Skipping past the telnet bollox pointed out by others, the next paragraph is the following gem:
"Browsers started using a centralized resolution system called the domain name system (DNS) to make things more user-friendly when the internet became more widespread. Instead of typing 208.65.153.238 into the address bar, a user could simply type "youtube.com," and DNS would do all the work for them"
Given that web browsers appeared in 1989, and DNS appeared between 1983 and 1987, we can safely say that DNS came before browsers.
Before that, folks kept a 'hosts' file updated.
Truly remarkable how much that page gets wrong. And to think it's just a blockchain variant of alt roots. :P You can read more about alt roots here:
Not that I think this project will go anywhere its antecedents haven't already, but...
> Given that web browsers appeared in 1989, and DNS appeared between 1983 and 1987, we can safely say that DNS came before browsers.
I mean the telnet thing is really wrong but this really just seems like picking on nothing to me. I don't think the intent there was literally to say "dns came along and changed how we used browsers" but to contextualize the use of names vs. ips in terms people in a modern computing context will understand more easily.
(That people had other means of managing host names before dns is a better criticism, but still kind of misses the point of what they're saying)
Nah. It's an attempt to cast the discussion about name policy as a simple technical concern --- before we landed on this weird DNS thing, it suggests, everyone was just typing IP addresses into address bars. But of course, there's much more to it than that: the names themselves are of immense value, and there is a massive commercial and public policy debate over how they should be administered. And when you think about how central those names are to how the Internet works, you're immediately left wondering how the hell some random coin should be in the mix.
I don't think it's ignorance motivating that text.
Axie seems to be the worst. This has become a huge fad in the Philippines. It's sort of like Pokemon, except that it costs about $1000 to get into the game. For a few months, poor people were buying into this and making enough money to quit their job. Then the supply of suckers ran out. The token is down 85% from peak.
There are now metaverse NFTs with "presales". The promoters don't bother creating a virtual world; they just sell "land" in one that doesn't exist yet, with vague promises of a future implementation. That's almost certainly an illegal public offering of an security under US law. (Look up "Howey test").
Even Decentraland, the NFT-based virtual world, only has 275 people logged in right now.[1] (Compared to about 40,000 for Second Life, over a million for Roblox...) People who bought "land" in Decentraland for a "virtual store" are not getting much traffic.
The SEC shut down most of the ICO scammers. NFTs were invented as an end run around securities law. Since each one is different, they're not a commodity, so the CFTC lacks jurisdiction. If they represent a real world thing, they're a collectable, not a security. But most of the newer NFTs represent nothing.
Getting away from uppity domain registrars would help. Originally, you owned a domain. Now, registrars take the position that you just rent it from them. But NFTs are not the answer. Too many scams and scammers.
I smell a false analogy fallacy. Axie NFTs are fad so just avoid all NFTs at all cost?
I think that the concept of an NFT is quite a useful one, and you should maybe not consider the value of it as a first thing.
Domain names were also speculated upon in the early days. Remember people buying up 3 or 4 letter dommain just to sell them for more money?
It's just a change in the trustmodel behind DNS that is really needed here. Instead of trusting a chain of certificate authorities you can trust a blockchain's consensus mechanism instead.
> I think that the concept of an NFT is quite a useful one
How so? Can you name a few useful applications? From my understanding NFTs are only as useful as "property titles", which means they have exactly one and only one application: speculation, which in my book is an absolute anti-feature.
> Instead of trusting a chain of certificate authorities you can trust a blockchain's consensus mechanism instead.
I think you meant registries, not CAs? How is "blockchain consensus" better, given all the blockchain failures we've witnessed so far with BTC/ETH and all the related hacks, monopolistic overreach from state-sponsored hardware vendors, and lack of consensus on the software leading to hard forks? Don't you think hard-wiring monetary incentives to something as basic as a naming scheme can only lead to bad outcomes (as ICANN gTLD programme has demonstrated)?
What's your take on the GNU Name System? Concerns about it being a GNU project aside ("it'll be ready when GNU/Hurd is ready"), what do you make of the features and threat model? Personal opinion: on paper, i find it's the least-worst naming system i've ever read about.
> Originally, you owned a domain. Now, registrars take the position that you just rent it from them.
I've never seen any mainstream registrar take that position. If you were going to argue that a domain name "belongs" to anyone other than the registrant, it'd be to the registry itself, not to a registrar.
