Well you are still at the mercy of whoever's maintaining the network, aka some random-ass people all over the world. Aside from the few main ones, crypto networks are infamously ever changing and you're more likely to register a domain on something that won't even exist in 2 years because it collapsed completely. Maybe Eth is to big to fail now, but maybe it isn't.
That is the kind of tired, cheap, wrong argument that it so easily debunked, but "skeptics" still keep using.
Let me see if you really interested in a healthy debate or just proud of your ignorance: do you know that when using web3 for domains, you'd still be running the email servers that use the traditional protocols? You won't be losing any of your messages. You'd still be authenticating against a traditional server.
There is nothing about using web3 for domains that would lead you to lost access to your data.
You're right, you won't lose access to your emails, you'll lose access to your address, which really is the bit that I care more about - losing ability to log into sites and services is what really scares me about losing my email.
But now the important question - how does a web3 domain name solve the domain problem any better than just buying a domain, especially considering that "Permission to fail" is being touted as a web3 feature?
> losing ability to log into sites and services is what really scares me about losing my email.
That is also not a problem. Maybe current "login with web3" implementations rely only on your primary ENS record for authentication, but there is nothing stopping a MFA system where the ENS name is only one of the factors.
E.g: you could build a system that uses ENS as the first factor and a regular PGP client cert as the second. There are also "social recovery" systems. Even if you lose your keys, you can have a secondary address that can vouch for you.
IOW, stop arguing like you are the only one aware of the current limitations and that the thousands of people working on the space are just stupid. Before spewing ignorant arguments, you should learn a thing or two .
Can you point me to a blog post that explains how to use web3 name assignment technologies, in combination with MFA, to own and manage a domain that is robust to DNS provider shenanigans, crypto hacks I can’t anticipate or understand, and targeted harassment?
This is a genuine question. I’m in the market for new tools to manage my personal and business domains.
not a direct answer, I think way to design this system would be to make an ENS (with a competing resolver technology) and point that to IPFS hashes
in parallel, you would also look into how to secure your address, where the ENS is stored in
IPFS doesn't offer compute nodes and databases, so you would design your web offering around not needing that, completely rule out that idea in favor of a different one, or incorporate a different solution in addition to your frontend
but basically the way you design web 3 applications is not the same as web 2 applications to begin with. trying to straddle both worlds results in the worse experience. most similarities I can think of are kind of like omnivores trying substitute tofu based foods and having a bad experience, and attributing that experience to non-meat diets when it is just worse than a pescatarian, vegan and vegetarian cuisine.
> how does a web3 domain name solve the domain problem any better than just buying a domain?
I don't see how you addressed that. Apparently we would need to implement whole new systems? I understand, you gotta bring try to bring it into the domain and get belligerent to obfuscate the answers to simple questions, but I don't see how this addresses my concern.
> stop arguing like you are the only one aware of the current limitations and that the thousands of people working on the space
Ah, ok, so this is all in theory. Simpler answer would be "it isn't, but maybe one day it is." Thanks, I read between the lines and answered my own question!
> Before spewing ignorant arguments, you should learn a thing or two.
> how does a web3 domain name solve the domain problem any better than just buying a domain?
Was that your question? The answer is simple: once you lease (ENS only works with leases, just like traditional domain registrars, so I'll void the term "buy") a web3 domain, there is no one that can unilaterally take it away from you.
> Ah, ok, so this is all in theory.
Not exactly. It's possible, but impractical. If you are willing to accept the really poor UX and the lack of general adoption, you could build such a system today.
> The answer is simple: once you lease (ENS only works with leases, just like traditional domain registrars, so I'll void the term "buy") a web3 domain, there is no one that can unilaterally take it away from you.
This is all in the context of someone who didn't pay their domain renewal, and so it lapsed a few weeks later. How would this system behave differently?
“Permission to fail” is a desirable property in a research setting, not a global name service. This justification is in tension with the one offered above, and there’s been no evidence so far that “web3” is becoming more resilient instead of less resilient.
This is a cop-out: of course there are people with absurd risk or reliability profiles, for whom just about anything might be suitable. The entire point of systems like DNS are that they’re universal substrates that don’t require any exceptional selection. Economies and ecosystems are built on reliable and universal substrates.
My perception of resiliency is based on this[1], which has reliably increased in volume and quantity by year.
An exploiter's bug bounty on a still functioning yield farm that flash loaned itself, getting frontrun by a MEV searcher because the exploiter didn't hide their transaction directly with a block producer via bribery? That sentence wasn't possible a year ago. And now we all know not to design a smart contract powering a farm that way. I don't view that as failure. A regulatory sandbox would have deliberated for 5 years before allowing a launch and still never found that problem, resulting in a following multi year pause, committee report, and disbanding. I view this much faster higher iterative version as a win.
