Ethereum Blog

The Ethereum Launch Process

Introduction

user

Vinay Gupta

Hexayurt Guy, former cypherpunk, release coordinator and general strategist


LATEST POSTS

Ethereum’s unexpected future direction 01st April, 2015

Getting to the Frontier 12th March, 2015

Uncategorized

The Ethereum Launch Process

Posted on .

I’m Vinay Gupta, the newly minted release coordinator for Ethereum. I’ve been working with the comms team on strategy, and have now come aboard to help smooth the release process (some of the content in this blog is out of date, please see this link for the most up to date information on Ethereum).

I’ll be about 50/50 on comms and on release coordination. A lot of that is going to be about keeping you updated on progress: new features, new documentation, and hopefully writing about great new services you can use, so it’s in the hinterland between comms and project management. In theory, once I’m up to speed, I should be providing you with the answers to the question: “what’s going on?” But give me some time, because getting up to speed on all of this is nontrivial. We have a very large development team working with very advanced and often quite complex new technology, and keeping everybody up to date on that simultaneously is going to be tricky. To do that well, I have to actually understand what’s going on at quite a technical level first. I have a lot to wrap my head around. I was a 3D graphics programmer through the 1990s, and have a reasonably strong grounding in financial cryptography (I was, and I am not ashamed to admit it, a cypherpunk in those days). But we have a 25-30 person team working in parallel on several different aspects of Ethereum, so… patience please while I master the current state of play, so that I can communicate about what’s changing as we move forwards. It’s a lot of context to acquire, as I’m sure you all know – if there’s an occasional gaffe as I get oriented, forgive me!

I’ve just come back from Switzerland, where I got to meet a lot of the team, my “orientation week” being three days during the release planning meetings. Gav writes in some detail about that week here, so rather than repeat Gav, read his post, and I’ll press on to tell you what was on that release white board.

There is good news, there is bad news, but above all, there is a release schedule.

There will be another blog post with much more detail about the release schedule for the first live Ethereum network shortly – likely by the end of this week, as the developer meeting that Gav mentions in his post winds up and the conclusions are communicated. That’s the post which will give you timelines you can start firing up your mining rigs to, feature lists, and so on. Until then, let me lay out roughly what the four major steps in the release process will look like and we can get into detail soon.

Let’s lay out where we are first: Ethereum is a sprawling project with many teams in many countries implementing the same protocol in several different language versions so it can be integrated into the widest possible range of other systems/ecologies, and to provide long term resilience and future-proofing. In addition to that broad effort, there are several specific applications/toolchains to help people view, build and interact with Ethereum: Mist, Mix, Alethzero and so on. Starting quite soon, and over the next few months, a series of these tools will be stood up as late alpha, beta, ready for general use and shipped. Because the network is valuable, and the network is only as secure as the software we provide, this is going to be a security-led not schedule-led process. You want it done right, we want it done right, and this is one of the most revolutionary software projects ever shipped. 

While you’re waiting for the all singing, all dancing CERN httpd + NCSA Mosaic combo, the “we have just launched the Future of the Internet” breakthrough system, we will be actually be releasing the code and the tools in layers. We are standing up the infrastructure for a whole new web a piece at a time: server first, plus tool chain, and then the full user experience rich client. This makes sense: a client needs something to connect to, so the server infrastructure has to come first. An internet based on this metacomputer model is going to be a very different place, and getting a good interface to that is going to present a whole new set of challenges. There’s no way to simply put all the pieces together and hope it clips into place like forming an arch by throwing bricks in the air: we need scaffolding, and precise fit. We get that by concentrating on the underlying technical aspects for a while, including mining, the underlying network and so on, and then as that is widely deployed, stable and trusted, we will be moving up the stack towards the graphical user interface via Mist in the next few months. None of these pieces stand alone, either: the network needs miners and exchanges, and it takes people time to get organized to do that work properly. The Mist client needs applications, or it’s a bare browser with nothing to connect to, and it takes people time to write those applications. Each change, each step forwards, involves a lot of conversations and support as we get people set up with the new software and help them get their projects off the ground: the whole thing together is an ecology. Each piece needs its own time, its own attention. We have to do this in phases for all of these reasons, and more. 

