Hacker News new | past | comments | ask | show | jobs | submit login
But what about blockchain apps? (heropunch.io)
36 points by xj9 on Jan 16, 2018 | hide | past | favorite | 52 comments

Blockchain apps aren't going to happen, just how BitTorrent apps never happened. There's a bunch of reasons, but most of them come down to the fact that decentralization isn't really a boon. Consider a new social network I want to build. Making a simple prototype and hosting it on AWS is easy. Getting people to use it is hard. Getting people to use it and make them self-host some kind of "node" in this decentralized universe -- that's exponentially harder. Diaspora clearly showed that no one wants to set up and host their own nodes. And then you have to solve all kinds of really hard problems: double spending, consensus, 51%+ attacks, and so on. Some of these problems are simply unsolvable and we have to compromise. SSBT, for example, has no consensus, so everything is subjective[1] (which makes it almost impossible to do anything useful with).

The people still fighting this war of decentralization remind me of the Japanese soldiers[2] who didn't realize WW2 was over. You lost. No one cares about decentralization. Google won. Amazon won. Facebook won. The great majority of people invest in BTC and ETH because they think it'll make them wealthy, not because of some noble desire to fight "the man." Besides, BTC isn't really decentralized anyway, the million-dollar-VC-backed miners basically control that ecosystem.

[1] https://github.com/ssbc/secure-scuttlebutt/issues/86

[2] http://www.cnn.com/2014/01/17/world/asia/japan-philippines-w...

You are technically correct...

> decentralization isn’t really a boon

Narrowly true. On its own, decentralization isn’t good or bad. In an ecosystem like ours, where most tools and services assume centralized architecture, it’s probably net bad.

But here’s a hypothesis for you:

In an ecosystem where decentralization was the default, decentralization would be a boon.

There’s a chicken and egg problem there, but I just wanted you to understand why I think decentralization is worth thinking about, even though I can see, as you do, that it is objectively the wrong way to solve almost any isolated problem.

I am building decentralized software, where there is no database, no global namespace, no source of truth, etc. My assumption is I’ll have to redesign everything, from basic authentication to employee workflows to accounting strategies in order for it to pay off. The reason I’m trying is I suspect it has very recently become technically feasible to take on a scope like that for a well-enough defined business logic. I.e. a “vertical slice” where all aspects of the business are radically decentralized. The downside being you can’t use any off the shelf software or development tools whatsoever.

If there is one thing we should learn from history, it is that making far-reaching predictions about the future of technology can result in feeling very, very silly.

I am reminded of that one IBM executive (I think) saying back in the day that no one would ever be interested in having their own computers at home.

In a similar vein, keep in mind that TCP/IP was invented in the 60s but did not start getting widely used until the 70s.

Email is decentralized and is probably the most used and most compatible system there is.

And yet what percentage of email lives on the servers of a handful of giant corporations?

But you can host your own emails. You can't host your own Facebook page.

>But you can host your own emails.

Sure, as long as you don't actually want to email someone else. AWS, DigitalOcean, Linode, anywhere you might want to host your email server, they're likely already blocked by all the spam filters so you're never going to get through.

>You can't host your own Facebook page.

You can't host your own Gmail either, but you can host something awfully similar to Facebook while not being exactly Facebook.

I've been hosting my own emails for 8 years and haven't had any problems.

If you host something "awfully similar to Facebook" you can't communicate with Facebook users. If you host your own emails, you can still communicate with Gmail users. Surely you weren't intending that retort as a serious argument?

The second argument follows the first. If the first isn't true, then the second also isn't true and it seems you've been fortunate in that regard. But if the first is true (that spam lists block addresses from shared hosting providers), then you can email but not to Gmail. Sort of like hosting Facebook without communicating to Facebook users.

Seems like a solveable problem.

Why would I want to host my emails there? There are things like proper servers.

Gmail is just the webinterface to Googles mail, I can host the latter, while using the former with a proper client.

I don't see why percentage is relevant here.

In that sense, the Internet is "decentralized," too. But most people don't host their own photos and don't host their own emails servers. People started mass-adopting the Internet precisely because of centralization.

Yes. But what "most people" do is not really relevant if you can do it yourself. Self-hosting is exactly why the Internet is great.

"The people still fighting this war of decentralization remind me of the Japanese soldiers[2] who didn't realize WW2 was over. You lost. No one cares about decentralization. Google won. Amazon won. Facebook won. "

Fuck no and I hope the real hackers (not the business people pretending to be hackers) fight this forever until they win. Decentralization is not dead, is the backbone of the principles of the internet, and I find your casual dismissal to be extremely miopic on the subject, and the subservient acquiescence of it disappointing.

Fuck google, amazon and facebook. They aren't the internet, and they aren't going to last forever.

