Web3 – A Vision for a Decentralized Web
blog.cloudflare.com
blog.cloudflare.com
So why does all talk of Web3 always shoehorn in blockchain and smart contracts? Who cares about that besides crypto buyers, who are mostly conducting their business via clearweb exchanges anyway?
The best part of Web 1.0 era was that you didn't have to pay to access information beyond what your ISP charged. Ethereum and its dApps are pay-to-play. I'd be interested to know how many of the cited 7000 dApps offer free information, and not just overpriced digital trinkets.
Nobody - those are exactly the people who push "Web3" (which has little to do with the web) as they're in desperate need of a Greater Fool in order to realise their "investment" in cryptocurrency-related assets.
That alone is enough to make me feel sure of my investments. However, there are many beautiful things that will come out of Web3.0 as well. End to end open source VPNs, containerized and running on a decentralized cloud is already a reality. You an use it today for free on Android and IOS (Velocity VPN - Running Sentinel back end.) Its open source and anonymous end to end. No one is keeping logs.
Don't look at the shit coins and the bubble and think the dream of Web3.0 is not real.
That's a little misleading, the app says it is "free and unlimited while in its testnet phase". Presumably it will require some cryptocoin eventually.
Are you saying sentinel is open source, or that Velocity is open source? Because I can't find any velocity repositories on the web, nor is the app on f-droid.
My guess is that number is zero.
It would make no sense otherwise.
No reasonable person would open a classic book store on Rodeo Drive. What's the difference?
I wish the NFT and blockchain enthusiasts better luck with the term than I had.
Blockchain is open by default. So all information is free. You can see who owns the trinkets, and the entire history of the trinkets from inception. You can't claim you "own" that trinket, but the information is free
Blockchain is getting the heat social networks should have up front. There are sufficient bad actors that blockchain will be dangerous for most people's financial well being, like social networks are for most peoples mental well being.
This is misleading. Very little on the internet is actually free. Right now (and back then) there is tons of content that is either paywalled or supported by advertising.
I think by-in-large most people consume less "usage" than what they pay for with subscriptions, etc., and generate much more revenue via advertising for the sites they visit than the cost of the bandwidth they are consuming. Imagine if instead of being bombarded with ads or subscription CTAs, you are simply charged $0.00001 for your visit to [insert cool blog here], or $0.0001 per second of video you watch on [awesome youtube equivalent here]. That is the dream.
It is? That sounds horrible to me. I don't want to have to constantly think about the ticking price meter every time I open a video. It also screws over anyone without means, or the developing world. The paid-for-by-advertising model lowers the barrier to entry and allows anyone, even those without change to spare, to participate.
I dislike ads, especially in their current state, but I dislike the "dream" you're describing even more.
How would that work? Blockchains are far too inefficient to host video files and nobody is hosting that much data for free so you still need to set up a paid hosting environment or learn why P2P video hosting has failed every time it's been tried in the past. You can charge people to host their video, at which point you'll learn that it's really hard to compete with ad-supported hosting because the number of people who say they want to pay up front for their stuff and actually do so is a rounding error of the number of users a major video site will have.
Many NFTs reference third-party hosting services for this reason (that's the real service; the part on the blockchain is the expensive vanity link) and anyone using IPFS for real will need to pay for ongoing hosting if they want their content to remain available.
Part of what dictates this will be abuse: if you provide free hosting to strangers on the internet, they will exhaust your capacity and some will try to host material which violates copyright or other laws. Over time, anyone not getting paid to deal with that will stop offering free hosting to random strangers on the internet.
It all went to hell in the 2.0 era when DoubleClick and Google introduced Real-Time Bidding and serving ads dynamically from their servers.
I’d guess it has nothing to do with that and way less dramatic.
They are a company that benefits and profits from having a large network. It makes sense to try and be part of alternate networks.
A lot of people, including myself, see their growth as also having made small businesses/sites/apps more competitive on an increasingly centralized web from way more evil players in the game.
