A new .self TLD promises everyone a free subdomain. The thread found the holes fast

What HCCF's .self proposal actually promises

A group calling itself HCCF published a vision for a human-centered top-level domain — .self, pitched as infrastructure for self-hosting and "reclaiming our digital selves." The core promises, pulled from the proposal by a reader, are deliberately radical:

"Everyone entitled to a subdomain at no cost … No parking, squatting, or reselling." — a reader quoting the proposal's core-features section

One person, one subdomain, free, for life — framed as an entry point to digital sovereignty, the kind of thing a school could teach kids to run on a cheap VPS. There's an obvious irony the thread didn't miss: the manifesto page itself buckled under the Hacker News traffic, throwing three different error messages on reload. "I guess it's self-hosted on something underpowered," one reader observed — a self-hosting pitch that couldn't survive its own front page.

The three questions the proposal can't yet answer

The skepticism wasn't snark. It clustered on three structural problems, each fatal on its own.

The first is money. The same reader who quoted the features asked the question the proposal skips:

"How are you going to pay for the (substantial) cost of running a TLD without registration fee revenue? … How do you plan to tell the difference between a parked/squatted domain and one in legitimate use but offering no public-facing services?" — a reader who quoted the funding gap line by line

The second is identity. "Free" and "free choice of domain" can't coexist, one reader argued, without inviting the squatting and phishing that got the free .tk TLD blocked everywhere two decades ago. The only fix is to tie a domain to a verified identity — which is its own privacy nightmare. A backend developer put the enforcement question in one line:

"One domain per person. How will you ensure this?" — a backend developer in Argentina

The third question is the most damaging: why a TLD at all? A veteran reader argued the whole scheme could hang off any domain already in hand — including the very onmy.cloud the proposal is hosted on — without the cost and ICANN ordeal of running a top-level domain. (Standing up a new TLD, another noted, runs around $600k with no guaranteed return.) An identity researcher who has spent a decade on human-centric internet identity offered a constructive path — zero-knowledge proofs to enforce one-domain-per-person without exposing who you are — pointing to Microsoft's Vega work as a model. The idea has a following; the comments are full of people calling dibs on your.self and treat.yo.self. But want isn't a funding model, and the proposal still has to answer who pays, who verifies, and why this needs to be a TLD.