It took bitcoin, a much less complex project, several years to cover that terrain: we have a larger team, but a more complex project. On the other hand, if you’re following the github repositories, you can see how much progress is being made, week by week, day by day, so… verify for yourself where we are.

So, now we’ve all got on the same page on real world software engineering, let’s actually look at phases of this release process!

Release Step One: Frontier

Frontier takes a model familiar to Bitcoiners, and stands it up for our initial release. Frontier is the Ethereum network in its barest form: an interface to mine Ether, and a way to upload and execute contracts. The main use of Frontier on the launch trajectory is to get mining operations and Ether exchanges running, so the community can get their mining rigs started, and to start to establish a “live” environment where people can test DApps and acquire Ether to upload their own software into Ethereum.

This is “no user interface to speak of” command line country, and you will be expected to be quite expert in the whole Ethereum world model, as well as to have substantial mastery of the tools at your disposal.

However, this is not a test net: this is a frontier release. If you are equipped, come along! Do not die of dysentery on the way.

Frontier showcases three areas of real utility:

  • you can mine real Ether, at 10% of the normal Ether issuance rate, 0.59 Ether per block reward, which can be spent to run programs or exchange for other things, as normal – this real Ether (This was not the case at launch – Frontier block reward is 5 Ether per block, and will remain that amount until Casper).
  • you can exchange Ether for Bitcoin, or with other users, if you need Ether to run code etc.
  • if you already bought Ether during the crowd sale, and you are fully conversant with the frontier environment, you can use it on the frontier network.
  • we do not recommend this, but have a very substantial security-and-recovery process in place to make it safer – see below 

We will migrate from Frontier to Homestead once Frontier is fully stable in the eyes of the core devs and the auditors:

  • when we are ready to move to Homestead, the release after Frontier, the Frontier network will be shut down; Ether values in wallets will be transferred, but state in contracts is will likely be erased (more information to follow on this in later blog posts)
  • switchover to  the new network will be enforced by “TheBomb”

This is very early release software: feature complete within these boundaries, but with a substantial risk of unexpected behaviours unseen in either the test net or the security review. And it’s not just us that will be putting new code into production: contracts, exchanges, miners, everybody else in the ecosystem will be shipping new services. Any one of those components getting seriously screwed up could impact a lot of users, and we want to shake bugs out of the ecosystem as a whole, not simply our own infrastructure: we are all in this together.

However, to help you safeguard your Ether, we have the following mechanisms planned (more details from the developers will follow soon as the security model is finalised):

  • if you do not perform any transactions, we guarantee 100% your Ether will not be touched and will be waiting for you once we move beyond Frontier
  • if you perform transactions, we guarantee 100% that any Ether you did not spend will will be available to you once we move beyond Frontier not be touched
  • Ether you spend will not fall through cracks into other people’s pockets or vanish without a trace: in the unlikely event that this happens, you have 24 hours to inform us, and we will freeze the network, return to the last good state, and start again with the bug patched
  • yes, this implies a real risk of network instability: everything possible has been done to prevent this, but this is a brand new aeroplane – take your parachute!
  • we will periodically checkpoint the network to show that neither user report nor automated testing has reported any problems. We expect the checkpoints will be around once daily, with a mean of around 12 hours of latency
  • exchanges etc. will be strongly encouraged to wait for checkpoints to be validated before sending out payments in fiat or bitcoin. Ethereum will provide explicit support to aid exchanges in determining what Ether transactions have fully cleared

Over the course of the next few weeks several pieces of software have to be integrated to maintain this basket of security features so we can allow genesis block Ether on to this platform without unacceptable risks. Building that infrastructure is a new process, and while it looks like a safe, sane and conservative schedule, there is always a chance of a delay as the unknown unknown is discovered either by us, the bug bounty hunters or by the security auditors. There will be a post shortly which goes through this release plan in real technical detail, and I’ll have a lot of direct input from the devs on that post, so for now take this with a pinch of salt and we will have hard details and expected dates as soon as possible. 

Release Step Two: Homestead

Homestead is where we move after Frontier. We expect the following three major changes.

