Web3 architecture and how it compares to traditional web apps
thenewstack.io
thenewstack.io
Internet is already decentralized for the most part. I run my own web server and mail server from my home. There's my "piece of the Internet".
Sure, DNS is centralized (even though it's heavily federated), and TLS certificates are centralized (less federated).
So, what we are looking to replace is simply "who is there?" and "is it the real you?" services with something less... centralized. This can live on the blockchain, sure, but blockchain is a database you sync across in its entirety, right? I am sure you can only get parts of it for some applications, but this, by definition, does not scale to the size of the Internet. (Hum, this is probably a decent dapp proposal: domain name system with certs, but it'd still grow very large if it was to cover entire Internet)
I don't get the connection with Google/Amazon/Facebook either: they are humongous services operating on the decentralized Internet today. Yes, people flock to them, but that's exactly the Internet you can avoid (a recent discussion on search.marginalia.nu gives a pointer as to how).
Making money on blockchain requires retrofitting the web to use blockchain to do what the web can mostly do already.
In "Blockchains Are a Bad Idea", James Mickens gives several arguments against blockchain-based systems like Bitcoin:
See his presentation here: https://youtu.be/15RTC22Z2xI
1. People have out-of-band trust relationships in real life which reduce the likelihood of malice. Bitcoin-style anonymous identities undermine trust relationships and are not needed for legitimate (not illegal) transactions.
2. Real life legal systems encourage good behaviors. If you have a dispute with someone, you can sue them. Bitcoin and related systems lack these protections.
3. Existing tools such as public-key cryptography and digital signatures can provide most of the functionality that applications need without the problems that blockchain-based systems have.
1. Zero trust is highly optimal for business. When it is systemically impossible to cheat your vendor/customer, nobody even tries. Happy path only.
2. Good luck suing someone in a foreign jurisdiction with incompatible legal systems. Maybe that sex toy would be embarrassing on ones credit card and tagged to ones identity. Cryptocurrency (especially private cryptocurrencies like Zcash and Monero) offers an alternative to doxing yourself.
3. You miss the great innovation of SN: Preventing Double Spends. You cannot enforce double spend without a source of truth -- which was traditionally central server/banks.
It was the Proof of Work that made the ledger distributed and censorship resistant, and it is the ledger that prevents double spends. And with Bitcoin, you have a limit of 21 million coins, which addresses SN's concern that central bankers ALWAYS inflate currency. So, no, existing tools did not suffice. That is why we have Bitcoin today.
I don't understand what point you tried to make, and frankly it sounds incoherent.
Moreover, the weasel word "highly optimal" means nothing. You failed to present any value proposition, specially when discussing a solved problem.
This is perhaps the main problem plaguing web3 proponents: they always fail to present any value proposition at all, and instead resort to all sorts of handwaving to try to distract away from that void
> 2. Good luck suing someone in a foreign jurisdiction with incompatible legal systems.
This statement fails to take into consideration the very notion of a jurisdiction, and somehow presumes that a shop in a country which chose to do trans-national business shouldn't have to deal with the legal implications of operating in another country.
> 3. You miss the great innovation of SN: Preventing Double Spends.
That is not innovation in commerce: it was already a solved problem before the inception of blockchain-based technologies.
Email interoperates, but messaging doesn't. There's no adopted standard for identity (except centralised Google/Apple), or for payment (except centralised PayPal/Apple, and the bad UX of entering credit cards).
Turning these (and more) into protocols would allow more innovation at a layer above that. This innovation just can't happen right now, as FAANG have to approve of everything you do that works with systems the populuation already use.
web3 doesn't seem good enough yet to help, but it is at least working on it. And it could potentially solve the lack of identity and payment standards.
This is incorrect but educationally so: there are many new distributed protocols (RSS/Atom, XMPP, various things around blog interactions like FOAF and ping backs, GraphQL, OpenID, etc.) but the ones which successfully saw adoption were adopted by a few large companies who saw advantages from doing so. Being open or distributed not only wasn’t enough to be competitive with the centralized platforms, it often prevented that since you needed to coordinate improvements across a range of competitors while Google or Facebook could just ship an update.
