How is it independent if we rely on a dns service?
When do we start building a decentralized dns network? /halfS (i will search if one exists after this comment and if not il see if duckdns is still free)
Decentralized DNS (DNS servers that is) as a concept doesn’t really work, we have centralized DNS and Certificate Authorities in order to establish a chain of trust, for better or for worse. Otherwise, there is no way for me to say, claim that I’m google.com or some other big domain.
And really that’s just from a security standpoint. DNS also serves as a glorified phone book for the internet - but if everyone tries to make their own phone book and there was no centralized copy to refer to, you (or rather your computers/devices/etc) wouldn’t even begin to know where to begin looking up google.com . Even if I wanted to advertise myself as google.com, without some sort of centralized solution there is no one to “advertise” to in the first place, if that makes sense.
I think at least for the foreseeable future we’re going to be always tied to centralized DNS, unless we want to have something like key signing parties for domain names - and even then, I still don’t even think that’d work unless every device on the planet goes to the “same party” and everyone promises to have a unique domain name somehow… and a way to trust that the first “person” we find to be “advertising” their domain is actually the proper owner of the domain, in which case we’re pretty much right back to centralized DNS.
Hopefully that all makes at least a bit of sense, I’ve just woken up so I’m not 100% here yet haha.
Good morning, Thanks for your explanation. I am curious towards your phone book analogy because a phone book is like a ledger of information and creating a decentralized ledger is the main tech behind blockchain (which ive always understood as being more about the technological promise rather then virtual currencies). In my personal theory having a blockchain adress book of domain names sounds realistic. Is the problem that it would only work for those connected to it (so completely seperate from the current internet) or is there another technical challenge.
Ive found handshake which is supposedly backwards compatible with current dns system but it uses a p2p system and coins to vore: https://handshake.org/
What are your thoughts on it?
Also good night. Cause while you just woke up i’ve been delaying to go to bed.
In my personal theory having a blockchain adress book of domain names sounds realistic. Is the problem that it would only work for those connected to it (so completely seperate from the current internet) or is there another technical challenge.
Well, while I normally tend to have a kneejerk visceral reaction to blockchain tech these days, this looks to be something that would actually make sense/benefit from a blockchain based system. I’m certainly intrigued by it, but a couple of issues I can already think of:
From a user perspective: As you’d mentioned, it would rely on people being connected to it already, and people already have a very difficult time with the onboarding process of Lemmy
From a developer perspective: You’d most likely need some sort of library to handle communicating over this protocol, since none of the major (or at this point, I’d assume any of the) operating systems won’t support this natively. This means that not only does the server side of Lemmy need something to handle this, but any client-side apps (whether on PC or mobile) would also need a way to handle speaking over this protocol
From a server admin perspective: I’m not too sure about this one, but I wonder what the resource usage would look like for a protocol like this? Right now, its very easy for instance admins to just pickup say a $5/very cheap VPS from <insert provider here> and get up and running. I feel like the resource overhead of the protocol alone would make this a bit more difficult to get started. That’s just my guess though.
So handshake definitely looks interesting in general, but unless it really takes off and becomes incredibly mainstream (which I’d love to see - I’m all for decentralized tech) I think it would only result in more negatives than positives.
This is correct. Other servers will not connect with you if you don’t have a valid certificate.
Damned.
How is it independent if we rely on a dns service? When do we start building a decentralized dns network? /halfS (i will search if one exists after this comment and if not il see if duckdns is still free)
Decentralized DNS (DNS servers that is) as a concept doesn’t really work, we have centralized DNS and Certificate Authorities in order to establish a chain of trust, for better or for worse. Otherwise, there is no way for me to say, claim that I’m
google.comor some other big domain.And really that’s just from a security standpoint. DNS also serves as a glorified phone book for the internet - but if everyone tries to make their own phone book and there was no centralized copy to refer to, you (or rather your computers/devices/etc) wouldn’t even begin to know where to begin looking up
google.com. Even if I wanted to advertise myself asgoogle.com, without some sort of centralized solution there is no one to “advertise” to in the first place, if that makes sense.I think at least for the foreseeable future we’re going to be always tied to centralized DNS, unless we want to have something like key signing parties for domain names - and even then, I still don’t even think that’d work unless every device on the planet goes to the “same party” and everyone promises to have a unique domain name somehow… and a way to trust that the first “person” we find to be “advertising” their domain is actually the proper owner of the domain, in which case we’re pretty much right back to centralized DNS.
Hopefully that all makes at least a bit of sense, I’ve just woken up so I’m not 100% here yet haha.
Good morning, Thanks for your explanation. I am curious towards your phone book analogy because a phone book is like a ledger of information and creating a decentralized ledger is the main tech behind blockchain (which ive always understood as being more about the technological promise rather then virtual currencies). In my personal theory having a blockchain adress book of domain names sounds realistic. Is the problem that it would only work for those connected to it (so completely seperate from the current internet) or is there another technical challenge.
Ive found handshake which is supposedly backwards compatible with current dns system but it uses a p2p system and coins to vore: https://handshake.org/ What are your thoughts on it?
Also good night. Cause while you just woke up i’ve been delaying to go to bed.
Have a good night!
Well, while I normally tend to have a kneejerk visceral reaction to blockchain tech these days, this looks to be something that would actually make sense/benefit from a blockchain based system. I’m certainly intrigued by it, but a couple of issues I can already think of:
From a user perspective: As you’d mentioned, it would rely on people being connected to it already, and people already have a very difficult time with the onboarding process of Lemmy
From a developer perspective: You’d most likely need some sort of library to handle communicating over this protocol, since none of the major (or at this point, I’d assume any of the) operating systems won’t support this natively. This means that not only does the server side of Lemmy need something to handle this, but any client-side apps (whether on PC or mobile) would also need a way to handle speaking over this protocol
From a server admin perspective: I’m not too sure about this one, but I wonder what the resource usage would look like for a protocol like this? Right now, its very easy for instance admins to just pickup say a $5/very cheap VPS from <insert provider here> and get up and running. I feel like the resource overhead of the protocol alone would make this a bit more difficult to get started. That’s just my guess though.
So handshake definitely looks interesting in general, but unless it really takes off and becomes incredibly mainstream (which I’d love to see - I’m all for decentralized tech) I think it would only result in more negatives than positives.