  • Ether mining will be at 100% rather than 10% of the usual reward rate (Frontier/Homestead block reward will remain 5 Ether)
  • checkpointing and manual network halts should never be necessary, although it is likely that checkpointing will continue if there is a general demand for it
  • we will remove the severe risk warning from putting your Ether on the network, although we will not consider the software to be out of beta until Metropolis

Still command line, so much the same feature set as Frontier, but this one we tell you is ready to go, within the relevant parameters.

How long will there be between Frontier and Homestead? Depends entirely on how Frontier performs: best case is not less than a month. We will have a pretty good idea of whether things are going smoothly or not from network review, so we will keep you in the loop through this process.

Release Step Three: Metropolis

Metropolis is when we finally officially release a relatively full-featured user interface for non-technical users of Ethereum, and throw the doors open: Mist launches, and we expect this launch to include a DApp store and several anchor tenant projects with full-featured, well-designed programs to showcase the full power of the network. This is what we are all waiting for, and working towards.

In practice, I suspect there will be at least one, and probably two as-yet-unnamed steps between Homestead and Metropolis: I’m open to suggestions for names (write to vinay[at]ethdev.com). Features will be sensible checkpoints on the way: specific feature sets inside of Mist would be my guess, but I’m still getting my head around that, so I expect we will cross those bridges after Homestead is stood up.

Release Step Four: Serenity

There’s just one thing left to discuss: mining. Proof of Work implies the inefficient conversion of electricity into heat, Ether and network stability, and we would quite like to not warm the atmosphere with our software more than is absolutely necessary. Short of buying carbon offsets for every unit of Ether mined (is that such a bad idea?), we need an algorithmic fix: the infamous Proof of Stake. 

Switching the network from Proof of Work to Proof of Stake is going to require a substantial switch, a transition process potentially much like the one between Frontier and Homestead. Similar rollback measures may be required, although in all probability more sophisticated mechanisms will be deployed (e.g. running both mechanisms together, with Proof of Work dominant, and flagging any cases where Proof of Stake gives a different output.)

This seems a long way out, but it’s not as far away as all that: the work is ongoing.

Proof of Work is a brutal waste of computing power – like democracy*, the worst system except all the others (*voluntarism etc. have yet to be tried at scale). Freed from that constraint, the network should be faster, more efficient, easier for newcomers to get into, and more resistant to cartelization of mining capacity etc. This is probably going to be almost as big a step forwards as putting smart contracts into a block chain in the first place, by the time all is said and done. It is a ways out. It will be worth it. 

Timelines

As you have seen since the Ether Sale, progress has been rapid and stable. Code on the critical path is getting written, teams are effective and efficient, and over-all the organization is getting things done. Reinventing the digital age is not easy, but somebody has to do it. Right now that is us.

We anticipate roughly one major announcement a month for the next few months, and then a delay while Metropolis is prepared. There will also be DEVcon One, an opportunity to come, learn the practical business of building and shipping DApps, meet fellow developers, potential investors, and understand the likely shape of things to come.

We will give you information about each release in more detail as each release approaches, but I want to give you the big overview of how this works and where we are going, fill in some of the gaps, highlight what is changing, both technically and in our communications and business partnership, and present you with an overview of what the summer is going to be like as we move down the path towards Serenity, another world changing technology.

I’m very glad to be part of this process. I’m a little at sea right now trying to wrap my head around the sheer scope of the project, and I’m hoping to actually visit a lot of the development teams over the summer to get the stories and put faces to names. This is a big, diverse project and, beyond the project itself, the launch of a new sociotechnical ecosystem. We are, after all, a platform effort: what’s really going to turn this into magic is you, and the things you build on top of the tools we’re all working so hard to ship. We are making tools for tool-makers.

Vinay signing off for now. More news soon!

 

profile

Vinay Gupta

http://hexayurt.com

Hexayurt Guy, former cypherpunk, release coordinator and general strategist

Comments
user

Author Joshua Davis

Posted at 8:32 pm March 3, 2015.

So basically you are saying that until you reach Metropolis stage don’t assume any financial contracts you create are in reality binding is that a correct reading of this article? Yes I understand that what you are doing is not easy and this is a new technology and you want to tackle the launch in stages but I need to know at what point will I be able to produce a commercial product on the Ethereum platform. In which stage will I be able to produce a contract where switching from stage A to stage B doesn’t result in the “state in contracts is will likely be erased.” Since the launch is late March I’m guessing we won’t arrive at binding reliable contract code operation until June.