The Crypto noise aside, I also think a decentralized Internet is exciting. At the very least, I’m glad they are throwing money at and will see where this goes research helps things go.
They need new faces and the most obvious source of them en masse is people who aren't welcome on other cloud platforms. That means cryptocurrency people, who have been effectively outcast from the rest of the internet's premium tier hosting providers, because when they were allowed on the rest of the internet's premium tier hosting providers they tried to steal shit while they were in there.
unfortunately not, as this was on the front page for a while
Cryptocurrency mining using integrated photonics (tue.nl) - https://news.ycombinator.com/item?id=28715881
I have one of these.
It's called my computer's local files.
It runs locally and I can even access my files when the Internet is down.
If all your stuff is DRM-locked on someone else's computer, it is not your stuff.
If your stuff is on someone else's computer, it is also not your stuff.
My money that is stored in some bank's DB is not on my local computer, but I sure hope it's mine. In some countries they aren't so lucky and it's exactly as you say, the money doesn't belong to them.
You have a contract with your bank which says that they owe you a certain amount of money based on your past interactions. That contractual claim against the bank is what you actually own. The number in their database is just a summary of those interactions. Just see what happens when the bank messes up and enters a larger number in their database than what the contract says they owe you: The DB is not authoritative; the contract is.
Physical possession is not, of course, a requirement of ownership, though it certainly helps.
There is also a difference between owning specific property, such as the contents of a safe deposit box, and an entitlement to be paid according to the terms of a contract. In a very real sense, until you make that demand for a withdrawal in accordance with the contract the actual money belongs to the bank, and as a depositor you are merely one of the bank's creditors, just as you do in fact own your home even while it's serving a collateral for a mortgage and could be claimed by the bank if you fail to keep up with the payments.
You can chose to let others do your computing for you, but don't be surprised when they want to charge you for it.
>> My money that is stored in some bank's DB is not on my local computer, but I sure hope it's mine.
I trust my bank to hold my money because there are legal protections in place to protect against bad behavior. I can sue them if they violate the laws that protect me.
What can be done if someone steals your cryptocurrecy?
Computing for yourself isn’t free either, you just paid the bulk of the costs up front.
One of the areas of decentralized computing that’s actually interesting to me - and I mean this in the idealized version, not whatever cloudflare’s thing is - is the idea of being able to “burst” compute when I need additional power. The fixed cost of having a personal machine capable of anything I might want (note: I might want - I’m an outlier, I know most people’s needs are met by the most rudimentary toaster available) is remarkably high compared to the cost of my average compute needs - being able to “rent” the excess could be both cheaper and less wasteful on the mean.
True, but I control the computers in question. The decentralized web is more a question of who controls what and what they can do. Cost does definitely come into play, but to a lesser extent.
>> the idea of being able to “burst” compute when I need additional power. >> The fixed cost of having a personal machine capable of anything I might want is remarkably high compared to the cost of my average compute needs - being able to “rent” the excess could be both cheaper and less wasteful on the mean.
Again, the larger part of the argument is control and decentralization. I completely agree that cloud computing is useful, especially when you have computing needs that change elastically and you are aware of the limitations of renting storage space or computing time on someone else's computers.
My concern about Web3 computing is where my local computing capabilities are reduced and I would need to rely on hundreds of computers owned and controlled by other people to get capabilities that are not worth the opportunity cost to me.
Not much, but at least you don't have to:
* ask a bank's permission, or
* divulge trade secrets, including private investment strategies, to a bank's employees, to meet the bank's AML "source of income" disclosure requirement
To gain access to banking services. Truly owning your own assets, with no dependence on others, has its benefits.
I don't think it would be that easy in third-world country or dictatorial government or country in wars
The concept of "mine" is not well defined in money. Perhaps the proof of having a debt can be in your hands, but the debt itself (=money) is a contract between parties, and owned as much by you as the society.
So I'm not really sure what point your example illustrates.
DRM only works if you run the code that implements DRM.
My copy of the bits are not impacted by someone's claim of ownership nor can their ownership token stop me from accessing my copy of the bits.
