How this differs from previous attempts

Portable, learner-centred records of a child’s work aren’t a new idea — the question is whether the mechanism behind them avoids the failure mode that has sunk or sidelined previous attempts.

inBloom: a centralised store parents wouldn’t trust

inBloom was a US nonprofit, launched in 2013 with roughly $100 million, mostly from the Bill & Melinda Gates Foundation and the Carnegie Corporation. It aimed to give teachers a single, centralised store of student data to personalise instruction. It began with nine participating states, and was wound down in April 2014, roughly a year after launch.

The fullest account of what happened is Data & Society’s 2017 “autopsy,” The Legacy of inBloom, based on interviews with 18 stakeholders. Two articles covering it pick out different parts of its findings:

• Education Week, 2017 focuses on the report’s disagreement over blame. Supporters blamed unrealistic ambitions, poor implementation and a misreading of the fragmented US school system; parent activists rejected that framing, arguing that treating opposition as irrational assumed inBloom was worthy of trust in the first place. The fallout included more than 400 state bills on student-data privacy.
• THE Journal, 2017 focuses on the report’s other findings: bad timing, launched amid the NSA revelations and high-profile breaches; no compelling story about benefits for teaching and learning; and a coalition that shrank from nine states to three within months.

The accounts disagree about how much was avoidable — inBloom’s own supporters cite poor communication and delivery, its opponents cite a reasonable refusal to trust — but they agree that the technical design was not what sank it. What parents and states rejected was a single organisation holding sensitive data on children at scale, and asking them to trust its safeguards. Better communication might have delayed that reaction, but it doesn’t change the structure that provoked it.

Satchel’s structural difference

Satchel’s structural difference is that Solid never puts the data in one organisation’s central database in the first place. Each child’s data lives in their own Pod, and sharing is a resource-level, revocable grant written into that Pod, not a policy promise made by an intermediary. That’s a difference in where the data physically lives and who can grant access to it, which is what can make it survive a vendor changing hands, going out of business, or being asked to share data in a way a family didn’t expect — in a way a centralised store structurally can’t.

The control isn’t blanket “family” ownership, though. The child and their guardian are separate parties, a guardian is a recorded relationship rather than something inferred from being a parent, and a share only takes effect when both agree, with either able to revoke it. The full model is in Satchel: adding Solid storage to Scratch.

Decentralised storage removes the single-point-of-trust failure mode, but it doesn’t by itself solve, for example, who holds a young child’s Pod credentials or what happens if a family loses them.