Now, to back it up a bit, I do have some criticisms of what I call extreme decentralization. A relevant past comment of mine: "As a sysadmin, any time I read some variation of; "At some point even data centers may become a thing of the past.", I know they don't know what they are talking about. As a matter of fact it has provided much joy through laughter (followed by required sysadmin scotch) at the show "Silicon Valley" for obviously parodying the issue. Datacenters aren't going anywhere, and this strange fascination in hipster-hackers with instant uber-decentralization-pushes concerns me because it ignores some of the more real (and fixable) issues at hand like dns centralization in favor of magical "p2p(+blockchain) will save us all" thinking not backed by much real world practical implementation.

Don't get me wrong, I'm a darknet, meshnet supporter. I love decentralization. That said, I support the establishment of the infrastructure required to support it independent of end-user devices, and I think for security and other purposes it's at least possible they should remain separate, and devs shouldn't assume so much right to cpu-cycles.

So in essence the topology I think that is preferable would be properly called decentralized-distributed.

Of course thats part of the reason I support things that go against that common grain, such as ipv6nat."

The key point being that I think the proper terminology of the current internet is decentralized-distributed. Those who ignore the distrubuted part of decentralization are going to have a tough time, but in the same vein those who ignore decentraliztion in favor of distribution are also going to have a hard time.

This post will not age well. I can imagine the same tone of a post circa 1995: "Microsoft won. AOL won. Alta Vista won." It's true until it's not. I personally really dislike the dismissive and defeatist attitude. Enjoy it while it lasts I guess.

Alta vista was replaced with Google. AOL was replaced with Facebook. My point was that corporations won -- I think that'll still be true in 50 years. Sorry to burst your cyberpunk dreams of a crypto-revolution.

corporations won

yes, they won the current round of games. 50 years is a long-ass time dog. nobody can see that far into the future. i'm actually more motivated to continue on my present course because of this type of attitude. see, if the things you do and say piss people off, you are touching on something... important or sensitive. i'm triggering some shit, and i'm going to keep it up. please, tell me i'm wrong and foolish. tell me i already lost. that's the battle i'm looking for. i got nothing but time.

i'm not exactly anti-corporate either. i'm anti-"the current crop of megacorps" mostly because i see them as my competition. impossibly tough competition. i love hard problems tho. i'm a scientist at heart. i want the money, but only if i can get it by being obnoxiously clever.

...or something idk just not boring and broken no point in living if you aren't fighting

Besides, BTC isn't really decentralized anyway, the million-dollar-VC-backed miners basically control that ecosystem

yeah that's what i said.

SSBT, for example, has no consensus, so everything is subjective[1] (which makes it almost impossible to do anything useful with).

i disagree, there are a huge number of (fun) applications that don't need global (or strong) consensus.

You lost. No one cares about decentralization. Google won. Amazon won. Facebook won.

the war isn't over yet fam. i'm just trying to reclaim my homeland. don't forget that the reconquista took 700 years. if i can make something fun for myself that removes my dependency on centralized computer systems, i'll be happy. its a big universe, i don't have to kill the beast to build a nice rural community in the countryside. i just have to break its grip on myself and mine.

you should like, read the post tho. i'm actually arguing against blockchains and dhts because they don't scale well and require expensive internet infrastructure to operate.

> the war isn't over yet fam. i'm just trying to reclaim my homeland. don't forget that the reconquista took 700 years. [...] its a big universe, i don't have to kill the beast to build a nice rural community in the countryside.

Hyperbole aside (lol Reconquista), if you prefer to live on an island, decentralization is 100% fine. But be warned: my grandparents lived a very hard life farming and toiling the fields. It's not as glamorous as it sounds. Hobbes' bon sauvage is illusory.

i didn't mean like a literal rural community (ok maybe a little). the internet is a big place in any case. i don't have to "win" to create a self-sustaining community of users, developers, and other creatives. my goal is to make it easier for people to play with silly mesh tech. it isn't like a super serious startup looking to make a million bucks. i mean, we'll try to make money if we can, but right now we're doing research and exploring ways to make decentralized stuff less stupid, because current-gen p2p is kind of shit tbh

we will probably fail, but its whatever at least we'll have learned something and had some fun along the way

> Hobbes' bon sauvage


Common misconception. Hobbes was the primitivist, not so much Rousseau.

Decentralized doesn't mean every user needs to run their own node. Getting people to run a node is solved once there's a (financial) incentive for that, which is what made cryptocurrencies possible.

I agree no service is going to get adoption if its main feature is decentralization, but that doesn't need to be its value proposition. "Post your cat photos here and get paid in cat-tokens that are worth real money" would be the incentive.

This is actually an interesting idea. Does anyone do this? Sort of like reddit but you'd get paid (fake internet money) for contributing quality content.

steemit.com is one of the most popular dapps and it does exactly that.

Exactly, the example I had in mind. I tried their app and at this point I think the friction to setup an account, wallet etc. is too high for the average user, but this business model is awesome: great incentive for attracting early users, and if the platform ever becomes a hit, the financial rewards will be distributed among users, not just to the owner.

What about something like Steem/EOS? The general idea being that anyone can build a centralized UI layer that plugs into a decentralized database. I'm genuinely curious and have not invested in either of those cryptos.

I believe Veoh tried to be like a decentralized YouTube using a bit torrent like viewing client. But this added user acquisition friction because they had to install something versus just pointing a browser to a url.