ICANN: "Remember, paying to register a domain name is not the same as "buying" it outright or permanently. You do now "own" a domain name. What you are doing is more like leasing the domain name from the registry operator that the domain name is associated with. Registration periods can vary depending on the registry operator. You cannot buy a domain name forever. However, you can continually renew its registration (before the registration period expires) to maintain control over it. While the term "buying a domain name" is also commonly used in relation to purchasing an existing domain name registered to someone else, the purchaser is not buying the domain name, but just acquiring the rights to register it to themselves and continue to manage and re-register it in the future."
I am mystified by Handshake. It's not just a "protocol", as I understand it, but also a coin. In other words, this project has essentially claimed the Internet root as its own, and then securitized it, selling in effect shares. But they don't own the Internet root; it isn't theirs to securitize. I always ask: if they can do this with DNS, can I do it with ARP?
That's the thing about claiming the root: anyone can try. There is no title to the root, of course, as you well know.
You can certainly try to do it with anything you like, Sealand included. The real work is getting others to agree or care. Handshake has some preliminary support, and more, it seems, than any other project in recent memory.
Alternic once tried, as did Namecoin. IIRC there was also an alternative PKI root that some hackers tried to get going, but no browser ever incorporated into the trust store.
Alternic and Namecoin foundered on the same issue. To me, this Handshake stuff seems like the Springfield Monorail. It makes sense as long as you don't think about it, but as soon as you do, you wonder why on Earth anyone would agree to new root that gives value to a bunch of pre-mined coins.
Anyone can try! I'm not saying Handshake is, like, illegal. But it still seems like... a scam? I'm not sure what the word is.
They've reserved the top 100k Alexa domains on HNS for the original DNS owners. They're actively thwarting ICANN's overreach and centralisation of control by redecentralising the internet again. But sure, everytime HN sees a legitimate use case for a blockchain it's a "scam".
Is it a coin-like? Can you speculate on the coin-like? Did they pre-mine the coin-like and gift it to themselves and people they wanted favor with? If so, do I care that they've "reserved the top 100k Alex domains on HNS for the original DNS owners"? How nice of them.
Bollocks. What ICANN overreach? Creating around 2000 new top-level domains over 20 years at the request of the "global multi-stakeholder community" that anyone can be a part of? You prefer Handshake be the sole determiner? Why Alexa 100k not 10k or 1m? Who made them god of the root again?
ICANN overreach? You mean like opening a gTLD program (why not) but only for the richest corporations on earth. By making a 100k$ entry ticket they have made sure that only bda commercial actors are able to operate a gTLD, leading to outcomes such as many unprofitable gTLDs closed down over the years [0] and certain TLDs raising prices by over 100% from one year to the next [1].
Moreover, ICANN-affiliated industrial cartel is trying to keep this for-profit status quo, as exemplified by njal.la registrar operator Peter Sunde being denied ICANN accreditation despite being well-known (despite my political disagreements with him) as a good-faith actor [2].
So only the top 100k domains get to keep their names, and the rest of us can find new ones because we'll be domain squatted by early miners with tons of coins?
> They're actively thwarting ICANN's overreach and centralisation of control by redecentralising the internet again.
I'm all in favor of that. ICANN and its affiliated TLD/registrar mafia is screwing everyone. However, Handshake depiste the good stuff (distributing lots of cash to free software project) is not effectively decentralizing DNS governance, it's centralizing it in different hands: who controls the Handshake source code? what is the treat model of its blockchain in regards to Sybil attacks? Can the persons with the biggest wallets control the network? If so, it sounds like ICANN all over again, with "coin" speculation added on top.
> a legitimate use case for a blockchain
A blockchain in itself has many usecases as an append-only ledger. However, i've yet to see a legit usecase for a public blockchain, because trying to establish global consensus among unlimited untrusted peers is "hard" to say the least.
DNS sounds like a good usecase for a DHT, where Sybil-resilience has been better researched in the past decades. The root zone would simply be a well-known entry in the DHT, much like it currently is a set of well-known IP addresses and DNSSEC keys.
I strongly recommend you watch ICANN's "Emerging identifiers" session [0] from ICANN66 where both Handshake and GNS are presented. I found it very informative that Handshake is more concerned with the domain acquisition process (auction) while GNS is more concerned with solving actual problems we have with DNS (security and privacy).