Consumers didn’t value client decentralization, especially when that meant incompatibility, lower performance, and fewer people they know using a service.
Putting things into a blockchain not only doesn’t help with any of the real problems but makes all of them worse: performance and reliability are completely out of your control, everything costs money (good luck getting users when the competition not only has better features, perform, and social graph but also doesn’t charge them to sign up or send messages!), and since you centralized onto a network run by finance companies the only way to get traction on any of that is to have enough money to be taken seriously.
Standard for identity - not sure what you exactly mean by this, but there’s WebAuthn / OpenID Connect / SAML
Standard for payments - W3C web payments
Whether it’s Bitcoin, DeFi, DAOs or NFTs Web3 allows you to participate in owning digital property at the protocol layer. Imagine if you could have profited directly from your early adopter curiosity and passion for the early internet by owning a part of the base layer, instead of waiting till the IPO where VCs and early shareholders take most of the profits before dumping their bags on the public.
That’s Web3
And it's already there it seems: https://ens.domains/
I wonder how does one integrate ENS into an existing DNS-based system — is there a DNS-compatible resolver that will look up names on ENS (and an IPv4/IPv6 proxy perhaps)?
It's also a missed opportunity to provide hierarchical certificates at the same time, signed by the registry, though in a sense, a smart contract is exactly that, but I am more interested in simple name-resolution and TLS communication for traditional web applications like HTTP and mail (name-to-IP).
We've decentralised transfers with BitTorrent pretty well, and it's actually reasonably fast and cheap to operate. Opera used to have it built-in, and there are some WebRTC versions.
The web has a decentralised database of domain ownership - Certificate Transparency. The same approach could have been done for NFTs if they were really about owning digital assets, and not gambling on a pump and dump scheme.
Basically there are lots of energy efficient and lower-latency ways to decentralise the web if we wanted to. But the web tends to clump around large services mostly due to economies of scale and user convenience (average user can't be told they've lost their private key and they're totally screwed).
The radical change here is that now anyone with some pocket change in crypto can deploy a decentralized application. We don't know how that's going to be used yet, it's still so early, but it's certainly exciting.
Web3 is a terrible name (as everyone else has already pointed out), but the magic that shouldn't get lost here is that a new decentralized internet is being born.
Oh man it'll be awesome. Decentralized applications running on someone else's hardware with no SLA or even a meaningful relationship between the application author and person(a) executing it. Terabytes of data will just sort of float on the blockchain...sorry, immutable data store. I'm sure it'll be super fast when uploading from people's residential Internet connections.
So exciting!
Genuine question, how is that different from renting a VPS or buying a Raspberry Pi and running your website on it?
What about decentralized malware?
Web3 seems like a perfect opportunity for malware applications and hosting illicit content.
Have some CSAM you want to host? Put it on Web3! Want to dox someone or leak sensitive info captured in a hack? Put it on Web3!
Edit: This also begs the question: what legal liability might one face for hosting illegal content on Web3?
It is like crypto is a solution looking for a problem. It feels like the whole blockchain hype all over again. More noise than actual substance.
Anyone know of an example for a lay user where crpto has a huge benefit compared to other tech in the space?
Disclosure: I own ~$1 in cryptocurrency I forgot the password to years ago
The 3 in Web3 is just there to wipe the slate clean in order to resell just about any original vision that previously failed.
But blockchain itself does not prevent copyright infringement of digital assets. I.e. the artist may sell his photos as NFTs, but I can still use them without permission or resell them through other channels.
See, for example, the many NFTs traced directly from artwork by real artists.
A lot of folks hopping on the hype train mentioning they are doing X with blockchain are missing the revelation. blockchain is just a piece of the innovation, imagine if when the car was invented and suddenly inflatable tires got all the hype … that’s pretty much the times we are in. The whole is greater than the sum of its parts.