This is news to lots of people. You are doing a good job helping us shift our concepts and expectations. Yes Ethereum is not going to fall out of the sky on March 20th fully formed and ready to live in. But it seems like somewhere between homestead and metropolis contracts will be seen as binding. For whatever reason I didn’t really think this would be the case because there has been a test net up and running with several PoC releases which I thought were enabling binding contracts to be available on 3/20. I think this is a wise decision but if March 20th will not be the moment at which binding contracts can be produced on the network I believe it would be important to know that and to make sure other people know it too.

Reply
    user

    Author gabriel munteanu

    Posted at 8:33 am March 4, 2015.

    That’s how I read this too. The contracts will be reset when they reach Metropolis. So, again, more testing environment. Even the ether transactions will not be 100% real. The author says these can be reversed too. We could see lots of fights. I have 1 question:
    Can I NOT import my wallet in this Frontier stage? and delay it until proper time?

    Reply
      user

      Author George Hallam

      Posted at 12:57 pm March 4, 2015.

      You don’t have to do ANYTHING with your wallet, and if you are non technical it is recommended to wait for Homestead or Metropolis to dive in. However, should wish to interact with your wallet on the Frontier network, any coins that remain in the wallet will be safe:

      – If you do not perform any transactions, we guarantee 100% your Ether will not be touched and will be waiting for you once we move beyond Frontier

      – If you perform transactions, we guarantee 100% that any Ether you did not spend will will be available to you once we move beyond Frontier not be touched

      – Ether you spend will not fall through cracks into other people’s pockets or vanish without a trace: in the unlikely event that this happens, you have 24 hours to inform us, and we will freeze the network, return to the last good state, and start again with the bug patched

      Reply
        user

        Author gabriel munteanu

        Posted at 1:11 pm March 4, 2015.

        Thanks George,
        1 more question: if I earn ether [mine/contracts/whatever] during Frontier stage, it will be available in Metropolis, right? It won’t get reset.

        Reply
          user

          Author Stephan Tual

          Posted at 4:35 pm March 4, 2015.

          All account balances will follow through across ALL releases. So yes is the answer to your question. The currently known ‘reset’ or, ‘re-genesis’ if you will, is between Frontier and Homestead, it won’t affect balances, but it will affect contract code and contract states, which is why we recommend Dapp developers to wait for Homestead before deploying ‘live’ versions of their dapps.

          user

          Author gabriel munteanu

          Posted at 5:00 pm March 4, 2015.

          Stephan, what about the contracts’ balances? Do they get “suicide” and, if there is an address there, the funds will be transfered to that address? Or just sent to the creator’s address?

          user

          Author Stephan Tual

          Posted at 5:30 pm March 6, 2015.

          @gabrielmunteanu:disqus: No that’s not how it would work, instead it will likely be a re-Genesis. Note that balances are property of accounts, and there are two types of accounts: those with associated EVM code, where the account represents an autonomous object, and those without EVM code, where the balance is controlled by an external entity. Both will have their balances preserved at re-Genesis.

        user

        Author Paweł Bylica

        Posted at 4:59 pm March 5, 2015.

        But if I don’t do ANYTHING, as suggested, I will loose because of mining going on in Frontier stage.

        Reply
user

Author Donald McIntyre

Posted at 9:56 pm March 3, 2015.

I think a good definition is that releases will not be determined by time schedule, but by security milestones.

Reply
user

Author pablox43

Posted at 4:09 am March 4, 2015.

Very well written Vinay. Thank you for the update.

Reply
user

Author Dima Starodubcev

Posted at 5:03 am March 4, 2015.

Very, very smooth roadmap. Proud to be part of it.
When and there will DevOne happen? It is critical for everybody who need to get visas.

Reply
user

Author cyber•Fund

Posted at 5:08 am March 4, 2015.

Cannot wait to embed myself to Ethereum.

Reply
user

Author Dima Starodubcev

Posted at 2:37 pm March 4, 2015.