Well if the thing they want is trust then lets "securitize" that. We can bundle up all the world's trust sell little trustcoins. Inevitably then whoever had the most coins would be the most trustworthy, and hence the source of truth for things like domain resolution etc.
“The internet was the evolution of telnet, which allowed users to connect to bulletin board systems via phone lines. However, connecting to those systems required that the user know the connection address and port.”
What? This is so confused that it’s “not even wrong”. Did GPT-3 write this? :)
As for the part that is just wrong, there’s never been an internet where you had to know the connection address (there were periodically updated host files) and you never needed to specify the port for telnet (it’s a “well known port”).
> As for the part that is just wrong, there’s never been an internet where you had to know the connection address (there were periodically updated host files)
Not necessarily. If your Internet connection was via KA9Q NOS over a DOS packet driver running SLFP (the predecessor to SLIP), you had to keep your own hosts file, and people would trade things like Scott Yannoff's list. If you look at http://www.nic.funet.fi/index/gnu/funet/historical-funet-gnu... for example, they'd list IP addresses and ports because some users didn't have systems connected to DNS.
People aren't talking about Telnet(23/tcp) as a protocol, they use that terminology because the command-line unix "telnet" client when connecting to a non-Telnet-service (meaning doesn't exchange WILL/WONT handshakes etc) on another port, it effectively acted like netcat. So if there was a cool game listening on 10.123.45.6 port 2000, people would type "telnet 10.123.45.6 2000" to connect to the game from many places.
I never understood why that sort of list wasn’t just a hosts file. But I wasn’t really involved in the SLFP world; maybe it didn’t support hosts files?
I guess my POV is that before DNS was introduced, the Internet had a hosts file you were expected to keep up to date, and after DNS, if you didn’t have DNS, what you had wasn’t quite “the Internet”.
> the Internet had a hosts file you were expected to keep up to date
I wasn't online back then so don't take my word for it, but my understanding is that different people maintained different hosts files. There was not a single Internet-wide hosts file, although one of these hosts files was somewhat more popular than others (due to who was maintaining it).
I used the Internet at CMU during the pre-DNS era, and my recollection is that the official hosts file came from SRI-NIC (this is corroborated by the ICANN history [0]).
Telnet is a protocol on the Internet, not a predecessor of the Internet. To connect to a BBS using phone lines, you just dialed its phone number; this has no relation to the Internet. Even if we generously consider a phone number to be a “connection address”, there’s no “port” involved and DNS has nothing to do with phone numbers.
I am sure this is not the objective here, but this article was so badly written and inaccurate that it made me lose respect for the IPFS project. There are a lot of technical assertions that aren't just wrong, they're absurdly false, within this article.
I was just about to submit an equivalent comment, so I'll piggyback on yours. I feel just like you, plus, I also hate to see them promoting another crypto scam attempt at taking over a piece of global infrastructure to make a quick buck.
I like IPFS as protocol. I didn't mind their work on Filecoin, as it made sense in context. But every time I check, they seem to be getting more and more in bed with dubious cryptocurrency initiatives. At this point I'm no longer willing to advocate for IPFS itself.
It's very disappointing, because I think the IPFS protocol has legs, but unfortunately they're promoting what I believe amounts to crypto grift here, and they're not even doing a good job at that promotion. There might even be a viable use case for an alternative root on blockchain, but this implementation with pre-mined associated coins and uncapped pricing is not it.
> Imagine if all Twitter users had 'verified' status and you could also find them across their other social apps without any risk of following fake or wrong accounts by mistake.
Without any risks? Ideas from self-sovereign identity are somewhat included in this promo. Does this service include hiring notaries to check state-issued IDs to verify identities? Risk-free IDs are unsolved I believe.
> Does this service include hiring notaries to check state-issued IDs to verify identities?
How would that be without risks? Corruption concerns aside, Nation-States are likely among the biggest identity fakers for their intelligence services.
I think "without risk" meant that discovery of alternative accounts/URLs is signed with a public key you already know of, so impersonation risks are reduced. I don't care to know "who" this is, as long as i know it's the same person.
So, silly question perhaps. I genuinely don’t know. Is blockchain handy for this decentralizing part of DNS or DNS entirely?
