[talawahdotnet]

Well said. Decentralization does not magically give you reliability or high availability. Sometimes it is quite the opposite. It is still just a bunch of software running on a bunch of miscellaneous computers over an unreliable network.

To me there are strong parallels between the cryptocurrency boom and the previous P2P boom. P2P technology was exciting and super interesting, but people assumed that it would magically be the solution to a million problems just because the tech was awesome and decentralized. But that didn't happen, instead Netflix happened. Netflix found a way to solve the actual problems that people were having and at the end of the day the underlying technology didn't matter.

I expect to see something similar happen with a lot of the current crypto use cases. More centralized solutions like Amazon QLDB[1] will provide the majority of the benefits that people really need: a transparent, immutable, and cryptographically verifiable transaction log, while also being performant and highly available.

Many people see decentralization as the solution to all problems because it will finally make things fair, a level playing field for all. I just don't see it happening. Not because of some sinister world order trying to retain control, but simply because that is not how human interactions and trust tend to work. I think we can continue to move towards more decentralization, but it shouldn't be the end goal in and of itself and we should always be skeptical when there is more focus on that than there is on solving the problem at hand.

1. https://aws.amazon.com/qldb/

[EthanHeilman]

>Netflix found a way to solve the actual problems that people were having and at the end of the day the underlying technology didn't matter.

This, one thousand times this, it is always this. People have problems and they want solutions, they don't care about particular technology architectures.

[CuriousSkeptic]

To be fair p2p was, and still is, a better technical solution, and had it not been illegal, would probably have won.

(p2p wasn’t illegal, but what people wanted out of it was)

[kabdib]

There are services that incorporate P2P (e.g., Spotify, and you can find others if you do some research). If you're writing closed-source software that operates on content that belongs to you (or that you've legally licensed) there's nothing preventing you from using peer-to-peer technologies in your stuff.

I can't believe that Netflix didn't seriously consider P2P. I imagine they didn't do it because there are issues with consuming limited upload bandwidth that doesn't belong to NF, and having poor control over the resulting complicated network flows and the ultimate customer experience (think about what your first customer service action for a customer having video quality issues would be).

No doubt there are many managers and system engineers who shy away from P2P because it's associated with bad actors. But there are sound reliability engineering and customer-facing reasons not to use it, too.

[chrischen]

Didn’t Netflix start providing ISP level caching vis boxes? In any case if they didn’t this method would make P2P pointless as consumer peers would still route traffic back to their ISP, and in worst cases another consumer could be at another ISP network.

[hndamien]

In this case it is a little stranger though. It seems that the problem is that the economic system is rigged against them, and their payment systems are spying on them. When the “solution” to this becomes evident it has been flowing to the participants in that economy. They then realize that Bitcoin was solving their monetary problem all along and the awakening happens - only when they rush to the solution it appears to be a mirage as the very act of them rushing on their realization turns into instant debt liberation for the long term participants. Then it all crumbles and they say it’s a scam and was never a solution. In reality, nothing really changed, and the slow crawl will continue for the long term participants.

[mynegation]

I am genuinely interested what are those problems that decentralized apps and cryptocurrency solve? Anonymity of payments? Bitcoin is a shitty way to pay for anything and I want convenience first and most of the time I do not care about anonymity at all. Smart contracts? After 2008 there was talk about including Python code into contracts to codify distribution of gains in MBS’s. Pretty low tech and workable. Did not happen, because finance is still largely a business of relationships and trust.

[EthanHeilman]

>Anonymity of payments? Bitcoin is a shitty way to pay for anything and I want convenience first and most of the time I do not care about anonymity at all.

Unfortunately Bitcoin provides much less anonymity than cash or credit card payments. Bitcoin provides excellent censorship resistance, in that Bitcoin lets anyone who has Bitcoin send Bitcoin to anyone that wants Bitcoin (assuming of course internet access and fairly technical users). Additionally Bitcoin lets people custody Bitcoins like cash or gold and also lets those same users send Bitcoins digitally. This is a new development and something people couldn't do before. We are still in the early stages of seeing how people use this new capability. Check back in 20-100 years.

Smart contracts, especially legally enforceable smart contracts, have existed well before cryptocurrencies, however cryptocurrencies provide a fairly natural setting for them.

>Did not happen, because finance is still largely a business of relationships and trust.

Relationships and trust mean a high cost of doing business, both because trust is expensive and because relationships are not able to be automated. Cryptocurrencies and smart contracts offer a technical substitute for trust. This substitute is not good in all cases and has some serious tradeoffs. However there are some areas in which it is a clear improvement.

[jjeaff]

I agree that cash is better for anonymous transactions ... Sometimes. But I don't follow that credit cards are more anonymous that Bitcoin.

You generally have to give up your ssn and lots of private data to get a credit card. Your name is tied to it.

Cash is great, but does require you or someone to physically deliver it.

But crypto can be laundered through a mixing service over a tor connection and when done properly and carefully, would be essentially untraceable.

[EthanHeilman]

>But I don't follow that credit cards are more anonymous that Bitcoin.

Credit card transactions are private between you, the merchant, the processor and your bank and anyone those parties sell your data to. Whereas Bitcoin transactions are recorded on a public immutable ledger.

>But crypto can be laundered through a mixing service over a tor connection and when done properly and carefully, would be essentially untraceable.

You can certainly obfuscate Bitcoin transactions. I even wrote a tumbling protocol for Bitcoin[0] and have been involved with other cryptocurrency privacy projects. I believe that cryptocurrencies can achieve a high degree of privacy, but we aren't there yet.

Note you can do similar things to make credit card payments more privacy. For instance by buying and trading gift cards or store credits.

At the end of the day if you buy physical goods online you have to have them delivered to your house, drop site or PO box. That means giving out a physical address which greatly reduces your anonymity.

[0]: https://eprint.iacr.org/2016/575.pdf

[xtracto]

> Bitcoin is a shitty way to pay for anything and I want convenience first and most of the time I do not care about anonymity at all.

As a side comment, just two hours ago I paid an IPTV service with BTC using coinpayments.net and it was easy (scan QR code and click send) and quite fast (at first I was going to pay for BCH because usually it is faster, but I only had BTC in my wallet).

Really cool as payments technology if you tell me.

[stale2002]

> I am genuinely interested what are those problems that decentralized apps and cryptocurrency solve?

Censorship resistant, electronic, financial transactions.

You can make an electronic payment to anyone, and nobody is going to stop you. No transactions are currently being censored on most of the big crypto platforms, so we have provable evidence that it works for this usecase right now.

[mynegation]

Thank you. So, electronic financial transactions is a problem already solved by other means (and much better at that). Censorship - I totally recognize my first world privilege - what are the contexts in which this is relevant? Even if payment censorship problem is somehow solved, the delivery of goods and services can still be blocked, be it by customs or firewalls.

[dmichulke]

> Bitcoin is a shitty way to pay for anything

I think following the reasoning of many BTC optimists, your quote is equivalent to saying "PPP (the one in OSI layer 2) is a shitty way to have globally distributed, secured real-time chats"

[shrimpx]

If people end up really wanting something like asset management that's not in the hands of a middleman, of course your Netflix and Amazon examples don't apply.

[talawahdotnet]

People want fast, cheap, easy ways to move money. Most don't particularly care whether or not a middleman is involved. They just want the job to get done and they want it to be simple.