Your Domain Is Your Identity: Lessons From the .name Fight
The .name fight shows why domain ownership and registrar risk matter — lessons on email identity, TLD choice, and infrastructure you don't control.

Twenty-two thousand people nearly lost their domain names and email addresses this year — not to hackers, not to lapsed payments, but to a business decision. Verisign planned to kill off third-level .name registrations, and only a long, expensive community fight forced a reversal: existing domains get grandfathered, new registrations close. That outcome is being celebrated, and it should be. But if you run anything that matters on infrastructure someone else controls, the celebration should come with a chill. The happy ending was contingent. The underlying risk is permanent.
I've registered domains for side projects, clients, and employers for about fifteen years. I've watched a registrar get acquired and double renewal prices, watched a client's domain lapse into redemption because a credit card expired, and watched a startup lose its primary domain for six days because of a botched transfer. None of that taught me as much as watching the .name saga play out, because it compressed every category of registrar risk into one story with a body count. Here are the mistakes practitioners keep making, told through what actually happened.
Mistake 1: Treating Your Domain as a Commodity Purchase
Most of us buy domains the way we buy office chairs: find the cheapest one that works, check out, forget about it. The .name people did something different without realizing it — they bought an identity. Neil Fraser has run his personal site on a .name domain for over two decades. That's twenty-plus years of inbound links, email contacts, OAuth callbacks, and account recovery addresses. When the termination was announced, he wasn't looking at losing a $12/year line item. He was looking at losing the canonical address for his entire professional existence.
This is the first fix: mentally reclassify your domain from "hosting expense" to "load-bearing identity layer." Your code can move hosts in an afternoon. Your database can be dumped and restored. But your domain is embedded in other people's systems — in their address books, their password reset flows, their ALLOWED_HOSTS lists, their DNS. You cannot migrate your reputation with a tarball.
Concretely, that means the decision of where to register deserves the same rigor as where you run production. Look at the registry operator, not just the registrar storefront. Ask: who ultimately controls this namespace, and what are their incentives? Verisign didn't want to terminate .name out of malice — it was close to zero revenue and nonzero hassle for them, so the rational move was to kill it. Rational for them. Catastrophic for the 22,000 people inside the blast radius.
Mistake 2: Registering on Exotic TLDs for Convenience
One of the most upvoted practitioner comments on the story was blunt: stick to the main TLDs. The commenter had bought into several new gTLDs, got hit with massive renewal price hikes, found transfers between providers difficult, and ended up retreating to .com and .net — calling .consulting the worst offender. That matches what I've seen with clients. A first-year promotional price of $9 becomes a renewal price of $60, and by then your email is printed on business cards and your domain is in every customer's inbox whitelist.
Here's the before/after. Before: you grab yourname.dev or yourstartup.io because the .com is taken, and it feels modern. After: you're five years in, the registry has raised prices 3x, you can't easily transfer, and the registry itself is a subsidiary of a company that might sell or shutter the TLD. The .name case is the extreme version — the registry didn't just raise prices, it announced the namespace was ending.
The .io situation deserves a mention here too, since it's the trendy TLD most of our industry actually uses. It's a country-code TLD tied to a territory whose sovereignty has been contested, and there have been credible questions about what happens to it long-term. I'm not saying abandon your .io domain tomorrow. I am saying that when I register a domain for anything meant to last a decade now, I check three things: who operates the registry, whether renewal pricing is capped or arbitrary, and whether the TLD's existence is tied to a single commercial entity's goodwill. .com passes all three trivially. Many fashionable TLDs fail at least one.
A domain you can't keep is a domain you're renting, and the rent can change to anything at any time — including 'get out.'
Mistake 3: Building Your Email Identity on a Domain You Might Lose
The most painful story in the whole saga wasn't the fight itself — it was a comment from someone who had let their first.last.name registration expire because the shutdown was announced. The grandfathering reversal came too late. Their domain and email forwarding were gone, and there was no realistic path back. They'd already done the miserable work of moving every account off their first@last.name address to something else. That's the cost of acting rationally on a threat that was later retracted.
This is the part of domain ownership nobody prices in: email is the skeleton key to your digital life. Every "forgot password" link, every bank notification, every tax document lands there. Lose control of the domain behind your email address and you don't just lose future mail — anyone who later registers that domain can receive password resets for every account you forgot to update. It's an account-takeover vector with your own history as the payload.
The fix has two halves. First, your primary email domain should be the most conservative infrastructure choice you make — more conservative than your hosting, your CI, your database. A .com you control outright, renewed for years in advance, with auto-renew and a backup payment method. Second, treat email-provider lock-in as a separate axis: your address should be portable across mail providers because the domain is yours. If your address is @gmail.com or @fastmail.com, you've made the provider the identity layer instead — different failure mode, same category of problem.