Artificial scarcity is artificial.
No I don't think so at all, but that's only a sliver of what is being discussed.
> My copy of the bits are not impacted by someone's claim of ownership nor can their ownership token stop me from accessing my copy of the bits.
> My copy of the bits are not impacted by someone's claim of ownership nor can their ownership token stop me from accessing my copy of the bits.
> DRM only works if you run the code that implements DRM.
Agreed.
-----
Here's an example:
You bought a digital item, say a sword used in video games. Yes you can keep that image file of the sword and do whatever you want with it. That's only a tiny piece of "owning" that item.
Imagine you could take the sword and bring it into lots of video games. The properties of that sword having value in those games can only exist with digital scarcity.
Now say you want to sell merch, t-shirts with your sword printed on them. You can own the copyright to your item, so it's legal and protected.
Again, "ownership" means more than access.
If you don't like this or care about it... that's fine. But it's unlocking a massive amount of capability that's never existed before. As a personal who likes tech because it's an enabler... it's exciting
It's still artificial scarcity. Unlike real swords, forging bits into magic swords requires no real work and requires no raw materials. I can make hundreds of similar digital magic swords with little effort, so there is no real scarcity.
>> Now say you want to sell merch, t-shirts with your sword printed on them. You can own the copyright to your item, so it's legal and protected.
Are copying, distribution, and derivative works rights included as part of the purchased digital goods or would those still be held by the creator?
"On the one hand, the transfer of an NFT associated with a work of art or other copyright-protected work would not itself involve the reproduction or distribution of a protected work, consistent with the first sale doctrine. But, if the new owner of the NFT accesses the underlying work, and this access involves the creation of a new copy of the work or distribution of it, then the transfer of the NFT may potentially fall outside the scope of the first sale doctrine and create potential copyright liability.
Despite this doctrinal murkiness, many NFT sales agreements appear to skirt the first sale issue by explicitly providing that NFT buyers have the right to resell the NFT. Some sales agreements go further and provide that the artist is to receive a set percentage of resale royalties. If NFTs remain a fixture of the art market, courts may be forced to decide whether to fashion a digital first sale doctrine that is responsive to new technological developments, including the rise of NFTs." Source: https://news.bloomberglaw.com/ip-law/what-copyright-lawyers-...
It seems that the rights of NFT buyers with regard to copying, distribution, and derivative works is not legally settled yet.
Yes, it's intentional scarcity. Why is it assumed that's a bad thing? When we license our code, there's a choice to limit it's use or not (intentional scarcity). The same concept hasn't been available for millions of creative professions in the online world (unless you go through a expensive middleman to manage rights).
Artificial scarcity is not necessarily bad, but it is not realistic on its own.
The amount of piracy on the Internet shows that artificial scarity alone is insufficient.
>> When we license our code, there's a choice to limit it's use or not
Yes and the choice to impose artifical scarcity or not impacts how widely the code is used. Code released under unrestricted licenses is more widely used than artifically restricted code.
This is where Web3 comes in. The last two decades have proven that building a scalable system that decentralizes content is a challenge. While the technology to build such systems exists, no content platform achieves decentralization at scale.
There is one notable exception: Bitcoin.
The transaction processing capacity maximum estimated using an average or median transaction size is between 3.3 and 7 transactions per second.
According to the Cambridge Center for Alternative Finance (CCAF), Bitcoin currently consumes around 110 Terawatt Hours per year — 0.55% of global electricity production, or roughly equivalent to the annual energy draw of small countries like Malaysia or Sweden.
I'm guessing that these problems get brought up a lot, but as it relates to this article, I wouldn't build my argument on Bitcoin being a pillar of scaling. Perhaps virality, but I'd argue that has more to do with financial speculation than solving end-user problems.
This is what Twitter is in the process of implementing through Strike.
And sorry if I don't trust the self-described "trust-less" money developers to get batching right on the first try. The only reason this isn't being exploited is because no one is using it.
