[SanFranManDan]

> trustless decentralized internet

That is the solution. What is the problem? I understand the problem they want to solve but I don't think that is truly the problem. I drank the kool-aid for over a year before I realized that the theoretical and practical realities are incompatible.

A truly trustless decentralized ledger is for the most part useless since as soon as you need to interact outside of the system (like change real world goods). Registering ownership on Ethereum and being trustless cannot coincide as ownership is a societal construct backed by governments and consensus. If I register stock on ethereum, I still need trust in the societal agreement outside of ethereum. Only transactions are trustlessm, but it is still very very far from clear that trustless transactions actually mean something without trustless assets that can be run on a truly trustless network.

I think this is evidenced by the fact that there are no compelling real world use cases for ethereum. There has been a huge downturn in turnouts at ethereum events and mindshare coinciding with the DAO attack after they shot themselves in the foot by doing the manual rollback which showed you still need trust in the network that runs the "trustless" system thus there is no such thing as a trustless system.

Yes it is still young, but I don't believe the theoretical case of having a trustless system (ethereum) interact with a trustful system (human society) while still being trustless is solved or even possible.

As soon as trust enters into the equation at any level the entire premise of ethereum is worthless moot. Could there be a narrow domain of applications that can use a "minimal trust network"? Yes. But that is different from being truly trustless.

Trust is a feature not a bug. Are there situations where you want less trust? Absolutely. But you are shifting the trust to another agent, not getting rid of it completely. Ethereum can exist, but it won't transform society and replace governements like everyone was hoping to ~9 months ago.

[heliumcraft]

Well.. for starters there is certaintly value in a decentralized infrastructure (just ask wikileaks or someone living in a dictatorship) as well as other interesting projects such as makerdao, golem, etc..

I'm not sure what downturn you've seen since ethereums devcon this year had over 700 participants attending despite being in China, and there is another conference next month this time in europe, so if anything it keeps growing.

The DAO was too much, too soon. you say "they" which I'm not sure who you mean, but please note that ppl behind the DAO are not the same as the Ethereum foundation, consensys, reuters, deloitte, santander, etc.. let's not conflate one group and their mistakes with the entire Ethereum comunity.

As for oracles, they are an interesting problem to solve. Note that not every project would necessary require outside interaction (such as a decentralized DNS), sometimes there are creative solutiosn to do certain things externally (http://www.ethereum-computation-market.com/), and there are ways to verify the validity of the data a oracle is providing, as an example you can watch this talk https://www.youtube.com/watch?v=9pDUobV8geI

[alistproducer2]

I feel exactly the way you do. I was initially seduced by Ethereum. So much so I actually read the white paper...all of it and started an Eth meetup. After a few meetings of being the only skeptic in the room, I realized the community's problem is not the tech - it's the community. The cryptocurrency world (at least the one I encountered) is a huge echo chamber filled with people who genuinely believe they are smarter than everyone else. It also only attracts a certain type of person so there is little diversity of thought. For example, there are basically 0 people in that world that think it's necessary to dumb UX down (like hiding crazy hash addresses and not denominating things to 13 decimal places) so it can be used by non-technical folks.

Eth's problem is exactly what you said: it's a solution desperately in search of a problem. I believe there are legitimate reasons to displace decentralized systems, but the market just isn't there yet. The value add just isn't there for anyone but a specific kind of nerd.

[DennisP]

Actually the Ethereum Name Service was launched recently. It's been in the works for quite a while, but took some time while they worked out a way to dissuade name squatters. The idea from early on was that most people would mostly use names instead of raw addresses.

The fact that things are a bit raw and technical right now is a function of the whole system being less than two years old, and still rapidly changing. But it's already at a point where an Ethereum app can have a nice web UI, and all you need to use it is a Chrome plugin and a little ether.

[Jd]

Same boat here. I co-founded a few Ethereum related meetups, including the Silicon Valley one. I think one of the big problems here is that a lot of the "business thinking" type folks are pushed to the side, as well as some brilliant but non-standard tech minds.

I notice that a few times when various people (including myself) raised critical feedback we were marginalized. I know a few brilliant people who were kicked out of various Ethereum related stack channels for non-conformity, for example.

[stale2002]

OK, fine, here is a usecase.

What about online gambling? With a trustless system, you can now prove that the game is fair, and isn't going to up and steal your money.

Gambling is frequently thought of as a "shady" business where trust matters a lot.

There is no interaction with real world goods. (the exchanges handle that).

[gertef]

Yes, a completely fair game for either "games of chance", or "games of robots skill". That's great for your poker-playing bot, but not for human endeavors (except bot programming).

Ether can provide some measure of server/network-side honesty, but makes no guarantees about what happens on the client side. So you must assume that all the other clients are super-powered bot-enhanced adversaries, or you need some way to trust and verify the client-side behavior of your peers.

The robots have come for oru jobs, and now they have come for our play.

[stale2002]

I meant more like blackjack where you are playing against the house.

The vast majority of gambling is not poker. It is slot machines or blackjack.

Client side stuff wouldn't matter for gambling Vs the house stuff (which is what most gambling is!!)

[gertef]

Pardon my moralizing, but I strain to find a situation where someone's life is improved by easier access to slot machines or blackjack games, even provably fair ones.

[openasocket]

How do you prove the gamling is fair? I know little about Ethereum, but how can you make the outcomes of games verifiable? Either you use a RNG, in which case the house can use some RNG biased in their favor, or the program in completely deterministic, in which case the player can compute in advance the permutation of the cards and play perfectly.

[zeroer]

Here's an example: Let's say you have a card game with N players playing off a 52-card deck. Since log_2(52!) is less than 226, then you can randomly shuffle a deck of 52 cards with, say, 512 random bits.

The dealer picks a random 512-bit number, X_0, and publishes a hash of X_0 to the players.

Each of the N players, picks another random 512-bit number and publishes it. These are X_1, X_2, ..., X_N.

The dealer xors X_0 through X_N to get a random seed to shuffle and deal out the deck.

At the end, the dealer publishes X_0 so the players can verify the dealer wasn't cheating.

[heliumcraft]

There are different ways, one is simply to do it in the blockchain itself , which is completly transparent so you know exactly what is going on, however this suffers from the flaw that the miners could potentially manipulate it (if the incentive is high enough), some proposals include doing a type of DAO in which participants commit a random value (hashed and revealed later). There are also approaches like the one used by satoshi dice https://www.satoshidice.com/provably-fair/?stage=1

[wyager]

https://en.wikipedia.org/wiki/Provably_fair