited with the Amsterdam Chamber of Commerce on 12 November 1997. Deputy civil-law notary F.J.J. de Jonge and W.K. Vietsch appeared for the act, with Vietsch acting as TERENA’s secretary general under written proxy. The deed converted the planned separation into a legal association and identified the instrument through which its powers would be organised.

The legal transition had begun with the September 1996 inquiry and moved through the three-person committee, Coopers & Lybrand’s advice, stakeholder criteria and RIPE-161’s proposed structure. Those stages explained the problem and the designers’ intended solution. RIPE-176 occupied a different level of authority. The proposal recorded reasoning and anticipated arrangements; the deposited deed established the association, its seat, membership rules, governing bodies, delegation and documentary obligations. Reading them together requires preserving that hierarchy.

The hierarchy also existed within the legal record itself. RIPE-176 supplied an English translation for accessibility while stating that the Dutch text governs wherever the two differ. Amsterdam was named as the association’s official seat, linking the legal person to the Netherlands and to the controlling instrument deposited there. The rule of textual priority matters because explanatory or translated language cannot displace the deed that carries legal authority. Accessibility and legal control were deliberately separated.

The association took the name Réseaux IP Européens Network Coordination Centre, abbreviated RIPE NCC. Its objectives encompassed registration, coordination and administration, together with new activities, and profit-making was excluded as an objective. The former TERENA service now had legal personality, formal purposes and a seat of its own. Amsterdam consequently carried two meanings: the city of the documented operating office and the legal home of the incorporated association.

Membership began from the standard service agreement. Subject to the deed’s provisions concerning legal persons and employees, entry into the qualifying agreement led to candidate membership. A six-month candidate period preceded full voting status and could be extended, or accompanied by restrictions on service use, under the stated conditions. The contractual gateway established who could enter the association. It created a defined internal constituency connected to service arrangements rather than treating everyone affected by Internet coordination as a member.

Enterprise, small, medium, large and supranational categories reflected contributor classifications in the standard service agreement. Annual contributions followed a charging scheme determined by the general meeting. Category, payment, membership and voting were linked while retaining separate legal functions. A candidate could belong without voting. A category could affect contribution and voting weight. The service agreement opened the membership route. These distinctions made the association’s internal constituency legible and tied financial participation to formal rules.

Authority rose through the general meeting. It held powers left unassigned elsewhere and had to convene annually. It appointed board members and acted on central questions of corporate governance. Members holding at least two percent of the votes had a route for proposing agenda items; those holding at least one-tenth could require an extraordinary meeting through the prescribed procedure. These thresholds gave minority positions formal means to place matters before the membership body or compel review.

Voting translated the categories into differentiated internal power. Candidate members had no vote. Enterprise and small members received one vote, medium members two and large members three, while supranational members followed the formula stated in the articles. Proxy use and concentration among related entities were limited. The deed thereby converted a service-agreement constituency into a countable system of corporate decision-making, with membership status and category determining the weight attached to participation.

The board formed the next layer. It consisted of three to five natural persons appointed through the general meeting’s nomination process, and employees of the association were ineligible. Two board members acting jointly represented the RIPE NCC. Excluding employees from board membership distinguished corporate supervision from the workforce administering the service. Appointment by the general meeting connected the board’s position to the members while joint representation prevented one director from acting alone for the association.

Professional management held a defined domain beneath that corporate layer. The board was required to delegate operational decisions under standard service agreements to management. It also had to establish an arbitration procedure for disputes after consulting members. The design combined member authority, board representation and professional administration through delegation. General powers belonged to the membership body, corporate representation belonged to the board, and individual service-agreement decisions were assigned to management within the legal structure.

Registers, minutes and accounts made the hierarchy reviewable. The board had to maintain a membership register recording names, addresses, admission dates, categories and voting entitlements. Board proceedings generated minutes, which had to be adopted and sent to members within two weeks of adoption. Financial records had to be sufficient to establish rights and obligations; annual reporting, financial statements and an accountant’s certificate were required; specified records were subject to ten-year preservation.

These duties created institutional memory in legal form. Membership status could be checked against a register. Board action could be examined through minutes. Financial position and obligations could be followed through accounts and annual reporting. The preservation period carried evidence beyond the tenure of particular directors or managers. Authority was therefore accompanied by records through which later members could review who belonged, how decisions had been taken and what financial responsibilities the association carried.

The executed structure did not simply reproduce every transitional feature discussed in RIPE-161. The proposal’s contemplated TERENA role, initial-board arrangements and first-year veto explain how the designers approached the handover and remaining liabilities. The deed controls the association’s established rules. Keeping those statuses separate prevents an intended transitional device from being mistaken for a permanent legal provision and keeps later legal precision from erasing the debates that preceded it.

Incorporation consolidated relationships assembled over several years. Staff belonged to an employer with its own legal identity. Contributors could enter a membership system through service agreements. The general meeting and board could act for the association, management could exercise delegated operational authority, and registers, minutes and accounts could preserve the resulting acts. The new organisation inherited the centre’s practical history while changing who held formal rights and obligations around it.

Place now carried legal consequences alongside practical ones. Moving an office would affect people, records, communications and contractual relationships. Altering an official seat would also engage the governing instrument and jurisdiction of the legal person. By fixing the seat in Amsterdam, the deed added a legal commitment to a city already associated with the NCC’s operation. The location endured through continuity, but the institution located there had changed from a RARE and TERENA service into a contributor-based association.