Agree with you 110 Terawatt Hours per year is likely too much, even though it's `less alarming than you might think` (quote from the last paragraph of HBR article). Core of the post was not much on the usage of Bitcoin rather than the trend towards decentralisation. Resource usage of proof-of-work based blockchain is definitely a discussion for a later post though.
I also invite you to look at the source for this number (https://cbeci.org/cbeci/comparisons), which is nuanced, and makes the distinction between electricity and energy consumption.
All of those are truly decentralized with a wide range of implementations and operators, not to mention proven real-world robustness as opposed to Bitcoin's hard requirement on a massive amount of always-on hardware and network capacity.
You're right that there may be better examples but how much of the world's email and DNS content is served by Google/Microsoft, I wonder. People are also complaining in other threads that Cloudflare is actually making the internet more centralized.
Maybe they're talking purely about dollar value, but I'd say more useful information has been passed back and forth with torrents.
That decentralized software aligns much easier with commons than with markets is something that I would consider a feature, not a bug.
We’d definitely have the most comprehensive and accurate catalogue of the musical output of humanity ever created.
Instead now we have Spotify and Apple Music. We’ve only scratched the surface of the good that torrent technology can do.
You're litterally sending this message accross a decentralized system that outscaled bitcoin by many many orders of magnitudes in a fraction of the time.
I don't think cryptobros and the rest of the world have the same definition of "decentralized" because clearly we don't understand each other
This is true of most of the common sales points: for example, you'll see people claim blockchains are censorship resistant or anonymous, beneficial to the global unbanked, etc. with little ability to explain how that's actually true in practice.
And I often tell people that my goal is that if you hear about an Internet technology and Google it you should find that Cloudflare has already written about it and/or implemented it. That way people can trust us as a valuable source of information and to be on top of a changing world.
Imagine a large customer of Cloudflare who's heard about web3. Lo and behold we've written about it, have a product offering, etc. Maybe they read all that and say "Nope, don't want web3". Doesn't matter, they still trust us to be their partner in whatever's next.
You have a point about proof of work in general though.
Which means everyone mining Ethereum can't switch to Ethereum classic, or it would no longer be profitable for anyone.
Spoiler alert: it won't happen.
(yes, I'm an employee)
I have almost always been impressed with Cloudflare's blog posts (and documentation). Good and explicit marketing, clear language, interesting and smart products. Just really attractive stuff that demonstrates competency.
This one feels like a mumbo-jumbo of cryptocurrency buzzwords that no one needs or even wants to hear. Just thinking about the term "Web3" pains me to the bones.
I guess I don’t understand the standard for etiquette.
One of the first comments of an announcement by Cloudflare is by an employee of Cloudflare publicly calling it “peddling bullshit” with no additional context to what or why.
It’s not fair to the authors and the people who are working on this.
Calling another researcher and team’s work “bullshit” in a forum highly trafficked by the CEO and CFO and then saying you will raise it internally, is not “raising it internally” but “slandering it publicly” and flexing your high value / can’t touch this employee muscles at everyone else. It’s a giant middle finger to the company and teams from a jerk employee.
Cloudflare is one of my favorite companies to follow, but I had no idea how toxic it was working there for some. Definitely enlightening.
Blockchains are way worse than that. I think PoW cryptocurrencies are way past any benefit of the doubt, are actively harmful on a large scale, and any promotion of them is shameful. Cloudflare crossed the line here.
For Cloudflare this may be just yet another marketing and SEO piece, but it is giving legitimacy of a large tech company to the blockchain nonsense. People researching the buzzwords are going to see that a successful company powering a big chunk of the actual Web says "Web3" and NFTs are legitimate. This blog post is going to end up in a thousand pitch decks for more buzzword-salad ponzi-scheme tokens, and drive even more coal-purchasing fiat money into the massive waste of energy and silicon.
The future of the Web may very well be in p2p and content-addressable protocols, and Cloudflare is likely to be exploring it, but it's absolutely irresponsible to even suggest PoW blockchains to be any part of it.
Oof, your dev god complex...
This isn’t complicated or hard to break down. Nothing you just said and nothing about the technology is an excuse for your behavior.