Suppose a blockchain network, or networks, knows all the TLD server IPs (I don't have DNS sharp in my mind, sorry for that).
You can transfer ownership of a domain to someone else. I think conceptually it’s an NFT right? Not that it is art, but domains aren’t fungible I think. So it is non-fungible. So you could wrap each non-fungible domain in a token on a network, making it an NFT.
I wonder what the actual issues will be. Striving for a decentralized DNS with whatever technique one would use seems to be a good thing IMO. The internet is for us all.
Edit: Ohhh this article is actually using a blockchain solution called namebase. When I was skimming, I saw IPFS so I thought somehow IPFS was used instead. That was more subtly interwoven in the story than I thought.
Conceptually, there is an important distinction between "mechanism" and "policy." There are many, many mechanisms for handling legal/technical control/ownership/system of record/etc. This is a problem with solutions that date to the beginning of civilization. Whether mechanism is pieces of paper in a vault or replicated records in different jurisdictions or hashes on blockchain matters only on the margins.
Policy is the hard part. What happens when mechanism fails or when the parties relying on the mechanism have disagreements about the implications. When there is disagreement about which blockchain is authoritative. (Sometimes blockchain people call this "governance" and wish to solve it also with mechanism.)
Anyway, there are good reasons for human organizations to act in a policy/governance capacity. One may think of any particular organization as flawed, which is no doubt true. But in any context where people depend on the mechanism working "correctly" people are the only participants who can provide that assurance.
The most popular decentralised domain name system is ENS. While domains are not called NFTs (ENS predates the term) it functions more or less how you described.
Of course it's not lol, like anything Blockchain it adds no value and makes everything it touches worse. If Apple loses their domain NFT should they what, just re-brand the business?
There's a reason we have a chain of trust and authority leading all the way up to the supreme court and not just "yolo, SFYL."
[edit] Like all the distributed identity plays, "not your keys, not your life" is such an asinine model to try and foist on people.
I've done published research in cryptocurrencies in its earlier days when I was in academia and I was very open-minded about it, but ultimately I came to the same conclusion: it's difficult to think of much that it does better than more straight-forward solutions and easy to think of downsides.
As I've gotten older and had to interact with various bureaucracies and paperwork, I've only become more convinced that lauded cryptocurrency features, such as (purported) immutability and decentralization, are a bug, not a feature. In my relatively short time on this mortal coil I've probably had to contravene some codified system four or five times by getting the right person in the loop so some rule, regulation, or contractual obligation can be lifted. No matter how hard we try, we can't anticipate everything. People will die, keys (and other vital paperwork) will be lost or destroyed, contracts will have mistakes, fraud will be perpetuated, money will be mistakenly transferred to the wrong parties, etc. All of these have happened to me and all were resolved by a sometimes annoying but ultimately effective system set up to handle such cases with some backstop authority empowered to fix things.
All of us will eventually get into some hell where we need a human to override the computer and just make the sensible thing happen. How often is it that we're on a support line just trying to get an exception to some stupid corner case into which we've gotten ourselves. It's always a relief when you can finally plead your case to the higher authority, whatever that may be, and get some damn relief. Why aren't more people concerned about a world where this isn't possible? It's an honest question. It totally baffles me.
Very good points. That's also my experience so far, and i think even the most dedicated ETH crowd will acknowledge they had to apply human authority to the source code when they found that "code is law" meant they would loose all their savings due to a bug.
However, i'm personally curious about mixed systems with crypto-identity and human-administered discovery. The best example i have in mind is the GNU Name System, where you can delegate specific names in your zone to someone else's public key, and each zone is administered by the associated public key. This way, you get technical reliability when addressing the public key directly, while still retaining user-controlled delegation of human-friendly names to build a global tree like DNS or a local friend-to-friend mapping (eg. `fatoumata.bob` to reach Fatoumata, whom you know is a friend of your well-known friend Bob advertised in his zone).
As someone who was involved in research in cryptocurrencies, what's your take on such developments?
I've never looked into the GNU name system in any detail, but it does seem to afford a reasonable degree of flexibility by allowing you to publish trust relationships and then delegate to different zones. If I'm interpreting things correctly, I can imagine people generally delegating names in their zone to ICANN in order to have something similar to existing DNS and its robust management, but additionally delegate to other entities for improved censorship resistance. For instance, dissidents could delegate to some private zone that likely offers fewer guarantees. Clients would need to be configurable and perform queries in some priority ordering.