Testing the path-dependence explanation

Path dependence offers a strong account of persistence, but location is only one possible cause within the historical chain. Personnel continuity may have preserved knowledge regardless of city. RARE and TERENA sponsorship may have supplied the institutional bridge that mattered most. Contributor finance may have sustained the centre as its responsibilities expanded. European network growth may have increased demand for coordination, while first-mover effects may have favoured the earliest established office. The later preference for a Dutch vereniging added a separate legal reason for retaining a Dutch institutional home.

These explanations can reinforce one another. An established team may be most valuable when it remains close to accumulated records and contacts. Sponsorship may support continuity while contributors provide finance. Network growth may increase the importance of functions already assigned to the centre. Dutch association law may make an existing Dutch operation easier to reorganise without requiring the city itself to be uniquely advantageous. Path dependence describes this layering, but a useful causal account must still ask which attachments did the most work.

Contemporaneous site requirements and submitted proposals would sharpen that inquiry. Comparable measures of communications, staffing pools, premises, costs and legal conditions would show what the organisers expected from a host and how available options addressed those needs. Deliberations, scoring, reasons or a recorded decision would reveal which considerations selectors actually treated as decisive. Such evidence could separate an Amsterdam-specific advantage from a general preference for any location capable of supporting a small professional centre.

Later evidence could test persistence from the other direction. Records of proposed relocation, continuity planning, staff retention, transfer of records, changes in sponsorship or assessments of alternative jurisdictions would reveal which dependencies constrained subsequent choices. A measured account of transition work could distinguish the value of keeping the office in place from the value of retaining the same personnel, contracts or institutional sponsor. Evidence of a later move accomplished without disruption would weaken some continuity claims; evidence of location-specific obstacles would strengthen them.

The 1997 legal-form choice is especially important as a competing explanation. Coopers & Lybrand’s advice and the committee’s stated preference for a Dutch vereniging concerned tax suitability, democratic character and resistance to unwelcome takeover. Those reasons support the jurisdictional form selected for the successor, not the earlier placement of the office. If comparable legal analyses from other jurisdictions existed, they could show how strongly Dutch law shaped the outcome. Without them, the legal choice explains consolidation in 1997 more clearly than origin in 1991.

A falsifiable account therefore treats Amsterdam as one factor among several. The city’s networking ecology supplied plausible start-up advantages; the operating office accumulated attachments; RARE and TERENA supplied organisational continuity; contributors financed a growing service; and Dutch law offered a suitable association form. The claim is not that geography caused every later development. It is that early location became embedded in a configuration whose components increasingly supported continuation. Better comparative and transitional evidence could revise the weight assigned to each component.

Where the historical explanation stops

The full boundary can be stated compactly. The material records a 1991 proposal for detailed site requirements, open solicitation and a mixed selection board, followed by an Amsterdam office in 1992; it contains no completed requirements, submitted site proposals, candidate list, comparative measures, deliberations, vote or final reasons, so Amsterdam appears only as the later documented operating location and never as a candidate or winner in an executed contest.

RIPE-057’s three permanent staff, more than 60 participating organisations, more than 170,000 reachable European computers and worldwide estimate above 800,000 systems are different units, not interchangeable measures of membership, workload, use or representation. The office, 1997 legal seat, European service field and service-agreement membership constituency likewise describe different institutional surfaces. Contributor feedback at Dublin and by email has no shared denominator connecting contributors with every operator, resource holder, government or organisation affected by registry coordination.

The association’s internal democracy vested rights in members under its rules; it conferred neither territorial authority nor Europe-wide authorization. Finally, the available record supports a qualitative persistence account but cannot isolate the effect of location from personnel, sponsorship, contributor finance, network growth, contractual relationships, first-mover effects or the later Dutch legal-form decision.

The archival responsibility created by permanence

The 1997 articles brought a new precision to the RIPE NCC. From 12 November onward, a reader could identify the controlling Dutch text, the association’s objectives and Amsterdam seat, the contractual route into membership, the distribution of authority among the general meeting, board and management, and the records required to make those relationships reviewable. Registers documented who belonged and with what voting entitlement. Minutes, annual reports, financial statements and preservation duties gave the incorporated institution a formal memory.

Institutions often outgrow the circumstances of their founding. An address used by a three-person centre can gather employees, records, sponsors, contributors and contracts; an inherited operating location can later be written into the governing instrument of a legal person. Legal precision in 1997 cannot retroactively supply contemporaneous comparative reasons for the earlier location. It does, however, show what archival accountability looks like once authority, membership and money have become substantial enough to demand durable rules and reviewable records.

Amsterdam’s permanence is historically intelligible: a significant networking ecology preceded the NCC, the 1991 plan treated people, finance and place as linked establishment problems, a staffed office was documented in the city in May 1992, and contributor agreements and organisational scale later raised the value of continuity before Dutch incorporation fixed Amsterdam as the official seat. The archival consequence follows from that history.

When a practical address may harden into a long-lived centre of institutional authority, the requirements, alternatives and reasons behind the initial choice deserve preservation alongside later records of membership, voting, delegation and finance.