I have seen many great devs but the best ones are humble and nice. You my friend are a clueless jerk.
You owe the authors an apology:
- Thibault Meunier
- In-Young Jo
And everyone else on the team at Cloudflare. I can’t imagine this is the first time has been difficult or confrontational working with or around you.
What is “bullshit” is you able to dance this disrespectful behavior publicly in front of eastdakota and jgrahamc without repercussion.
It is not fair to all the awesome hard working people at Cloudflare.
Edit: This isn’t my fight though. I am done commenting.
Are you an employee? Or, speaking generally here to a company this community commonly roots for?
I guess either way. What’s your beef with them experimenting to try new and wild ideas?
Edit:
> (yes, I'm an employee)
If true, wow. If not, troll…
Publicly trashing another co-workers work/team is in poor taste, immature, and creates a hostile work environment — if you agree or not with the tech or direction. I hope you get what’s coming to you. Classic Dev God ego and suggest maybe disconnecting online a bit and apologizing to them. I can’t imagine Cloudflare’s hiring team likes to see this type of hostile banter allowed.
Only "wild" as in "stupid". Ethereum has been around for ages, and it still has nothing to do with "decentralizing the web". Its advanced functionality is only used for building things like automated Ponzi schemes.
Decentralization is great, but what is even better is if you host all of your decentralized projects on one central platform.
If some people host content on GCP and others Amazon is it suddenly decentralized?
Now, to actually add to the discussion, it would be great to see clearer use cases that Cloudflare has to help publishers and users of web3, as the blog post describes what’s behind the name but goes only very briefly on what Cloudflare actually wants to bring to the table. A very fast, yet privacy-focused or anonymous, and affordable decentralized CDN would be interesting.
* DHTs are "invented" at Berkeley/MIT. Online storage is too expensive for most end users, so the target market is helping large companies index data in their filesystems.
* A bunch of content distribution protocols get built on top of the concept, the most popular being Bittorrent. Torrent protocols run into issues with the free rider problem, resulting in slow downloads. Meanwhile, storage on the web is becoming cheaper and cheaper.
* To solve the free rider problem, IPFS was created. A brilliant incentive structure was created so that asset hosting no longer relied on the benevolence torrent seeders. Meanwhile, companies like Cloudflare, Mega, Google and Amazon make online storage essentially free.
* Cloudflare starts hosting IPFS assets for free. Decentralized storage still exists, but is still slow in comparison and, in IPFS's case, is more expensive than free, so people only use IPFS addresses to download assets from large, centralized services.
* IPFS is mostly used as a distributed network of hashed addresses that can be used to look up data in a large company's filesystem.
Integrate that with Cloudflare registrar to setup DNS automatically such that it's proxied to the old Web through the Cloudflare IPFS gateway, and I can host my entire site on Cloudflare. The current one-click push to deploy integration is key, and I'd like to see that for IPFS as well.
They could use the same edge servers as IPFS nodes, offering the same benefits of DDOS mitigation and such. Though as is the benefit of IPFS, they'd receive less traffic since other IPFS nodes might pin the content as well. It'd probably cost them less, since currently each page revision is hosted indefinitely at <rev>.<domain>.pages.dev which would be the same as keeping the old versions pinned on IPFS.
Today's web have an obvious problem - it is run by oligopolies.
CF/AWS and Google don't like you (like with Parler case or recent youtube shields on discussing elections on several countries) and you have a problem.
Torrent-like static distributions (IPFS) with distributed backends like Ethereum basically broke this oligopoly - even US gov wouldn't be able to shut down truly "web3" service, which is pretty cool, like in the old days when you could setup a server in third-world country. These days are different - there are state backed DDOS gangs in town with an INSANE amount of resources. You are banned from several CDNs, Google and AWS - and you are done if they want you to be shut down. For example, just a week ago Russian "elections" have happened, opposition sites were under complete ddos the whole time and Google+Apple just deleted opposition apps from the appstore after the small talk with gov guys, so there is no trust in big tech - they are not interested in anything but moneymaking.