All the pieces that made bitcoin were already there. Satoshi was just the first to arrange them in the right order
There are also some new and interesting paradigms however, that enable different systems and possibilities on the web than what we can achieve with our traditional, centralized walled gardens and payment services. See here for an idea:
Digital artwork is not magic: someone holds the copyright and needs to pay for hosting. NFTs don’t help the first problem because blockchains are too inefficient; IPFS helps only if you find value in paying for your hosting using that protocol rather than something else — nobody is hosting DRMed content for free so it’s just a question of what storage characteristics fit your budget. Since blockchains are inefficient, more expensive, and require you to maintain a complex operating network coordinated with other people just to use them, that’s unlikely to be a preservation success.
Rights is similar: NFTs are a distraction by people trying to get you to buy their Ethereum. What actually matters is a legal statement transferring ownership – if you have an NFT and the artist says they didn’t sell the rights, you’ll get laughed out of court but the reverse is true if you have any statement from the artist saying they did – we don’t need PKI, even, a witnessed email or note will suffice. That’s important for the discussion about forking: just as nobody is offering truly free hosting, whatever “the community” says is noise — what matters is what a court accepts, and a community which tries to ignore that will get DMCAed off of the internet.
Not sure what you are arguing about hosting. IPFS is already working well for many NFT projects, because the artist + collectors have an economic incentive to continue to host and maintain the media pointed to by the blockchain tokens.
Also, you may perhaps be aware that NFTs don’t usually convey ownership but a high percentage of the discussion confuses that. Whether or not that’s an intentional sales tactic there are a lot of people who’ve been confused on that point and left wondering why the price is so high for a link to content owned by someone else hosted somewhere else.
Ownership in the context of NFT is not about ownership of the artistic intellectual property.
Further reading:
https://nathanielstern.com/text/2021/custodianship-copyright...
How does pricing compare to the existing options? On-chain hosting already limits you to public works, which many people are not going to be comfortable with, but this usually comes down to cost and reliability.
> Ownership in the context of NFT is not about ownership of the artistic intellectual property.
I'm aware but that's not how most of the marketing sounds and it also gets into the question of value. If an NFT doesn't convey real rights and requires paying more for hosting (which is also separate in the case of almost all NFTs currently), the ceiling for long-term value is pretty low.
The decentralised part acts as the security and it would be hard to forge records, could make the process more transparent and therefore less corruption.
As a side note, I'm shocked by the amount of cynicism in this thread. It seems a lot of people have been put off from the negative media surrounding crypto rather than looking at the technology behind it with an open mind. It reminds me of when P2P file sharing was demonised by the media because of pirating.
As for the cynicism, note that this is an audience with deep technical knowledge and the ability to reason about systems, years into salespeople making big claims which have not been delivered. Given the conflicts of interest inherent to a system which requires outsiders to put fresh money in before insiders can cash out, skepticism seems warranted.
Also, I really do think that NFT ownership serving as an access token into a service is a very neat feature that we'll start to see more of as people start getting more creative with NFTs instead of simply being an art asset. There's a lot there and we're barely scratching the surface imo.
From an end user perspective, all I need to do is press a button to connect my wallet, and I'm in. The site doesn't even need to reload or hit a callback uri. It's a much simpler flow than any of the oauth authentication flows. Also, as anyone can create a wallet and it is not tied to any other identifying information, it's a breath of fresh air to have some vague sense of anonymity than hooking everything into a centralized email account.
From a developer standpoint, all I need to do is use an API with a callback function that on success takes whatever data I need and update my application state appropriately. It really is smooth, and I don't need to do things like maintain session or handle access token expiration and request refresh tokens. I can sign messages on the frontend & verify signatures on the backend to ensure the user is who they claim to be. It's so simple compared to the stuff I had to implement in the corporate world for OAuth (granted, the projects I've been working on in the space are much less complex).
If you want to try it out yourself as an end user, just download a wallet extension and check out an nft marketplace such as solanart or opensea. You simply have to connect, and that's it. I find it fascinating personally and it's one of the pieces of web3 that I'm most excited for. I believe we're going to continue to see a lot of innovation in the space.
Having set up a lot of SSO and other auth mechanisms, I think there is a good opportunity for a combination of NFT-access-token + WebAuthn [1]. The nice thing about that would be a secure login without email/passwords combination that you can access with your biometrics. Imagine the same flow as login with 1Password without having to fill the form.
