With much of Metropolis finished, save for one or two complicated features, a decision was made in 2017 to split the upgrade into Byzantium and Constantinople.
Byzantium went into effect without any problem in 2017, while Constantinople was pretty much forgotten about.
That’s because all attention turned to Hybrid Casper. Full Proof of Stake (PoS) was foreseeably going to take some time, so Vitalik Buterin, Chief Scientist at the Ethereum Foundation, came up with a simplified version of hybrid PoS and Proof of Work (PoW).
Testing was undertaken for months with the launch of Hybrid Casper very much imminent. Then Justin Drake, a someone new eth developer, “announced” deep in a reddit thread that Hybrid Casper had been called off.
The bombshell surprised even the individual/s that had been working on the Hybrid Casper testnet for months. Like a little baby, their reply was something like erm muurm murm waaa, so disappointed.
For a decentralized network, no one bothered to say something like: well you can do what you want, we’re going ahead with it.
Some ethereans did make noise, but more like reee rather than: who are you to call anything off? How do you just make such decision without public consultation first? How do you then just announce it like that without some thorough explanation?
As Buterin explains it, Justin Drake and Karl Floersch suggested scraping hybrid for a new approach. Buterin was not convinced, but they persuaded him.
Instead of then having some sort of, at least private, discussion with other devs, they then just said we’re not doing it.
This utterly peculiar decision becomes even stranger when later on some other dev says that Hybrid Casper can be implemented in the current PoW chain.
As we now know a lot more about the plan, such suggestion sort of makes sense because what has been proposed is two blockchains running in parallel. The Beacon chain, which is fully PoS but not quite a full blockchain before sharding, sort of links to the PoW chain.
The PoW chain will keep on running, so it isn’t very clear why that itself can’t be a hybrid – considering the PoS aspect is just a decentralized checkpoint – with the Beacon chain then linking to the hybrid.
In other words, it remains unclear why they scrapped hybrid casper if it is the case that the Beacon chain sort of has nothing to do with it.
In any event, Metropolis is brought back on the table. This is finished, they said, little bits here and there, with that statement being during summer.
As PoS was delayed, an issuance reduction was proposed. Eth holders voted almost unanimously in favor of reducing it with the vast majority asking for 1eth per block.
Devs, however, were quick to say that Coinvote isn’t a decision making mechanism without explaining how else do you get to a yes or no and who exactly says that yes or no.
Anyways, Hudson Jameson, who is in charge of devs relations and communications, organized a number of dev calls on the issuance reduction matter, with one of them inviting the miners.
Why on earth should the miners be consulted remains an unanswered question. In any event, in a devs only call, they agreed to reduce it to 2 eth.
Ethereum’s price at the time was about $300 or $400. The snail speed movement towards the issuance reduction, however, probably contributed to a significant fall against the bitcoin ratio and a reduction of its price to now $120, effectively the then 1 eth per block.
There are two dates we remember as very brutal where price is concerned. September 14th and November 14th or so. The former coincides with the decision to go for 2 eth, the latter…
The “imminent” issuance reduction found some testnet bug. The upgrade is immediately called off to January. There was going to be Devcon, too stressful for devs partying at ravecon0. Then there was going to be Christmas at the end of December, so the whole of December was too stressful. The price crash in September and in November, however, was presumably quite relaxing.
Fine, once, twice, they the technical ones. Give them a chance. Then suddenly Jameson “tentatively” gives the go-ahead to ProgPoW.
What? Had this not been settled last year through public discussion? Do we have some sort of council now that just makes decisions without some general public debate?
Debate followed of course, rumors circulated, then they effectively admitted Nvidia’s gear performs better, but that’s somehow AMD’s fault.
Bringing us to yesterday, the bug that “did not find any cases of this vulnerability in the wild.” Imminently ready two years ago, but somehow we have two last minute bugs.
Now we say the bug did not affect anything. Perhaps technically, but finding or exploiting some bug, especially if harmless, is one of the most effective social attack vector in public blockchains.
Now there will probably have to be another testnet, another main-net block number, with it probably taking at least two months.
Coincidentally, the ProgPoW team was pressuring devs to implement ProgPoW in two months. Now they’ll probably ask for it to be bundled together with the issuance reduction so as to force a controversial matter rather than it be judged on its own accord.
With millions on the line, greedy miners probably don’t care what the effect of that would be. Primarily, some estimate 60% of ethereum’s hashrate is asics. Meaning a 51% attack is foreseeable to the point of very likely if the algo is changed.
Where it concerns a major coin like ethereum, what the effect of such 51% attack would be, no one knows. Especially considering that for a gpu coin, the gear wouldn’t really be at risk. They can just move it to some other coin.
Meaning yesterday’s delay may have more significance than it appears as there is clearly here an interested party pushing for an algo change with this now extra two months or more clearly being just a coincidence.
Buterin is right however that there is the difficulty bomb which has kicked in. Meaning miners reward is gradually slowly reducing already. So from an issuance reduction perspective, the delay might not quite matter.
In fact, it may well be for the better because the ecosystem may well wait until block rewards fall to the equivalent of 1 or 1.5 eth per block. Usability might be affected a bit there, but miner’s refusal to increase capacity has affected it considerably in any event.
So whether there is an attempt at sabotage might not really matter as the clever ones have taken precautions in the way of the difficulty bomb.
In that fashion of: it’s good for ethereum, it might even really be so because at this stage one has to reasonably ask why should the difficulty bomb be delayed at all?
The purpose of it is obviously to move the network to Proof of Stake. With the issuance reduction now delayed and delayed to this point and with the Beacon chain testnet potentially out by March, is it not reasonable to ask for the Beacon chain to launch first and then delay the difficulty bomb?
Bitcoin operates finely with 10 minute blocks. Ethereum can too temporarily if need be to effectively force the miners’ hand, miners who now might be interfering.
Moreover, the risks of a 51% attack once the Beacon chain launches are irrelevant to the point of non-existent, if we are not mistaken, due to the checkpoints.
As such, one wouldn’t care at that point, besides miners, whether to ProgPoW for the PoW chain as it wouldn’t have any foreseeable effect on network security.
Before then, one can see reasonable opposition to ProgPoW because it makes a 51% attack very likely, something which could be devastating for a major coin.
So it may well be better perhaps to scrap Metropolis all together until the Beacon chain. Leave it at Byzantium for now and live it up to the difficulty bomb to do its designed job in light of recent events.