Vinay! Naming idea for release inside Metropolis.
Heroes of Might and Magic tell us that any town begins from tavern.

Reply
user

Author Anthony Di Iorio

Posted at 9:11 pm March 5, 2015.

Glad to have you onboard Vinay!

Reply
user

Author Irontigar

Posted at 4:22 am March 6, 2015.

Hmm….. Burst has AT’s with UI already….AT’s = SM’s….and they have been running good for a while. First ever cross chain tx. will happen between burst and some other coin, perhaps quora, or ora, dunno which one lol.

Also, coming soon, AT’s will be able to have their own ui with metadata, plz ether don’t delete this, plz everyone else check it out. We had 0 funding.

Reply
user

Author Daneo

Posted at 4:23 pm March 7, 2015.

Frontier stage…. 10% mining reward, why?
Where is advantage for those firing the rigs at launch? Why not 1% or 50%??
Till Homestead difficulty rises times and again no advantage for those who were prepared!
We have already removed dust from our 400 gpu’s, and now….. this very dissapointing announcement..,.

Reply
    user

    Author hexayurt

    Posted at 4:46 pm March 13, 2015.

    For a short while. Then 100% mining. We don’t want people piling too much value into the system before it’s been thoroughly tested – and there’s no way around that. It has to be before it’s 100%.

    Reply
user

Author BitBul

Posted at 11:18 pm March 8, 2015.

I recommend increase the mining reward by smoothy 1% each day instead of having one single step.

Reply
    user

    Author hexayurt

    Posted at 4:45 pm March 13, 2015.

    Interesting. Don’t know if that’s feasible or not technically, but I like the idea.

    Reply
user

Author Pat Jack

Posted at 11:43 pm March 8, 2015.

Vinay, please consider assigning a native English speaking “copy editor” to check this and future documents for spelling, grammar and punctuation, thanks. Language is code, and errors in language are very much like code bugs to potential investors or participants, sure, we can forgive you, but I don’t think that’s what you want.

Reply
    user

    Author hexayurt

    Posted at 4:45 pm March 13, 2015.

    Specific errors? Let me know, I’ll fix.

    Reply
      user

      Author Pat Jack

      Posted at 12:40 pm March 29, 2015.

      A copy editor is assigned the job of fixing it BEFORE it is published. When the function of a copy editor is valued and understood, then maybe the Ethereum project will have an opportunity to get one. It’s not a lack of funds, that’s for sure, I just don’t think the Ethereum project sees the value of having a professional, native English speaking copy editor.

      Reply
        user

        Author hexayurt

        Posted at 5:35 pm March 29, 2015.

        I have lived my entire life in the UK and speak only English. Could you point out any errors you saw so I can fix them? Thank you!
        Sent from my iPad

        Reply
user

Author J T

Posted at 5:41 pm March 10, 2015.

This was very helpful!!! Thank you.

Perhaps consider names of transition networks such as the Oregon Trail>>Pony Express>>Telegraph etc. 😉

Good Luck!!! Look forward to more posts like this.

Reply
    user

    Author Cabbot Sanders

    Posted at 9:39 pm August 11, 2015.

    I was thinking more on the line of Space frontiers, such as Mercury, Gemini, Apollo, STS, etc…

    Reply
user

Author madtechnician

Posted at 3:49 pm March 13, 2015.

“Reinventing the digital age is not easy, but somebody has to do it. Right now that is us.”

As soon as I got to that I stopped reading.

Sorry but I think this is a load of B-U-N-K.

I certainly hope I am wrong for the sakes of all the people who have invested in this project.

Reply
    user

    Author hexayurt

    Posted at 4:44 pm March 13, 2015.

    Try reading the rest 🙂

    Reply
user

Author ken

Posted at 12:08 am March 17, 2015.

great names are cobotica coboticas etherist etherizer etherizing hexa hexo

Reply
user

Author Elie Fabsu

Posted at 11:50 am January 31, 2016.

Interesting suggestions – BTW , if you is requiring to merge two images , my boss encountered a tool here bitsy.in/5bfdc

Reply

Leave a Reply to Pat Jack
Cancel Reply

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>

View Comments (33) ...
Navigation