Yes, crypto has a lot of scammers and bastards but those bastards would setup ANY website for you for one another coin. Any torrent, any porn payment processing (hello Christian fanatics and blocked card payments on pornhub), any content distribution (hello Apple in China and blocked Hong Kong flags !). Maybe that market-based approach is better if those "scammers" broke censorship and white-suite Silicon Valley boys prefer to deploy censorship mechanisms because that will make them another investment round without a scandal. Maybe censorship-free global distributed backends for critical apps are more important than a few bucks stolen. Freedom comes with a price.
Classical stack: User pays company via a third-party payment provider. Then the company pays their infrastructur provider via another payment provider.
Web3 stack: User pays infrastructure provider via the infrastructure's own payment system. User pays company via infrastructure's own payment system.
I worked with Stripe, PayPal, and whatnot. Some are simpler or harder to integrate, but none were really a joy to work with.
Solidity and in turn the Ethereum platform makes this really nice, since the infrastructure costs are always paid implicitly, because every transaction costs money.
Based on my observations (ETH gas) every transaction costs a ton of money and in a lot of jurisdictions you have to track capital gains / losses based on fair market values whenever you participate in a transaction.
I’m sure it’s easy for Cloudflare and other big companies to play around with crypto protocols because they can send the complicated financial end of it to their accounting department. For normal people like me it’s insanely difficult to deal with.
Cloudflare is, like, one of the most centralized parts of the web. How many sites/apps route all of their traffic through Cloudflare? How many TLS terminations do they do? How many DNS zones do they control? How many crosswalks do I have to find over and over because of them?
They seem to be going quite long on centralization.
[1] https://www.inc.com/jason-aten/cloudflare-outage-takes-down-...
Should the web be distributed? I'd like to see a good distributed search engine for instance. I know there are a few peer to peer ones already, although the content isn't high quality and there is a need for a way to deal with the high volume of spam.
Streams at a quality I haven’t seen from any broadcaster over the internet, comparable to 4K 60fps cable AND IT GETS BETTER, THE MORE PEOPLE THAT JOIN!
A truly decentralised social network for content is the stuff of lore. There are so many things to be said about use cases that arise from infrastructure such as this, it’s impossible to fathom the potential.
Do we really need bulletproof consensus on Wikipedia edits? Let alone your blog...
Idk what "Web3" will be, but the future of the internet is a tolerant(fuzzy) mesh network.
Although there is a bunch of overlap via DIDs and such.
Heck, one possible outcome is blockchains become free databases for the entrenched providers, paid for with customer bandwidth and power.
Honestly I think something like this is what will happen; we saw the original decentralized web1.0 as democratizing access to information, but in reality economies of scale have created Facebook and Google - the complete opposite. No reason to suspect this isn’t just the way of things.
(In particular, there is no mainstream way to keep an important private key safe enough to use as your primary identity. Cellphones break, and even Yubikeys are a niche.)
There’s also no real announcement to this blog post, or enough meat to it for “Web3” to catch on. What’s the real purpose? Is it just about hiring?
Or what to take of:
- captchas so things like recently discussed marginalia.nu can't easily index sites behind CF?
- per-country restrictions for visiting their customer web sites (stopping legitimate users from visiting parts of the web depending on their current location, but also reducing the value of VPNs)?
Basically, coming from Serbia, I've been stopped from visiting US-based web sites by CF increasingly commonly in the last few years: they make this all too easy for their misled customers (companies thinking that their US-based-businesses might not get significant business outside US).
And sure, each individual company could block visitors based on GeoIP data itself, but if this was not a few clicks, they would probably not bother. A large player like Cloudflare thus provides a net negative for the decentralisation of the web.
ssh like crypto works. If you want to bring in blockchain, do it later as an added layer.
In what way does this alert you that some fundamental thing is wrong with the company?
Overall bullish, but you make money by taking profits. Buy the rumour, sell the news.
if Cloudfare is replacing CDN keyword with Web3, may be it's fine.