On the developer side, you don't have to setup the OIDC flow, database, or anything. You just rely on a generic service to provide you with a token. Less opportunity to mess it up, and it can work on localhost too.
Now, would be the question of a similar exercise for authorization.
I'm writing an app that utilizes NFT ownership to authorize access to a service, and it's been really easy to implement, though I'm certain I'll need to continue making improvements to it. But my theory is that I can sign a message with my wallet on the frontend using NaCl, send that message and the signature to my backend and verify it, then verify token ownership using RPC. My PoC implementations have been working pretty well, but I am not sure of any holes that may exist in this approach, so it's still a work in progress.
From the user’s perspective I see the (dubious) benefit of being able to use the same key for your money and online accounts, but that’s about it …
The only way I can recommend getting started is the way that I did a few months back - I strongly encourage checking out and entering random hackathons! https://gitcoin.co/hackathons
There's a lot going on in the space currently - many projects are trying to attract development talent so they will sponsor / participate in hackathons. imo they're fantastic opportunities to learn something new and potentially win some extra money.
I'd probably recommend ethereum as a first ecosystem to really dive into - the tooling is more mature than other dev-centric blockchains. However, keep in mind that ethereum as a base can be a tricky place to launch your app because of high transactional fees, and you would likely want to integrate a layer 2 like Arbitrum or support alternative coins in a final application. I'll link some pages to a few technologies I used when building some ethereum dapps
https://www.trufflesuite.com/truffle https://www.trufflesuite.com/ganache
for wallet integration, either of these work https://github.com/ChainSafe/web3.js https://github.com/ethers-io/ethers.js
I highly recommend checking out The Graph too - it allows for you to leverage graphQL on your smart contract data by creating a dedicated subgraph. With some careful integration, you can replace a dedicated backend that would normally be required for various APIs with this. It's a very exciting project imo. https://thegraph.com/en/
And of course, the solidity documentation is very helpful. Coming from a scripting heavy background, solidity was very easy for me to pick up and write a simple escrow service. https://docs.soliditylang.org/en/v0.8.9/
I'd also recommend checking out other chains once you become more comfortable with the system. I'm currently working on Solana and it's a bit more difficult to work with due to the tooling not being as mature, but it is a very interesting chain to work with and I hope that it can scale better than Ethereum has.
A simplified example: You own an NFT and want to prove that you own it.
You can produce some chunk of data (not a transaction) that includes a reference to the NFT (be it the UTxO, Tx, or some hash) and sign it with the same key that owns the NFT (or a derivative key depending on the design of the system).
Since it's not a transaction it doesn't cost anything to produce and you can share it with the interested parties however you wish.
In a sense you can think of it like GPG (and GPG based SSH keys) except instead of the web of trust being a federated set of servers and key cross signing, it's a decentralised ledger.
It's not simpler but it can be easier because it largely comes out of the box set up. The PKI is already well established and you don't need to really spin up or configure anything complicated to use it. You just launch a wallet with the supported features and it "just works".
How complicated and expensive can you get. Engineering is about solving problems with the right tools and with the least amount of maintenance.
This is inventing a whole lot of stack for what? Your pipedream of "decentralised" internet. They fork Ethereum chain continueslly for their own gain, aka it's managed centrally!
/endrand
What is the problem and solution here? If you say illegal activities, gambling, and evading state regulations, I can see it. Great product.
Citing market cap isn’t that convincing. China has a property market with similar value that’s being threatened as we speak. What does a very large number prove except there’s been investment? It certainly doesn’t prove it’s a sound investment or that it has a future beyond a vehicle for the uses I mentioned.
I can see that people enjoy gambling, but that doesn't mean the actual tech is good.
In short: who is paying for the Ramen?
Let's take a multi-user chatting app. The compute power necessary to serve 1000 users is X. Running this costs us $Y. Now we distribute this app, having the same features, I now need to spend X plus the overhead of storing data on-chain plus - likely - extra network hops since my users form a peer2peer network of sorts amounting to $Z.
So now the cost is $Y + $Z: Who is paying for that?
