Unseen Labor

Sustaining our digital infrastructure is a new topic for many, and the challenges are not well understood. In this report, Nadia Eghbal unpacks the unique challenges facing digital infrastructure, and how we might work together to address them. pdf

The Sloan and Ford Foundations would like to fund a set of research projects to further study these dynamics, with an eye toward better understanding the economics, maintenance and sustainability of digital infrastructure. post

There is a misperception that this is a “solved problem.” The pervasive belief, even among stakeholders such as software companies, that open source is well-funded, makes it harder to generate support. Some infrastructure projects operate sustainably, either because they have a working business model or sponsorship, or because their required upkeep is limited. An unfamiliar audience will also associate open source with enterprise companies like Red Hat or Docker and assume the problem has been solved. However, these situations are the outliers, not the rule.

There is a lack of cultural understanding and awareness about the problem. Outside of the open source community, nearly everybody remains unaware of infrastructure’s funding issues, and the topic is perceived to be dry and technical. Developers needing support tend to have a highly technical focus and aren’t comfortable advocating for the business side of their work. Taken together, there is no momentum to change a broken situation.

Digital infrastructure is rooted in open source, whose volunteer culture discourages talk of money. Although this attitude has made open source what it is today, it also makes it difficult for developers to openly discuss their needs without feeling guilty or worrying about not being perceived as a team player. Open source’s highly distributed and democratic nature also makes it difficult to coordinate and sustain institutional actors who could act as advocates for their needs.

Digital infrastructure is highly distributed, compared to physical infrastructure. Unlike planning the construction of a bridge, it’s not always clear which projects are useful until after they have already taken off. They cannot be planned beforehand by a centralized entity. At the other end of the lifecycle, some projects are meant to decline as other, better solutions take their place. Digital infrastructure is distributed across hundreds of projects, large and small, built by individuals, groups and companies; it would be a behemoth task to catalog them all.