Aaron Parecki

Until you can answer that, please stop pretending to understand handshake and lead discussions on it.

So you have *not* used Namebase.io? That seems to be the level of entry for yourself. Work your way up to the above.

You can access handshake in even more ways. At first glance it seems complex. That’s okay. So is ethereum. To learn more read this to remove your skepticism, you’ll be a SME: hackernoon.com/everything-you… medium.com/amentum/the-ca… medium.com/amentum/launch…

Not at all: as with Ethereum you have options for security tradeoffs. Operate your own fullnode, use a wrapper wallet like Bobwallet.io, a light client resolver, or a service like Namebase/Coinbase. Simplicity abstractions take time, and HNS is 6 months old /1

There’s only one implementation, HSD. Can you answer my questions so I can probe better? What do you mean “account” and why did you say the names were $49...then I can properly reply.

kind of proving the original point here ^^

Yep. That’s what I did to determine that is a mess.

I installed the code for several implementations / apps from github. And I’ve got experience building and deploying to production a bunch of decentralized tech from solidity contracts to ipfs/Scuttlebutt. I don’t see how non-technical people will ever use this.

The right way to use Handshake: > Download HSD or the HSD/Wallet wrapper, BobWallet.io > Get some HNS and send to your wallet > Bid on any <63char string you can think of by opening an auction for it on handshake > Buy a TLD for <$1 docs.google.com/spreadsheets/d…

Sir this isn’t how handshake works. Where did you “setup an account”? And...the chain itself you bid and purchase a TLD from. You can bid any amount... Let me guess you used Namebase (this is Handshake’s Coinbase) and you then tried to buy a name from the user marketplace...

A little of both, I guess? Probably also limited response sizes? Weird how culture dictates so strongly what's possible with technology, way more than the tech itself.

Yeah, right now through our service. We’re very much planning on associating them in a more decentralized way, but just so much to stand up for the system in general. It’s on the list for sure, though!

Is it held back by the idea that it’s cached? Or because dns is a sysadmin type thing vs application?

Never mind DNS software that supported dynamic / programmable responses!

For some reason DNS providers haven't iterated on the UX around configuring records much. If as much time and money were spent on making DNS easier to use without understanding all the terminology as was spent on all the blockchain tech, we'd be in a much better place right now.

@expede has done a bunch of nice work on establishing decentralized identities; currently I think the email address -> decentralized ID bridge is just a centralized service, though?

Yup, exactly. In my mind, that's the only real question to answer to address the (arguably theoretical) concerns that identity isn't "decentralized" with this approach.

So if you’re able to control a domain name and set Mx records you’ve got portability. It actually works pretty well but it’s not used for individuals so much as organizations. What’s the equivalent or way to extend it to individual users.

Totally - it's one of the things I love about email addresses. I think the standard complaint is that most addresses are managed by e.g., gmail. Probably the easiest way to fix this is an EU regulation mandating ownership of "@gmail.com" addresses.

in a way, email already is portable. My email address is on my domain, parecki.com, but it’s actually a Gmail account under the hood. If i want to move that to another provider, I can, and I don’t need to tell anyone a new email address. This… aaronparecki.com/2020/09/24/32/

It'd also be easy to imagine a decentralized store of "notarized" attestations of moving email addresses [to a different provider]. That's basically the root zone, come to think of it.

For sure; it just requires specific functionality on the provider's side. With webfinger, we always imagined that you could enable "portability" by setting a redirect or a pointer, again provider-dependent.

LOL. 90 days for trademark owners to register. Seems legit.

the “I’m locked out of my account” problem is not going away any time soon, and is a real hurdle in building this kind of thing in a decentralized way

I mean email forwarding is a kind of portability.

with that description, I don’t see how abnormal people would use it either!

I walked through setting up an account and identity on handshake. First the cheapest one was $49 and took an unbelievable amount of technical knowledge and steps. It made dns seem simple. I don’t see how normal users would ever use it.

ant colonies

i.e., is that something that *must* happen with consent from email providers/domain controllers? Or is it something that could be baked into ssb/deltachat/similar? What would ownership verification look like in that case?

Yup. I've said similar things in the past, but now more than ever, I wonder if the way to think about this stuff is to take a different tack, and instead ask: "Given all the potential negatives of email-style identities, what would email address portability look like?"

Ahh, ok! Yeah, agreed. I guess in that vein I'd argue that all those side effects/properties are *features* - thinks like user@domain imply the ability to cognitively and logistically offload all the responsibilities associated with managing "domain" to someone else.

Agreed! And all the blockchain versions of a global namespace have nowhere near this level of experience in maintaining the system in the long term, and are very likely much more fragile than DNS

I'd go further and say that *if* someone comes up with either a non-global approach to naming that works in terms of UX, regulatory alignment, and technology *then* it'll be pretty easy to migrate to it at that point for any existing system (like DNS/email).

I wasn't saying either one is better than the other. Just that whether it's a global namespace is a property that then has all these other side effects when trying to maintain that property in a system

The reality is that the only real question of a global namespace is a regulatory one - who gets ultimate control of it? And in that, DNS is actually not a bad approach and has spent ~30 years ironing out regulatory bugs and has found a pretty stable spot.

heh social distance

I know this is maybe something we'll never agree on, but I don't see the problem with a global namespace? Isn't it a feature? It makes the web work! It makes the phone system work! It makes email work!

Sure and that social distance way works for searching online but I hear folks saying their account names all the time. It’s important we do find a solution.

I don't think there's a clear answer yet. But I'm thinking something that matches more how people communicate identities in person. Plenty of people share the same name and it's not a problem the majority of the time during normal communication.

so, like URI s and URLs? but with relative namespaces?