I’d guess for something like real-time chat you’d want another chain that could resolve transactions much faster and cheaper than Ethereum could and would have to buy their tokens to use it. Then hope that speculators don’t push up the price until using it for the intended purpose is in feasible.
Ethereum gas is the miner fee, they're not two separate things. More compute-intensive transactions cost more gas, and the price of gas is set in the market by supply and demand.
A more apt example that exists already is a digital art marketplace[1]. The burden of maintenance for the media lies on the users (those publishing & collecting the assets), and hosting is done in a peer to peer manner through IPFS[2]. The maintenance for the site (servers, indexing) can be funded & incentivized through (for example) a small % cut on all trades as a maintenance fee. In an ideal platform, these fees would be directed to a wallet that is collaboratively controlled by a number of shareholders who have an interest in continuing to support the project.
Most importantly though, and one of the core aspects that tends to be lost in these discussions, is that the assets themselves are typically decentralized. If the platform fails or a fork is desirable (eg: community wants a different direction than the original founders), the assets still live on the blockchain, and a new platform can easily emerge on top of the same data if there is enough interest. This is where you get the idea of “users own their data, and websites must compete to display/facilitate it”—a radically different paradigm than what we have seen in the past.
[1] - https://restofworld.org/2021/inside-brazils-diy-nft-art-mark...
[2] - https://gist.github.com/mattdesl/47f4ea12ea131eed8401bdacf95... (I believe this guide is outdated now as I haven’t been keeping up with their tech stacks, but it gets the point across)
It's really almost too simple.
I expect it'll be just as successful as Web 3 was.
The tooling around FOAF and similar semantic web microformats wasn't very good, and Atom/RSS wasn't needed or widely supported when the web became increasingly siloed. Web 3.0 died when personal websites and blogs stopped being the most common publishing medium.
"New web 3.0" is web + NFTs. The analogy I see thrown around is this :
Web 1.0 : read-only
Web 2.0 : read + write
Web 3.0 : read + write + ownership
Which is stupid, because we already have ownership.
Wikipedia is hosted centrally, because running infrastructure in a decentralized way is worse in every technical sense. The actual content and maintenance is as decentralized as you can possibly have.
I'm sure you haven't really put much effort into actually looking into this, but you can volunteer to work on the underlying software, and infrastructure as well. The wiki and infra code are open, the infra is documented openly, and the test/dev/staging environments allow volunteers to directly participate. You can have your code end up in production after review.
To be fair, we already had the "write" at the Web 1.0 stage. This is like "what the unwashed masses" can do.
Web 3.0 is doomed.
In other words, a guy heavily invested in speculative technology X, is pushing X.
This is clearly a solution in search of a problem.
I found it quite hilarious though to see the diagram in the article that explains how a decentralized Web3 app is supposed to replace a traditional centralized system, which ended up being plastered with the logos of companies whose proprietary commercial solutions you need to operate a dApp.
You can even call it a niche.
If you can capture part of the value of this market by creating value for those users... what's wrong with that? How is this a non-existent problem?
NFTs are a good example of what you're dealing with. There are so many people with - arguably - apparent "poor taste" exchanging large sums of funny-money because that's part of their vision of an end game, and maybe that "poor taste" it's just because there's nothing else to do with the funny-money.
The problem here is that they want to spend it, exchange it, do things with it, and dApps are here to serve them.
From my understanding of this: the Web3 is not about/for you, or me, it's more for them.
If they are hyping it to be the future, let them be, it's their thing and we "don't get it" - that's ok. We've seen this over and over again, some stuff stuck around, others faded away. But it's ok to not understand some social movements and not be part of them.
You're probably experiencing what some folks experienced when Rock Music came about, it's confusion and it's an odd feeling. I'm not saying crypto will have the same popularity of rock music... I could say another example like EBM music... the point is that we don't understand a lot of things.
If you want to appreciate art (of whatever type) what additional benefits of that art does the NFT provide? It's not access to the original image, those are just files, it's only certification of ownership of a particular instance of that image, AFAIK.
The concern I see with NFTs and other similar applications of blockchain tech, is that it's not about the application as much as speculating on them to make money.