Take a look at beaker. I'm not sure that decentralized apps wont happen. I agree that centralized servers have many benefits, but there are also benefits to giving control to users.

I had a look at beaker earlier today. It doesn't appear to do anything IPFS doesn't do, it just has a pretty user interface.

IMO it would be a much better project if it was IPFS-compatible so that it could benefit from all the existing work that is being done on IPFS.

Do you really believe that there's not going to be a market for TRUSTLESS wills?

For tax-free, trustless estate transfers?

Rich people just don't know that Ethereum does this, that it's built into it.

Everyone (Main Street) just thinks it's a crypto currency right now.

What about TRUSTLESS gambling?


Maybe there'd be a market for that?

We could literally set up a company that does this for rich people right now and charge them a percentage of their estate. Some people are so untrusting that they would pay a percentage of the estate for such a service. Not to mention the tax advantages!

Now, the high volatility of crypto is an impediment to this model, but after the blood bath (possibly starting today), values will hopefully stabilize a bit and make the values a little more stable and predictable.

How does a mechanism change tax rules?

Everytime you park the crypto in fiat, you face tax burderns.

The Etherium Network imposes no such burdens.

Depending on where you live... it still does. It's just hidden. No try buying a house with those ethers and get away with it.

So your libertarian utopia of cryptocurrencies being uncontrolled has now become an instrument of former off-shorers to hide their dollars/euros/yen/kittens.

Yeah, yeah, we'll write all that stuff in the fine print.

I am seeing this:

Surge protection

<strong>Warning:</strong><p>You triggered the wiki's surge protection by doing too many requests in a short time.</p><p>Please make a short break reading the stuff you already got.</p><p>When you restart doing requests AFTER that, slow down or you might get locked out for a longer time!</p>

Google cache: https://webcache.googleusercontent.com/search?q=cache:qGQvzK...

i'm being hugged to death! maybe i should add come caching to my nginx config

AFAICT, with current blockchain technology, in order to process PayPal level volumes, every single participant in the network would need to process PayPay level volumes... 1,000 transactions per second is not a problem for a single payments company; it seems like a giant problem for each of the 100,000 nodes on the network also to process 1,000 transactions per second.

Certainly I am missing something obvious?

This is one of the biggest problems with Blockchains right now. If it cannot be solved then their utility is extremely limited. There are two main approaches that seem likely to work:

- "Layer 2" networks: Most transaction volume happens off-chain. POW blockchain remains as a settlement layer. Quite a few proposed protocols. Lightning network [1] is the most well known. First mainnet lightning transactions were sent in December [2]

- Sharding: Split the blockchain into lots of different blockchains with a subset of miners validating transactions on each sub-chain. Miners are randomly assigned a new sub-chain on a regular basis to stop collusion. Each sub-chain has it's own address space. Sub-chains can communicate using asynchronus message passing. Ethereum has a draft specification [3][4] and is starting to implement some test clients [5].

[1] https://lightning.network/

[2] https://twitter.com/alexbosworth/status/946175898029395968

[3] https://github.com/ethereum/sharding/blob/develop/docs/doc.m...

[4] https://github.com/ethereum/wiki/wiki/Sharding-FAQ

[5] https://blog.ethereum.org/2018/01/02/ethereum-scalability-re...

There won't be 100,000 full nodes; more like 1,000 full nodes and 99,000 light nodes that only verify their own transactions.

I had a real hard time reading this post without being disoriented by the lack of capitalization.

Yes, this is a stylistic choice some bloggers are using these days and it's pretty annoying. I'm all in favor of learning how to break the rules but in this case it just makes it harder to read without adding any benefit.

I ran into a grizzled veteran on our company's IM platform who refused to use anything but properly capitalized and punctuated sentences inside of the IM chat. He converted me.

i'll do the next one in all caps ;)

I agree with some of the authors points. One item that is missing is the incentive structure.

Let's take mesh networks for example (as mentioned in the article). Substratum pays users in coins to share their network, which creates an incentive to grow the network (as long as the coins are worth something). If the coins are worthless, the model doesn't work.

It would take a lot of money to create the incentives without the coins.

DAG (Directed acyclic graph) based currencies don't need a blockchain; Byteball/Dagcoin, IOTA, Rayblocks (Nano) and others.

Not sure I entirely understand where this post is going. What does centralization versus decentralization have to do with a blockchain? Just because Bitcoin and others are using it as a centralized data store for a decentralized currency doesn't mean the blockchain has to be tied to any of that.

I feel like people keep getting really far away from what a blockchain really is (a fancy linked list that can ensure integrity from start to end) and are confusing it for how Bitcoin uses it.

I read it as:

Blockchains(PoW/PoS/etc) are not the solution for decentralized applications because blockchains are a global state and most decentralized application don't _need_ a global state. (Specially non-fintech ones that don't have a double-spend problem like social/websites/photos/etc)

It briefly mentions Secure Scuttlebutt, a gossip protocol that works over meshnets without global internet access as an example of a decentralized technology without a global state.

Surge protection warning

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact