23 highlights from the Ethereum core developer Q&A at EDCON 2019
Why is Vitalik Buterin watching BCH? How about those zk-Snarks? Where's Ethereum going next?
EDCON 2019 in Sydney, Australia, featured a Q&A panel with Hsiao-Wei Wang, Vitalik Buterin, Virgil Griffith, Danny Ryan and Justin Drake.
The forty-five-minute session brought a range of insights.
23 lessons learned
The highlights are distilled here, but make sure you watch the video to get the complete picture.
1. It took years to figure out whether proof of stake was even theoretically possible
As far as anyone knows, the first public reference to proof of stake was in a 2011 Bitcoin Talk thread. But no one was sure the concept was even feasible. Vitalik Buterin started really diving into it in 2014 and came up with a solution... which didn't work.
"The first version of proof of stake I came up with... it technically was not even secure. It had liveness failures." Buterin said. "Since then, we've just done a huge amount of research and figuring out that proof of stake actually is possible, and with what trade-offs."
It crystallised a lot over the coming year, and in 2015, Buterin suggested a more workable model that was much closer to what Ethereum ended up going with.
"In 2015, I wrote this big, huge paper on scalable blockchain design which ended up being pretty close to what we've implemented at a super high level, but there are a lot of details we needed to figure out."
2. Ethereum's proof of stake has only really been at the pointy end of development in the last year
The precise design for proof of stake has solidified a lot in the last year, in part because Ethereum's other developments such as sharding affected the final form it would have to take.
"We've spent years trying to make Casper FFG work, make sharding work on top of Ethereum chain, so we've definitely gone through at least four or five iterations since the basic idea," Buterin said. "But especially over the last year it's solidified a huge amount."
This highlights how complicated Ethereum development is. You're not just building a single thing to certain parameters and a set deadline. Rather, it's more like trying to assemble a complex, highly-precise machine made of interconnected gears, even while each gear itself is still under construction and subject to change.
"A lot of things seem to be three steps forward and two steps back," Griffith added. "I guess that's common for most scientific enterprises."
3. What bear market?
Hodlers gonna hodl, but more importantly buidlers gonna buidl.
The Ethereum team's experiences in the bearish market have been the same as many others who are experiencing crypto from the inside. Despite some of the headaches and downsizing associated with a crypto funding crunch things have, if anything, gotten more productive rather than less.
"There's definitely been a reduction of interest in just all blockchain products because... people who are just following cryptocurrencies for the price... stop following because the price stops moving in a happy direction for a while," Buterin said. "In general, people that are building things and have been building things for a long time are just continuing to build things."
4. Practical Ethereum applications are hitting a tipping point
The blockchain world was dapp-happy in 2018, but it wasn't dapp-ecstatic. That might be changing with the evolution of tools available to Ethereum developers.
"Right now there are more things to build with and build on top of than there ever have been before," Buterin pointed out, "and I expect there to be even more over the next few months. For a developer, the experience is becoming more and more interesting very quickly."
You don't have to look far to find examples of this change in the real world.
"For example, at the [EDCON 2019] Sydney hackathon, you start seeing, instead of people trying to build components to build things with, you start seeing relatively sophisticated projects coming out of a two-day hackathon because there are so many building blocks you can start piecing together to do interesting things," Ryan observed. "So I think in production we're... definitely hitting an inflection point in respect to the applications we're seeing."
5. It's not all institutional. Viable indie blockchain projects abound
"The most interesting stuff right now that's really hitting people might not be the institutional stuff at the beginning," Buterin said. "For example, these smart contract parametric insurance projects that are starting to pop up. Like, there's HurricaneGuard... there's one out of Sri Lanka, they're basically just using smart contracts to give people the simplest kind of insurance. So if there's a flood, here you get a payout. If there's not enough rain, here you get a payout."
In the case of HurricaneGuard, for example, the idea is that claims can be automatically filed and paid in the event of a government weather station detecting hurricane-force winds within a set distance of the insured property. As long as the financials can be made to work, there's little reason why you can't automate away much of the middle office.
"There's a specific kind of product that just seems really easy to build using a blockchain and smart contracts, and possibly much easier than it is now, and it seems like people are starting to actually use it," Buterin said.
Insurance is one example, but others are more ambitious. Elsewhere in Ethereum-land, people are seeing if they can use smart contracts to automate the rideshare companies out of the rideshare businesses.
Griffith also points out that algorithmic stablecoins might be old news in blockchain, but they're still extremely novel.
"Has there ever been an algorithmically stable thing?" he asked.
He also cited SpankChain as a great example of real-world Ethereum use.
6. The oracle problem isn't unsolvable
The oracle problem is seen as one of the major challenges for smart contracts.
Smart contracts are immutable contracts, and oracles are the things that feed smart contracts their information. So in the case of HurricaneGuard, it would need an oracle to say whether the hurricane occurred, whether someone was insured and so on.
The oracle problem is basically the problem of feeding shoddy data into smart contracts. On the one hand, smart contracts are just oracle problems all the way down. Smart contracts aren't immutable if you can tamper with their data inputs.
On the other hand, this isn't an entirely unsolvable problem, and Buterin gives some examples of how oracles are improving.
"Oraclize is interesting," he muses. "They've managed to extend their system and make it more secure by adding trusted hardware components in a couple of places. So they have a nice security model that basically means in most cases there are several things that need to break for the oracle to stop working, which is really cool."
Rhombus Network is another example of a group focused on solving the oracle problem to the extent necessary to make smart contracts more usable, Griffith adds.
7. Legally binding smart contracts are still a very fuzzy area
Smart contracts are a legal grey area, and there's still a big gulf between smart contracts, and smartness and contracts.
While it's easy to say that smart contracts are immutable, the fact remains that unless it's legally recognised as such, it's not actually immutable. In the event of a discrepancy between the legal state of an object and the blockchain state of an object, the law says the legal state is the correct one.
And while the panel said Ethereum team members had been in discussion with regulators in various places, there's nothing solid yet. This has necessitated some interesting solutions from startups that want to actually use blockchain.
Griffith gives the example of Mattereum, which basically realised it had to actually put lawyers on the blockchain, so to speak.
"So if you want to have there be an interface between the smart contract and the physical world... you transfer the physical asset to a legal firm, and then the legal firm signs a contract saying "whatever the blockchain says we will do. So if the blockchain says we will transfer it, they'll transfer it. So, it's that way you kind of make the legal firm's actions kind of like an appendage grafted on to the smart contract."
So, in this example, would the law firm be extraneous to the blockchain or would blockchain be extraneous to the law firm?
Fortunately, Australia is looking ahead at this kind of eventuality, and solving this problem of legally binding smart contracts is actually one of the key goals of the Australia National Blockchain project.
8. Don't expect private transactions in Ethereum any time soon, zk-Snarks notwithstanding
zk-Snarks caught the imagination of the Ethereum community recently. While the system is usually more closely associated with privacy coins, Ethereum developers were quick to realise that they can also serve as a potentially highly effective scaling solution for Ethereum.
Privacy is still very much on the backfoot though.
"There's definitely quite a way to go before strong privacy can really work well, and part of it is just the inherent limitations of the Ethereum 1.0 chain," Buterin said. He gave the example of a smart contract coin mixer, where the parties would inevitably re-identify themselves by paying the mixer gas costs.
"It's definitely not a very short-term thing at this point," Buterin said. Although "there definitely are teams that are working on it," he added.
9. zk-Snarks currently have a lot more to offer than privacy
"The existence of the technology [of zk-Snarks] was definitely a surprise to me. I'm like 'wow, you can do that?'," Buterin said.
It was first introduced as a privacy solution, and it's taken some time for the Ethereum community to digest the potential applications since then, Ryan noted.
"A few years ago, it was pushed as a privacy thing. It's taken time for the community to digest that, and then digest the fact that you can do other things with it, specifically scalability," he said.
"I'd say scalability is still underrated, privacy overrated," Drake agreed.
zk-Snarks for scaling are very much still part of the plan for Ethereum as an accoutrement for other scaling solutions like sharding.
Sharding alone won't realistically be enough for the needed scalability, Drake pointed out.
10. No, there is not a partnership between Tron and Ethereum
Recently Tron founder Justin Sun alluded to a partnership between Tron and Ethereum. This apparently came as something of a surprise to the Ethereum core developers, none of whom were sure what Sun's going on about this time.
"It's possible we've committed some collaboration with Tron," Griffith suggested.
"I don't think we have," Ryan said.
"I don't think we have either," Buterin added.
"You're the media," he told the assembled reporters. "You're the one who's supposed to be telling us these things."
But the panel still took time to compliment Sun's marketing prowess.
"If Justin Sun was a less-compromised asset, I would love to have him as chief marketing officer for Ethereum," Griffith said.
11. There's a move towards more regular forks
On the social side of things, Ethereum is aiming to normalise hard forks a bit more.
"There's definitely movement on the 1.0 chain to do more frequent and smaller hard forks to kind of get into a rhythm," Ryan notes.
The idea is to normalise more consistent and smaller updates. But on the Ethereum beacon chain (the still-more-experimental Ethereum 2.0 chain), the objective is still to move fast and implement larger changes. But if all goes well, this will likely settle into a rhythm of its own in the future.
12. Ethereum's developers are envisioning a multi-chain future
In an increasingly competitive world, Ethereum's key developers aren't too concerned.
The Ethereum developer community is still by far the largest and most active of any blockchain in the world, Ryan pointed out. But that doesn't mean it's functionally going to be the best at everything.
We don't need to eat everything."
"Ethereum errs on the side of maximum flexibility. If you just want to send tokens around, Ethereum is kind of overkill," Griffith says. On occasion, he's even advised people to look at other chains such as Stellar for their needs.
"We don't need to eat everything," he said.
The decline in ICOs has given way to a new rise in IEOs (initial equity offerings) among companies that want to raise funds. These are similar to ICOs, in that they are token sales, but different in that the tokens actually represent equity and in that they are usually sold through a single exchange.
The decline of ICOs is more than welcome, Buterin said, but he's still not sold on IEOs.
"Having that hype reduction is definitely a good thing for people that actually want to build," he said. "IEOs I'm more down on. Basically they seem just like ICOs but more centralised."
14. It's tough to be worried about ASIC miners
In response to a question around whether ASIC miners are a concern, Buterin explained that the network is going to proof of stake anyway. Even if Bitmain wanted to release an Ethereum ASIC miner, he said, he doubts whether it would be a sensible business strategy.
On the whole, it's just really tough to be overly concerned with ASIC miners coming to Ethereum, despite all the noise around it.
This has been borne out fairly well to date. Buterin's opinions on this issue haven't changed too much, and previous Ethereum ASIC miners have already decisively failed to change the world.
15. Things could get interesting for the old chain after the PoS fork
When people start migrating to the new proof-of-stake fork, things could start getting interesting for the old proof-of-work fork.
The old proof-of-work chain could very well be 51% attacked someday once people start abandoning it, Buterin says. It won't really matter though.
Although from one angle, this has already happened.
16. There's a lot of development and discussion happening in the background
Although the development of Ethereum is certainly much more transparent than most other blockchains as well as most businesses, that doesn't mean everything happens in plain sight.
There's a lot of discussion preceding each new development of Ethereum, Ryan says, and much of this is invisible outside of the actual development circles. New developments go through several rounds of proposing, discussing and feedback before being thrust into the public eye.
"Even though it is radically transparent, the iterative process helps us manage the complexity rather well," he says.
17. The community is more unified than it might seem
"Almost all changes that are technical improvements that are good for Ethereum just pass through uncontroversially," Buterin notes. "Governance has so far only been a problem for a relatively small set of controversial issues."
He mentioned the DAO fork, of course, as well as two issuance reductions, but also said that in all cases there was a relatively clear community consensus on which way to go.
Even the most devastating split following the Ethereum Classic fork was clearly a small minority going its own way, rather than a devastating fracture.
While Ethereum's governance situation isn't perfect, it's difficult to imagine a situation that really does effectively split the community, Buterin says.
"It's definitely possible that there will exist some kind of wedge issue that will divide the community in some way," he says, but it's tough to see what that could be.
18. Ethereum Classic is in the past, but its developers are still pushing the envelope
No one on the panel that day had any doubt that proof of stake was the correct direction for Ethereum to go.
"Proof of stake is great, man"
"Proof of stake is great, man," Griffith said.
Although this does raise questions around why Ethereum has partnered with Ethereum Classic if both chains are so set on radically different directions.
In response to them, Griffith noted that there's still a lot for Ethereum to learn from its estranged sibling chain, and he emphasised that the split was always about ideology more than technicalities.
"They have a bit more purist sensibilities," Griffith said, "but they're not incompetent. When they publish software, 30 or 40% of the time I'm like 'hey, that's really handy'."
Ideology notwithstanding, there's still a lot of value in the relationship. That said, ideology is sometimes a luxury for projects that don't have to worry about the real world as much as Ethereum does.
"I do think Ethereum Classic would probably have a little more trouble if they get larger. The larger you get you have to get more pragmatic because the real world impinges on you more," Griffith suggests.
19. There may be room for improvement in Ethereum's marketing
Ethereum has never been about self-aggrandisement to the same extent as other projects, but the panel concedes that it might have its place.
"We can certainly do better in communication," Ryan noted. "There are some internal efforts to communicate our successes a bit better."
But maybe it doesn't really need to. There are plenty of people happy to shill it with or without prompting, the team agreed.
20. Bitcoin's value is meme value
"The meme value of Bitcoin is very high," Ryan said. "People definitely bought into the meme of, whatever it is, because it was the first of what it is... the store-of-value meme, the king-of-crypto meme, the first-and-thus-better meme.. the gold thing."
Its brand cachet is through the roof, and that might be all the value it needs.
As for Bitcoin's obvious unsustainability in the face of falling block rewards, all the panellists seemed in agreement that it wouldn't break the 21-million rule. That is, after all, part of its meme value.
Some kind of hybrid proof-of-stake system seems the most likely outcome for Bitcoin, Griffith suggested. But it's not necessarily competition, he added.
"When I first joined the foundation, I thought of Bitcoin as the opponent, and thought it was kind of zero sum. I no longer think that," he said. "It's kinda surprising how quickly when you think someone's your opponent, they turn out not to be."
21. The Bitcoin Cash team really knows what it's doing
"Things I'm watching? One of them is Bitcoin Cash. It seems to have recovered a lot from its slump recently, but also technology-wise, it seems like they're beating Bitcoin to integrating Schnorr Signatures, which was just like super out of the blue surprising."
"Maybe two years ago, you might have thought that it's just, like, jokers, but there seems to be some real technical talent there."
22. Nothing's set in stone
The exact reward numbers for Ethereum's impending proof-of-stake mode are still well and truly up in the air – as are many other elements. Nothing's set in stone yet, and there's a lot of testing to go still.
23. There is no revenue. The Ethereum Foundation's time is limited
The Ethereum Foundation does not have revenue. It only has runway, and that runway is designed to eventually run out.
It's extended in tiny bits and pieces by selling T-shirts and event tickets, but it's getting shorter every year.
Where it goes in the future is anyone's guess, but the general consensus is that the clock has a lot of time left on it, but it's definitely ticking.
"There's a lot of opportunity for consulting, but as a non-profit, the Ethereum Foundation will continue to operate as such," Ryan said.
And although there were some thoughts of how to fund projects in the future, it's still too early to start seriously considering another DAO, Buterin said.
Eventually, when the money runs out, the party will be over and we'll get real jobs."
"We only go red. That's all we do," Griffith said. "We want the foundation to eventually have an expiry date. Eventually, when the money runs out, the party will be over and we'll get real jobs."
"I'll probably write my blog more," Buterin mused.
And you can judge for yourself whether Ethereum core developers should bank on a musical career after all of this.
Disclosure: The author holds BTC at the time of writing.