Anyone can register any domain name by providing a public key. This name + public key (hash) is the actual internal network address. For every domain name, the network maintains a ranked list of such addresses. Rank here is simply a numeric value equal to the amount of SafeCoins paid to the network for that address. By default, the browser resolves a domain name to the highest ranked address. Additional SafeCoins can be paid for an address by anyone to proportionally increase rank anytime.
Hostile domain name take-overs are possible with this system, either with the goal of deploying malware, to harm a competitor, or simply to bully. This may be countered by the following measures:
If the two highest ranked addresses under a domain name have both relatively high ranks, the browser doesn’t resolve to the highest, but instead presents a choice to the user, who has to figure out which public key to trust.
Client software could (privately) store in your SAFE account a list of known (trusted?) public keys. When multiple addresses under a domain name have a high rank (due to a take-over), the browser automatically resolves to the one with the public key that is known (from a previous visit for example).
It requires wealth to assure your domain. Not everyone may like that.
Squatting isn’t profitable and will thus likely be non-existent.
SafeCoin’s deflationary nature causes take-overs to become relatively more expensive over time. It may only cost $10 to get a rank of 1000 in the early days of SAFE. Later it may cost $1000 to get a rank of 1000, because SafeCoin has become more valuable.
It creates another source of income for the network, so it improves SAFE’s economic sustainability. Network rewards (farming, dev updates) can be higher, and/or data upload prices can be lower.
It’s a system with little complexity, so it’s less likely to malfunction or to be gamed.
Websites with a strong community (a la reddit) can probably rely on their community to prop up the website’s rank when necessary.
I don’t think cost is an issue as you don’t need a glitzy name to provide a useful service.
I think you would need a grace period before the name was lost (maybe even months), as it could mean you would have very little time to respond to a take over attempt. However, if the opposing bids are irreversible, then you would need a strong desire to take the domain to try to wrestle it away - you couldn’t just be calling their bluff, for example.
I actually like the idea that big brands will need to potentially pay big money to keep the domains too. This is much better (IMO) than leaning on trademark law to strong arm competition.
Furthermore, this doesn’t need to be the only way to resolve addresses. You can use native URLs, other resolvers, etc. May the best resolver win!
This is the best proposal for the safe network DNS that I have read. The problem was hard and I had doubts that someone could ever solve it. And finally you came and you did it elegantly. This is brilliant!!!
May I add, with great humility, one supplementary element in your long list of pros:
The naming remains user friendly (no additional numeric characters)
Glad you guys like it! I think the best thing about it is that it’s likely that high profile DNS disputes would be quite rare. There’s hardly an incentive to spend valuable SafeCoin on a domain that you’re not going to make serious use of yourself, if you can’t even sell it at a profit.
Very good work there and as others said, sounds like a winner at the moment. It may be possible to better it, but I sure as hell don’t know how. My first DNS server I set up was in the mid nineties. Sh*t that is 20 years ago.
There is one significant Con that struck me straight away was that if I went to a site “goodiegoodieyumyum” one day and got the history on the Goodies, then the next day it sent me to a lolly site. That might piss me off a little if it happened too often.
Like, they pay SafeCoin to get their rank higher, but also get some sort of rank points based on how many people are using it?
Like if the real Reddit has a huge amount of people using their instance of safe:app.reddit, shouldn’t that count for at least something (in terms of rank) compared to some rich attacker who just paid a million SafeCoin to try and take over?
EDIT: oh I see what you mean. The community pays SafeCoin to prop up rank. I guess I was thinking something automatic, without the people having to donate/pay. Seems a bit like taxing/burdeoning the people lol
The way I see it is that the safe’s native DNS follows Seneca’s idea, but web/app devs can sidestep it by making the domain name a long random set of numbers and letters as another hinderance to domain squaters, and then use a normal app that functions similarly to DNS/domain names (sorry, my geek speak still needs improvement) that allows the developers to use the name that they actually want to use while the DNS/domain app redirects people to the actual site itself, so I imagine that this problem wouldn’t occur past the early days.
This could turn out to be a huge annoyance for full screen apps.
Imagine this stuff popping up on a housemaid while she’s watching a cooking video.
The need to store trusted keys would mandate the need to interact with the system and upload the list after every session from a different device. One PUT per login (if a new public key is added).
Which makes it possible for 2 users to go to g00gle.safe and end up on different sites.
google.safe would probably never become a target (because they’ll make a crapload of money), but personally I would attack sites like savethechildren.safe and such small-timers - you could bust them for peanuts.
One would never know how much he has to pay if the method of calculation was simple.
A sophisticated formula could be created but where would it run?
It’s not the “big brands”, it s everyone.
So an app vendor hardcodes his app to get some stuff from secure.safe, and a week later secure.safe gets bought by a hacker syndicate.
Ugh, as if there already aren’t enough ways to end up on a wrong site!
I think you’d tire of it quickly enough, it’s a never ending battle attacking domains you don’t like, and it quickly burns a hole in your pockets. Besides, the actual address always keeps working, so it’s not like you can remove the website itself. The only thing you gain is that new people who haven’t visited the site before have a harder time finding it.
No, that vendor would hardcode the actual address of course, hash(domain + public key). Possibly accompanied with plain text domain as well for readability.
It’s either a flexible system with that possibility, or rampant squatting. There’s no perfect solution here. I think squatting should be prevented above all, it would ruin SAFE’s DNS on day one.
A capable one yes. But you know how it’s be like. In recent years a myriad of mobile apps didn’t (and still don’t) verify SSL cert.
All users - not vendors, bug regular people - who use this system won’t navigate to address.safe and then link their page todasf0670324ophsudf.safe to make sure the link remains time-resistant. It’d be a mess.