Your One Question About The Constantinople Hard Fork, Answered

Literally no one knows when Constantinople will hit the mainnet, but it won’t be in November.

The question on the tip of everyone’s Twitter feed and reddit thread seems to be, “When can we expect Constantinople?” The assumption underlying this question has often been that someone out there had the answer but just didn’t feel like sharing.

However, that was not – and is not – the case. The Ethereum community is all about transparency. For chrissake, the core developers livestream their meetings. There are countless channels, threads, feeds, chatrooms, and repositories chock-full of conversations, debates, collaborations, and decision-making processes of all sorts.

To be fair, that might be where the confusion stems from: There’s so much information out there but a lack of organization. Finding information on Ethereum-related developments is like searching the internet in the 90s before Google. It’s impossible.

I have never done so many web searches to find exactly zero results as I have while researching Ethereum, and let me tell you: I am a pretty good Googler. I was raised by my mother, who’s been doing SEO since before I hit puberty.

But hey, that’s what the media is supposed to do, right? Inform people. Organize information. So here we are, trying.

So! Today there was a core devs meeting that sought to answer everyone’s (including the devs themselves) Big Burning Question. But also, and more importantly, the attendees sought answers to the question: “What needs to happen before the Constantinople hard fork?”


The earliest that Constantinople will arrive is January, but further discussion on the topic will be had at the AllCoreDevs meeting scheduled for November 9.

It’s not unlikely that the hard fork will come in February, though. Or later. (I see you eyerolling.) March was even thrown out there as an option, as that would make the inclusion of ProgPoW a possibility.

Now, as to that other Big Burning Question? There’s a lot that needs done before mainnet launch.

For one, the snafu with the Ropsten hard fork was a good indicator that there needs to be better communication and coordination between clients and miners. When Constantinople rolled out on Ropsten, it happened two days before expected and there were no miners for the new chain.

Also discussed was the need for a hard fork-specific forum to quickly and easily discuss any issues (they refer to this as the war room). For now they’re just using the AllCoreDevs Gitter channel.

The need for more organized and comprehensive, navigable repositories of information was abundantly clear throughout the meeting, but so was the explanation for the deficiency: decentralization. I hate to say it, but there are some things centralized entities – and the internet they built – do pretty well, and straightforwardly disseminating information is one of those things. On this topic, core developers Lane Rettig and Hudson Jameson went back and forth about the need for any communication and coordination efforts to remain decentralized and not “too coordinated.”

Other than issues with communication, there was some discussion around coding the hard fork block number into the client software, as to minimize the need for coordination. However, not all clients allow for this, and some raised concerns that this was unduly complicated. It’s also notable, though no one said, that this might be considered somewhat of a governance overreach – a concern that some have raised about on-chain governance mechanisms that automate updates.

They also discussed the need to fix the fork monitor, which was not present for the Ropsten fork. Apparently having an active, functioning fork monitor would have helped clients identify the canonical chain to mine.

Moving forward, those on the call seemed to agree that a minimum of one month on the testnet is important prior to mainnet launch. Dmitrii of Harmony suggested that for the next hard fork, Istanbul, a hard fork should be rolled out on testnets as much as six months prior to mainnet launch, and that each EIP should be rolled out one at a time. Hudson Jameson and Lane Rettig seemed on board with this idea, but no conclusion was reached.

So, in conclusion, maybe don’t hold your breath for the next hard fork. And hopefully better communication is coming soon – between developers and miners, between developers and each other, and between developers and all of us.


Leave a Reply

Your email address will not be published. Required fields are marked *