That speculation, like all speculation, is susceptible to people overhyping it, leading to lots of losses for late adopters.
Supreme is for the "cool/boasting" factor. Pokemon/MTG is a mix of nostalgia with boasting on social media. Idk what Beanie Babies was about but it had a short run.
Is it an arbitrary choice? Does it require some (little) taste? I don't know, probably A or B, or both A & B.
We all buy stuff with the hopes that people will pay more for it in the future, the difference is that the majority of people exchange broadly accepted symbols like stocks, precious metals, and that renders those symbols as more safe bets - but the motivation is there if you strip off the symbol.
With that said I would never buy a "Bored Ape" NFT. Like I would never buy old Air Jordan shoes for thousands of real money units.
>If you want to appreciate art (of whatever type) what additional benefits of that art does the NFT provide?
Maybe the main value that NFT seems to provide is that is exchanged with crypto. Like I said in my previous comment, it's something people can do with the funny money.
>That speculation, like all speculation, is susceptible to people overhyping it, leading to lots of losses for late adopters.
I do agree with you, yet I need to recognize as well that this thing has momentum and a life of it's own. Even if it is simply because too many people have invested too much on it to go back (time/money), or even if they just believe in that vision...
The truth is that still this is a market that is self serving itself, to the point that they seem to be creating new protocols for that purpose. Itself has it's own problems and there are solutions for those problems, that's how value is created and exchanged. Even if the problem is "we have all this funny money and we want to spend it".
Can it all crumble down? I think it can, like many other cultural movements. But I need admit that I don't understand a big part of it, and I think that a lot of this tech literate user base of HN is on the same boat as me.
I reject efforts to try to force Blockchain into it.
Can someone explain like I am 5 what the benefit of Blockchain is for something like dat or IPFS? Seems like they work just fine without Blockchain ?
Decentralisation, open collaborartion, opportunity for entirely new paradigms of entrepreneurship free from the constraints set by the rulling professional managerial class. Weren't those core hacker values?
The Nakamoto consensus algorithm is the single biggest contribution to CS so far in the 21st century. Choosing ignorance of novel technology is really sad to see on a forum intitled "hacker news"
So the solution is centralizing a currently decentralized network onto a blockchain controlled by a few major players?
Remember: honesty is also a core hacker value. Before saying that “open collaboration” hasn’t been happening for decades, you might want to consider how credible that unsupported claim sounds when your audience knows that you will personally profit every time someone buys into the system you’re pushing. Hacker culture has traditionally found that kind of sales pitch disdainful.
>centralizing a currently decentralized network onto a blockchain controlled by a few major players
No, the basic idea is simply decentralization of trust. I do not know what specific blockchain you are refering to.
>someone buys into the system you’re pushing
I'm pushing that we come together to develop novel systems that fundamentally change how governments and human collaboration in general work. Speculative markets do not interest me.
> No, the basic idea is simply decentralization of trust. I do not know what specific blockchain you are refering to.
All of them? If I host a regular web app, I have a choice of thousands of hosting providers around the world and can self host, with the primary constraints being low-level network issues (e.g. sending spam will get you black-listed). If I use a blockchain I still have the network level concerns but am also ceding control at a higher level to a much smaller, far less experienced community dominated by a few major players. If I adopt someone's distributed app framework on top of that, I'm further reducing my choices and taking on support responsibilities for their decisions when those impact things like performance, security, abuse management, etc.
> I'm pushing that we come together to develop novel systems that fundamentally change how governments and human collaboration in general work. Speculative markets do not interest me.
I'm not opposed to that but the current blockchain hypefest has been a bunch of people pushing technology which cannot work to build those systems. That's why the rhetoric always boils down to the “buy in now or you'll wish you had later” needed by people who have already bought in and are worried about maintaining their ability to cash out. If they were building something novel, you'd know about it because they'd be showing new things which couldn't be done with existing technologies.
Unless I'm missing something.
Yes, both of these things exist in web2.
But, the exciting thing about web3 isn't decentralized payment. It is that you can participate in decentralized markets.
For example, if you go to any website and get an ERC20 token, you can use that token without permission from the site where you got it.