I've done the account-migration grind once, moving a client off an address they'd used for a decade. We built a spreadsheet of every service tied to the old address. It had 140 rows by the time we stopped finding new ones, and I'm certain we missed some. Budget weeks, not hours, if you ever have to do it — which is exactly why you should structure things so you never have to.
Mistake 4: Assuming 'Grandfathered' Means 'Safe'

Let's be precise about what the .name community actually won. Existing domains and email redirects survive indefinitely. New third-level registrations are closed. The namespace is now a museum: maintained, but not growing. Several commenters made the sharp observation that if they relied on one of these domains, they'd still move away — because an operator that tried to kill the thing once has told you everything you need to know about its priorities.
And there's a harder question hiding underneath: is the promise binding? One commenter asked whether Verisign's commitment to keep at least one registrar supporting third-level domains and email forwarding was written into an actual agreement, or whether it's just an announcement. As far as anyone can tell, it's the latter. Announcements are cheap. We've all seen "we're committed to maintaining X" age into a deprecation blog post three years later. If the survival of your identity layer depends on a press release, you are one reorg away from a bad quarter.
The transition period also exposed how fragile the support ecosystem is. Registrants whose domains came up for renewal during the uncertainty found that Tucows/OpenSRS wouldn't renew them, and almost no other registrar would accept transfers of both the third-level domain and the email forwarding service. One community member reportedly found exactly one registrar that would. Some domains fell into the redemption period — a state where they can't be transferred at all, only redeemed through the original registrar, who was refusing. That's a purgatory created entirely by process, not malice. If this sounds familiar, it's the same class of failure we covered in registrar lockdown vs layered defense: the lock that protects you from hijackers can also trap you with a registrar that won't cooperate.
So the lesson isn't "grandfathering is worthless" — it plainly saved 22,000 identities. The lesson is that grandfathering is a mitigation you didn't design and don't control, granted at someone else's discretion after a fight that cost real money and lawyers. You want to never need it.
Mistake 5: No Runbook for the Day Your Infrastructure Turns on You
Every production system I respect has an incident runbook. Almost nobody has one for their identity infrastructure. The .name holders had roughly two months between the termination announcement and the supposed end date, and even with that warning, people lost domains to expired renewals, uncooperative registrars, and redemption-period limbo. Two months wasn't enough because the escape paths were clogged.
Here's the runbook I'd write for any domain that matters, and it's cheap enough that there's no excuse not to do it this week:
- Renew critical domains for the maximum term allowed (often 10 years for .com) and enable auto-renew with a backup payment method on file.
- Set your registrar account email to an address on a domain you control, not on the domain itself — otherwise a lapsed domain locks you out of the account that would fix it.
- Know your transfer path before you need it: where's the auth code, is the domain locked, is the registrant email current? Transfers take days; disputes take months.
- Keep DNS records exportable and mirrored so you can redeploy to a new domain in hours if the worst happens. Test it once.
- Maintain a live inventory of every service, certificate, and OAuth app keyed to the domain. The 140-row spreadsheet is a lot easier to build before the fire.
- For anything critical, prefer registries with structural stability — large legacy TLDs, multi-year pricing visibility, and multiple viable registrars.
Notice that none of this is technical sophistication. It's hygiene, the same way backups are hygiene. The .name folks who sailed through were the ones whose registrations were paid up far out and whose domains weren't mid-renewal when the announcement dropped. Timing luck, mostly — but you can buy luck with a 10-year renewal for about a hundred bucks.
What the Fight Itself Teaches
One more thing deserves saying, because it's easy to miss in the risk-analysis framing: the reversal happened because people organized. Formal reconsideration requests, someone flying to an ICANN meeting in person, lawyers throwing the book, letters of support from across the industry. Registry operators respond to cost, and the community raised the reputational and legal cost of termination above the cost of maintaining a legacy namespace. That's genuinely encouraging — resistance, it turns out, isn't always futile.
But build your architecture assuming you won't win that fight. Most people on the wrong end of a platform decision don't have a Neil Fraser rallying a coalition and a legal team. They have a deprecation notice and a migration deadline. The entire discipline of running your identity on your own domain — a boring, stable, maximally-boringly-stable domain — exists so that your fate never depends on whether enough talented strangers show up to argue on your behalf.
We spend enormous effort on redundancy for compute and storage, and then put our single most irreplaceable asset — our name — on infrastructure chosen in a checkout flow in ninety seconds. The .name grandfathering is a win worth celebrating. It's also the best argument I've seen in years for treating domain ownership as the load-bearing decision it actually is. Pick boring. Renew long. Have the runbook. And be glad someone else fought this battle so you could learn from it cheaply.