Under this scenario, the ICANN zone could just refuse to publish an entry generated by a key that is deemed invalid (e.g., key is hacked, original private key lost, so revocation cannot by issued but authenticity verified in real life), so you'd probably end up in some situation where government blessed sources are canonical, but there's a reasonable fallback when official sources are not trusted. When these sources aren't trusted users would have to decide for themselves or do independent verification, which seems sensible to me.
I'd like to be able to have some indication that some private non-government controlled "root" zone and the canonical zone diverged so that I can then see if this is something I need to worry about.
Realistically, many buying a domain on something like godaddy would probably have their key managed by godaddy or perhaps tied to their credit card issuer, but more technical people would have the option of safeguarding their own keys.
There are other downsides to this approach in that a DHT is going to have different failure modes and scalability issues than DNS, but at least the obvious ones that occur to me are probably mostly solvable.
My takeaway from Web3 so far has been a mixed bag. On the one hand, having a wallet means you already have a pseudonymous account with payments for every dApp that has ever and will ever exist, and conversely the developers of said app reap the benefits of that fact for basically free. But this dynamic opens a whole can of worms in terms of risk.
There are plenty of ways to avoid 1:1 mapping between keys and access to/control of things on decentralised networks and blockchains, so really these are not arguments against blockchain just against a specific way you can choose to manage your keys.
Simple multisigs are a good start, as is key recovery via m of n (Shamir's secret sharing). More sophisticated systems have been built on top of this to allow, for instance, users to use their (singular) keys and to also allow a process, that remains decentralised, in which multiple [somewhat] trusted parties participate to recover access if they lose their keys or die.
These solutions are obviously more complex to build and interact with than a simple public/private keypair, hence why it has taken longer to develop them, audit then, and get to a point where they have a decent and intuitive UX (check out Argent, Gnosis Safe, among many). But if a big part of your argument against blockchain technology in 2021 is about lost keys, I'd highly recommend digging deeper into the state of the art before getting too stuck to a conclusion.
There was a time "I wouldn't use my credit card on a website because it's too insecure" made sense too, but HTTPS became the standard and we moved on. So it goes.
I'm well aware of the state of the art lol. But that's not a better system in any way. It's slower, less efficient and provides zero incremental value. And a ton of complexity and risk.
Is it possible to solve the problem? Maybe? Maybe not? I don't know. But so what if we do? In what concrete way is this better? It feels like at best we'll be right where we started.
It's a solution searching for a problem. A solution to a question nobody is asking.
[edit] personally I couldn't care less what database you choose to use to store DNS records - but it should probably be better than what we have now right?
> There's a reason we have a chain of trust and authority leading all the way up to the supreme court and not just "yolo, SFYL."
I partly agree with the argument, but i sincerely hope "your" supreme court will never be responsible for Internet governance. For what it's worth, i sincerely hope "my" constitutional council (France) will never be responsible for that either. Both are equally corrupt and reactionary entities, and i personally don't have any sort of faith in any Nation State whatsoever, as these entities were formed and architectured to oppress people and promote elites.
Well, I agree that we put our trust in technology an awful lot when we do this. But don't we put this trust into technology already?
One obstacle you described seems to be very much alive in the blockchain world today which is: what if you lose your key? Who can you turn to? Solving this (somehow) might help us to a more decentralized internet.
> But don't we put this trust into technology already?
No, that's why we have courts as the ultimate arbiters.
> One obstacle you described seems to be very much alive in the blockchain world today which is: what if you lose your key? Who can you turn to? Solving this (somehow) might help us to a more decentralized internet.
Anything short of "not your keys not your life" is a centralization risk and quickly, at the limit, approaches exactly what we have right now but less efficient.
The courts present a large problem though. Most digital crimes happen across borders. Courts are pretty much useless for ransomware, identity theft, and everything else online when the criminals are in an inconvenient jurisdiction.
Blockchain if anything shows the failure of our judicial system. Most of the cybercrimes are enabled by cryptocurrency and our justice system is hopeless to stop these payments.
The country in which businesses are being targeted. Crypto is a payment implement for crime. If the US outlaws buying crypto, companies in the US can't pay their ransomware bills and so will no longer be targeted.
