Pattern no. 10 — “Those who prosper maintain”: Institutionalisation · Maintenance
Translated from the French, which remains the reference text. This translation is under review.
Pattern n°10 · Metabolism (primary), Stratification, Thirdness · tensions T11, T7, T8 · Legitimacy · volets I.3, II.3 · ↑ navigation
Schema F10.1 · Context of pattern no. 10
Tension(s) at work: Coopetition; public, private and citizen; long term and agility — and, implicitly, the challenge of sustainability: crossing the valley of death, then never falling back into it. This pattern answers the question bootstrapping leaves open on the day it dies out: the network runs, the igniter has withdrawn — who pays for upkeep, forever? One particularity sets it apart from all the other fiches: no ecosystem in the corpus has fully resolved it. It is the pattern in search of its canonical embodiment — and, as such, the first item on the agenda of Euro-Indian convergence.
Context: The commons has passed critical mass. It has become what every successful infrastructure commons becomes: invisible. Yet invisible infrastructure degrades in silence — dependencies to update, vulnerabilities to fix, technical debt to repay, a standard to evolve — and this work has three properties fatal to its funding: it is permanent (never “finished”), thankless (no ribbon to cut), and non-excludable (its benefit cannot be reserved for those who pay, since the commons is open by construction). Meanwhile, the value passing over the rails grows — and those who capture it have, individually, no interest in paying for what they benefit from collectively.
Forces at play: On one side, the rationality of the free rider, at its point of incandescence: unlike bootstrapping, where waiting has a cost (one misses the network), maintenance free-riding costs nothing — the commons stays open, maintained by others, until the day it no longer is. On the other, the visibility asymmetry of funding: funders — public and philanthropic alike — finance novelty, and never or almost never upkeep; a maintainer who asks for funds to “keep it running” goes unheard, while the same maintainer promising a “disruptive innovation” is eligible. The median compromise — one-off rescues after an incident — is the real regime of today's world: reactive life support, triggered by crisis, that subsides as soon as attention does.
The pattern: Institute a regular flow of resources, indexed on the prosperity the rails generate — not on the pity they inspire nor on the novelty they promise. The corpus and its surroundings reveal four forms, which can be combined: the levy on the rails — an at-cost micro-levy on the transactions the commons makes possible, the most congruent form since the funding flow mechanically follows the flow of value; the congruence contribution — members' dues proportionate to size and to the benefit drawn (the model of the Eclipse working groups or of the foundations of industrial open source), a direct application of the principle Elinor Ostrom identified as vital: the correspondence between what each draws from the commons and what each brings to it; the pooled maintenance instrument — a fund dedicated explicitly and exclusively to the upkeep of what exists, sanctuarised against the temptation of innovation (the German Sovereign Tech Fund is its prototype, the EOS Maintenance Instrument its scaling up to European level); the contribution obligation — the emerging regulatory route: public procurement clauses imposing upstream contribution, the responsibility of integrators for their dependencies (in the spirit of the Cyber Resilience Act). Two cross-cutting rules: fund people, not projects — maintenance is work, and its unit of account is the maintainer, not the deliverable; and sanctuarise — any maintenance fund that is not statutorily protected ends up diverted towards novelty, for novelty always has better advocates.
Operator(s) mobilised: Primary: metabolism — this pattern is its constitutive rule. Support: thirdness — it is the guardian (fiche n°9) that collects and administers the flow, and congruence requires an arbiter to assess who prospers by how much; conversely, this fiche pays the debt contracted by n°9: the guardian's four locks hold only if they are funded. Support: stratification — funding can be indexed on the perimeter of the commons only if that perimeter is clear-cut (fiche n°4); commons with blurred boundaries are unfundable by construction.
The invariant under test: The pattern's own risk fits in three words: who pays commands. A metabolism resting on the largest beneficiaries insensibly hands them the key to governance — pay-to-play is capture through the funding window. And the levy itself casts its shadow: the guardian that levies on the rails can become the rentier it replaced — ONDC's founding promise (“the not-for-profit status guarantees that the infrastructure will not tax sellers abusively”) says exactly what is at stake. Associated criterion of legitimacy, twofold: contribution buys permanence, never sovereignty — strictly decoupling financial weight from governance weight, through the locks of fiche n°9; and proven cost price — public accounts, transparent pricing, surpluses reinvested in the commons or returned. A legitimate metabolism is recognised by the fact that one can verify that it nourishes without fattening.
Schema F10.2 · Those who prosper maintain
Cases of embodiment: Indian: NPCI is the most accomplished embodiment — and the most instructive through its limits. A utility held by a banking consortium under not-for-profit status, it is funded by the very actors who prosper on its rails: congruence by construction. But the case also reveals that the pattern remains unresolved even at its best: the imposed free-of-charge status of UPI transactions for merchants — politically untouchable — shifts the metabolism onto public subsidy, whose permanence is debated in India itself; one of the most successful commons in the world lives on openly accepted State life support. Beckn, maintained by the FIDE foundation on philanthropic resources, illustrates the earlier stage: the protocol that is spreading worldwide rests on a metabolism of donors — generous, and precarious. European: the German Sovereign Tech Fund is the conceptual breakthrough of the decade — public money dedicated by mandate to the maintenance of critical open source components, with a statutory prohibition on confusing it with innovation funding; the EOS Maintenance Instrument sets out to make it a permanent European capability. The dues of the Eclipse working groups (the industrial players of the Data Spaces funding the connector that equips them) provide the congruent form; the NGI/NLnet cascade funding, admirable for bootstrapping, shows by contrast the limit of the “per project” approach applied to upkeep.
Counter-example: OpenSSL, before April 2014: the library that encrypts world commerce — the banks, the States, the web giants —, maintained by a handful of volunteers and one employee, funded by a few thousand dollars in annual donations. Then Heartbleed: a vulnerability in this under-maintained code exposes half the Internet, and within a few weeks the Core Infrastructure Initiative is born, endowed with millions by the very companies that had never paid a cent. Seven years later, the film replays identically: Log4j, a logging component present in almost the entire global Java estate, maintained by volunteers — Log4Shell, planetary panic, emergency funds. The theorem these two crises demonstrate: crisis is maintenance's only lobbyist. This pattern exists to take that monopoly away from it — to replace rescue with congruence, haemorrhage with metabolism.
Signs of misapplication: The commons' funding is renewed annually, on criteria of novelty. Maintainers disguise upkeep as innovation to remain eligible (feature-washing, the symptom of a sick funding window). Core maintainers are employed by a single actor — the metabolism is outsourced to a particular interest. The commons' upkeep budget bears no relation to the value passing through it (the OpenSSL ratio: thousands of dollars for trillions). A “sustainability” fund finances prototypes. The levy exists but the guardian's accounts are opaque. Major contributors sit in proportion to their payments. The commons' funding plan was thought up after its launch. Two symptoms warrant an alert; three, a diagnosis.
Related patterns: Upstream: “Paying for the first tank” (n°3 — the bootstrapping of which this pattern is the necessary relay, the exit of the one being the entry of the other) and “The guardian who does not own” (n°9 — the legitimate collector of the flow). Downstream: “Turning the hourglass over” (n°7 — the evolution of the standard presupposes a permanently funded engineering capacity) and “The armour of the common” (n°11 — a starving commons cannot defend itself: metabolism is the first line of security).