Those tokens can be traded on that site, or another marketplace, or another dapp can use them. The original site can't control that.
If you go to uniswap you can see the prices of all those tokens. It's useful for speculators now but if those tokens have true utility value, it starts getting very interesting and very valuable.
Marketplaces are where things are traded freely. Web3 doesn't make payments any more free. It just makes things more freely tradable.
But consider:
- if Google were to stop functioning tomorrow, would you still be able to identify yourself to the web sites you use?
- if dang decided to censor your HN comments, could you or I easily spin up an instance of HN that ignored that censorship?
- if someone were to come up with a better algorithm for generating your Twitter feed, could they implement it and attract users to their site?
The way the web works today, we don't have equal access to the data that powers the applications we use.
I'm not saying which is better, but the type of decentralization envisaged by web3 proponents goes further than what we have today.
“The early Internet was useless too!”
“Everything on the web in the 1990s was a scam too!”
It’s demeaning towards the online pioneers, and it’s depressing if young people actually start to believe this stuff just because some dweeb wants to sell them multi-level casino tokens.
--
1. at least its Ethereum-based implementation, which is the vast majority of web3 stuff.
- Price
- Convenience
The problems so far we've solved with crypto seem to be around bypassing nation state regulations and as financial asset to gamble with. In those areas crypto has solved for convenience.
It's plausible that some time in the future users will need / want decentralized platforms with different ownership models for content. Is that time now? I'm not convinced. Being too early isn't great even if you're right eventually.
It might be cheap unless you actually want to use it for something heavy. Azure is really good at making itself look cheap by assuming that no one ever really needs to use its pricier performance options. It's cheap for CRUD, but once you want to crunch some numbers, like process financial statements for one month, you either wait a week to finish or pay a fortune.
We looked into replacing one (1)local SQL-Server that is able to crunch our monthly statements in a few hours. It would have cost us around USD $15,000 on Azure per pop.
The local machine cost $30,000 to buy + about $1000 in service plans / month, and will be used 2-5 years. And it would run other stuff for the rest of the month, at no extra cost.
Using a social recovery wallet (Argent specifically) has been my top Keanu Reeves "WHOA!" moment in crypto. Years ago when I started coming to these forums and everybody was talking about how this or that chain's TPS is the highest my argument was always that none of it matters to end-user if the UX is not there. I would repeat this ad-nauseum and get downvoted. Feeling like wallets are the key I started a website and twitter account that analyses and compares the UX of different contract wallets. When Argent came around I stopped what I was doing because everything else was pointless. The UX comparing to other wallets was incredible. But even more importantly the user experience was better than my bank.
You see, I fly a lot. And when I fly to different countries my credit cards (no matter how "international" they are) constantly get rejected and locked. When I try to transfer cash from my bank in country X while in country Y I get prompted for PIN numbers which I do not have, because I do not have my phone number. I walk around clutching my credit cards, got forbid they are stolen. And still every couple of years I have to cancel cards cause my credentials are stolen, so its not like the security is so awesome.
When I started using argent (especially when they had daily-limits, which I feel will make a comeback with lower gas fees) I realised that it has none of the problems I had with banks. And also, I can get yields on my assets and stake straight from the wallet. Send money to charity? Easy! Send some ETH to friend so that they can play with it, or cause I owe them? Easy! No seed phrase to worry about! Wallet gets stolen? No problem!
In every way that I can think of I would prefer a social recovery wallet. I barely even have any funds on my Trezor (hardware wallets are awful UX. telling people to use them is a sure way to dissuade them getting into crypto. Even I, a power-user, take out the hardware wallet once every few months cause it sucks. And then there's the seedphrase, gah!). But the obvious elephant in the room is - gas fees. Argent even changed their transact/recover logistics cause a wallet that could be perfect for payment was never used that way, it was only used for long-term safe storage.
Back to Liberosist's write-up and zkSync. There is probably a long way to go in terms of speed and adoption of crypto payments. But if transactions get cheaper I do no see any use from my bank any longer - zero! I am even willing to pay money for L1 recovery. This would be the true unbanking of the banked.