I don't think you can put this genie back in the bottle. The US is a corporation cooperative and the banks are already in crypto and only raising their holdings.
Even if the US bans it, it just moves the market more into the dark. Ransomware will still happen and still require crypto payment. The only difference is the hoops companies will do to acquire this crypto in a semi-legal way (probably by using multinational entities to route payments).
That would be devastating to any country that doesn't follow suit. They would be the ones targeted by ransomware. So not only would they get all the ransomware attacks, they would also get all the negative value crypto creates. Since it doesn't create any positive value, that would be an excellent big brain move.
This is a pretty bad take. If you want a 100% trust less system, then sure. Go for it. But you can incorporate whatever logic you want into your smart contract. This can include giving a specific party the ability to revoke ownership.
HNS looks a lot like a protocol designed to make Namebase some quick cash. So it's all on the blockchain, but to get a name you have to buy it auction with tokens minted by Namebase. They've managed to make the costs of a domain uncapped, except the domain is also useless because no one but Brave respects their root.
In case you have any doubts that this an NFTified pyramid scheme, check out this gem from their "Earn HNS" page:
> For any HNS your friend buys, you get +20% HNS and they get +10%
Their giveaway didn't go to well. They planned to give away "70% of the initial supply" to developers, but...
> To date, around 25 million HNS (1.83% of the initial supply, 1.23% of the fully diluted) has been claimed through the developer airdrop.
They can't even give HNS away, however...
> The discrepancy between the total possible number of airdrop claims and the actual number of airdrop claims can be explained by the fact that eligible developers need to have their SSH keys from 2019 in order to complete the claim. ... Without their old SSH keys, it's impossible for eligible developers to claim the airdrop.
You’re mistaken. Namebase is an independent registar for HNS domains and it’s not the same as the Handshake project, which did the actual distribution of the coins.
I’m surprised people haven’t heard about this project, they have been linked on HN many times before.
I'm sure I and many others have heard of it before. It's just that there are so many of these 'X but Blockchain' ideas that come and go all the time that is just turns into a blur.
Nothing stops you from creating others TLD besides .bit.
We just have to formally agree to use, say `/something/`, for storing name records for .something.
Naming is solved by petname systems - just using a local address book that decides what name (if any) will be shown for what public key - it costs nothing, it doesnt need a blockchain. Unlike all other solutions, it actually _is_ secure. The only reason you'd ever arrive at something that is not petnames, after learning of the existence of cryptography, is because you are overengineering and falling for boogeyman SSL CAs want you to believe in. Commerce is not good or healthy in P2P networks, it is a last resort. Any time you hear about an economic model, incentivizations, etc, you can be 99.99% sure the company has designed it this way with money as their first interest, before utility. And not only that, but money will be practically their only interest. Thousands of P2P products and secure messaging have been astroturfed by people (corporations, and startups, which have no practical difference) "solving" Snowden's problem since the NSA revelations.
Nobody actually uses blockchains. They go on whateverscan.io and lookup their transaction there to confirm they were paid.
If you visit a bank and sign up, they should be giving you a public key, not a domain name. Not a name in namecoin or whatever this article is talking about that you need to download a blockchain for and have a computer fast enough to keep up with on 24/7. Also, you give them your public key. You shouldn't be giving them a TWITTER (seriously wtf, this is like using Steam for chatting with your girlfriend), no not your facebook either, not even your Google. I should give them a fresh public key I can use to authenticate myself.
Any blockchain aside from a small few (like Bitcoin, early Bitcoin at least) are essentially a series of logos and nothing more. They have not contributed one thing to actual decentralization. A good giveaway would be that not a single one has a proper CLI aside from Bitcoin. Not even Ethereum does. They all want you to use a bloated, intrinsically insecure web browser for everything, which cannot even work on more than two pages without billions of man hours. Naming is NOT solved by blockchains - it is not known whether blockchains will ever be trustless. Aside from the trend towards centralization from pools, at some point the data moves too fast for an individual to run a validating node.
IPFS seem to friends with Cloudflare. I don't what their relationship is (the former do truly seem to care about making something good, but are just naive on top of ignoring existing solutions), but I would not trust anything remotely related to Cloudflare ever. They have broke Tor for LITERALLY 10 years (~2011-2018, and after 2018 you still need to muck around with user agent and other headers) for no reason other than to market a snakeoil WAF (yes WAF, this is the WAF doing this shit. The Tor blocking was never about DDoS) to their users (aside from the possibility of them being NSA'd). If I was in charge of IPFS, I'd be writing blog posts explaining to Cloudflare while basic best practices for the internet include not blocking IPs for a decade. You know? You can't communicate with Cloudflare and keep a straight face when proclaiming to be making the decentralized web replacement.
IPFS sucks. Multihash, multibase, multiaddr, multiwhatever is a TERRIBLE idea.
So what's the alternative to IPFS? Freenet, another thing that has no use without webshit, as they forgot to implement a document format (which is also needed to implement petnames).
I have no idea why one would be enthusiastic about a tech company (not IPFS, the other crap) creating a P2P product, when tech companies in America have not produced a single good piece of software in 20 years.
> but bro, you are not thinking about my wallet, everything you say is bad for business
No shit. I am speaking as someone who wants to have a pseudonym and exchange information online without dealing with corpocrap and webshit. I don't need my services to send someone to my house to make sure it's really me. I want actual real decentralized technology. I like exchanging information. I don't give a shit about commerece. I like what the internet is actually for: Hosting a hacking blog, talking to people far away, sharing code, sharing cracks for stupid software that wants me to buy 10 copies of the game to have a 2 hour LAN copy, etc (yes, I have made all of these). Without laws and other things that only exist to serve stupid people. I should be able to code an entire conforming implementation myself. This doesn't have to be user friendly. There needs to be something that's actually good and secure for people who know what they're doing.
Handshake? WTF is this? It's named a verb. That means they have a whole actual marketer (or one of those "hackers" that is really more into fads and brand names than programming).
Petnames are great but the issue is initial discovery. Passing around public keys isn't fun.
Another IPFS alternative is BitTorrent v2 (The main changes are that it dedups files between torrents and uses sha256, not sha1) which does basically everything IPFS does. The only catch is that the spec is only a year old so most torrent clients don't support it yet. Also, RE: cryptocurrencies Monero is not a "series of logos and nothing more". It is private (Nobody can see your transaction amounts or account balance), just does payments (No contracts), has lower fees than BTC, and it has a (reasonably good IMO) CLI wallet.
an example of a working example is the p2p Secure Scuttlebutt (SSB) social network: it can work completely without name servers (you can even sync over bluetooth)
What's your take on GNU Name System? Globally-unique public keys with human-friendly nicks (hyper hyper local root) sounds like a good mix. See also, ICANN presentatin about emerging identifiers, introducing Handshake and GNS projects: https://icann.zoom.us/rec/play/QA_d2NDBBhbecI0Yg4WcEorM0DlVt...
Definitely we need a mix. These look great, the more alternatives available and explored the better for us all. Hopefully these systems and standards will organically diverge, converge, and diverge again (into infinity), as is key to any kind of technological development/experimentation.
How does content-addressing help implement a better name server? If the system is fully distributed you're gonna have a lot more than one person deciding that they are @bob
so you'd all have to be playing by the same id-generation algorithm/rules/'game', which will mean that no two id's can be generated (at the same time). if you don't use the same id-generation algorithm as your friends (yet you claim you did) people would be able to see that by validating your chain [1].
in other words: my main critique is that this name server system is not as distributed as it could be, for example if it was built using the holochain framework.
in a holochain system if two communities get disconnected due to a natural disaster, they would still be able to continue in two (temporarily) partitioned communities. with blockchain this wouldn't be possible.
"Browsers started using a centralized resolution system called the domain name system (DNS) to make things more user-friendly when the internet became more widespread. Instead of typing 208.65.153.238 into the address bar, a user could simply type "youtube.com," and DNS would do all the work for them"
Given that web browsers appeared in 1989, and DNS appeared between 1983 and 1987, we can safely say that DNS came before browsers.
Before that, folks kept a 'hosts' file updated.
Truly remarkable how much that page gets wrong. And to think it's just a blockchain variant of alt roots. :P You can read more about alt roots here:
https://en.wikipedia.org/wiki/Alternative_DNS_root
The Internet Architecture Board spoke out strongly against alter roots in May 2000, in RFC 2826: https://datatracker.ietf.org/doc/html/rfc2826
It's weird how folks forget history.