Ethereum Blog

Olympic: Frontier Pre-Release

Introduction

user

Vitalik Buterin


LATEST POSTS

Roundup Q2 08th July, 2017

Roundup Round III 24th May, 2017

announcement

Olympic: Frontier Pre-Release

Posted on .

What began all the way back on midnight of February 1st 2014, is coming to a close: we are happy to announce the release of the ninth and last in the Ethereum Proof-of-Concept series. We invite the community to participate in the ongoing Proof-of-Concept IX testnet in our present release, Olympic, made available now.

The purpose of Olympic is to reward people who try to test the limits of the Ethereum blockchain during the pre-release period, spamming the network with transactions and doing crazy things with the state, so that we can see how the network holds up under high levels of load. At the same time, application developers, data providers, exchanges, and users are encouraged to develop and deploy on the testnet and run nodes – and if you have multiple virtual private servers, spin up as many nodes as you can.

Olympic will feature a total prize fund of up to 25,000 ether. There will be four categories of prize as well as a grand prize for the first person to create a substantial fork on the testnet.

The four categories of prize will be Transaction Activity, Virtual Machine Usage, Mining Prowess and General Punishment. Each category will include a main prize of 2,500 ether, as well as one or more smaller prizes at 100 – 1,000 ether and possibly tiny rewards of 0.1-5 ether simply for participating. Each of the categories will be judged by Vitalik, Gavin and Jeff, likely with substantial help from automated blockchain analysis tools. In addition to the Ether, like the bounty programme, all prize-winners will be entitled to have their name immortalised in the Ethereum Genesis block. Please note that to be considered for a prize, send an e-mail describing a claim to olympic@ethereum.org.

This is also the last phase of the Ethereum development process before the Frontier release, as the network is currently proving to be highly stable at its current size of 20-100 nodes, all major clients have been staying in consensus and we are approaching code freeze pending testing and auditing inputs. It is expected to last close to 14 days, though we reserve the right to shorten or lengthen it based on technical considerations. When we deem that we are ready, we will provide a 48 hour countdown for the Frontier 1.0 launch.

Binaries and source are available here: https://github.com/ethereum/go-ethereum/releases/tag/v0.9.18

A work-in-progress guide to Frontier is available here: http://ethereum.gitbooks.io/frontier-guide/

Transaction Activity

This includes activity related to sending and receiving transactions. Examples of things we might reward include:

  • The account that sends the highest total number of transactions
  • The account that contributes the highest total of gas usage
  • The account that receives the highest total number of transactions
  • The account that contributes the highest bloat (as measured in bytes) to the blockchain in terms of transaction trie contents
  • The account with the lowest address (in terms of lexicographic order or numerical representation; the two orders are equivalent) that sends at least one transaction
  • The two accounts that send at least one transaction whose addresses are closest to each other

Virtual Machine Usage

This includes activity relating to using the virtual machine. Examples of things we might reward include:

  • The account that makes the highest number of calls of any opcode
  • The account that makes the highest number of calls of any opcode within a single transaction
  • The account that sends the single transaction that takes the longest for a particular client to execute
  • The account that sends the single transaction with the highest ratio of time to execute per unit gas consumed
  • The account that receives the highest total number of messages
  • The account that reverts the highest total amount of gas usage
  • The account that creates the contract with the lowest address (in terms of lexicographic order or numerical representation; the two orders are equivalent)
  • The account that sends the transaction with the largest (as measured in bytes) encoded receipt RLP
  • The account that produces the largest number of transactions whose bloom is a (distinct) prime number
  • A creator of a contract that contains, for example, a sudoku solver, taking an array of 81 values as input and outputting the array of 81 values corresponding to a solved sudoku

Mining Prowess

This includes activity relating to mining; miners will be identified by coinbase address. Examples of things we might reward include:

  • The miner that produces the largest block in the main chain (as measured in bytes)
  • The miner that contributes the largest total number of bytes of block data to the main chain
  • The miner that mines the largest number of blocks in a row
  • The miner that includes the most transactions
  • The miner that produces the block that takes the longest amount of time for a particular client to process

General Punishment

This includes mucking around with the State in crazy ways, getting contracts to do crazy things. Examples of things we might reward include:

  • The sender/miner of the transaction/block that adds the largest number of items to contract storage
  • The creator of the contract with the largest code (externally-owned accounts only)
  • The creator of the contract with the most filled items in storage
  • The creator of the contract that contains a key/value pair with the highest merkle tree depth (ie. the largest number of hash lookups before getting to the value)
  • The creator of the contract that suicides while containing the most filled items in storage
  • The creator of the contract that makes the largest number of consecutive storage updates such that the storage root at each step is a prime number.

Note that the above categories are by no means an exhaustive list of things that are eligible for prizes. If you do anything outside of the above that you feel merits consideration, please email us.

The Grand Prize

A grand prize of at least 5,000 ether will be shared between the miner(s) who manage to create a substantial fork between the Go and C++ clients. The fork must be on the main chain; one client must accept the block with the other client rejecting it. Smaller prizes for forks between Go/C++ and Python may also be available.

Good luck, and we look forward to seeing and hearing what you all come up with!

– The ethereum core dev team

profile

Vitalik Buterin

https://ethereum.org

Comments
user

Author cyber•Fund

Posted at 4:26 pm May 9, 2015.

So much of emotions, Vitalik! Congrats The Team!

Reply
user

Author Paul Paschos

Posted at 4:53 pm May 9, 2015.

Here we go!

Reply
user

Author Brad A

Posted at 5:51 pm May 9, 2015.

Excitement!!! Excellence! Launching ahead to better humanity! Ethereum!!! Congratulations!
EXCITING 🙂

Reply
user

Author William Mougayar

Posted at 6:10 pm May 9, 2015.

Very creative set of awards….very much aligned with your objectives of validating the overall richness of the Ethereum blockchain.

Reply
user

Author Brian L

Posted at 8:29 pm May 9, 2015.

Are there test coins to use and if so how do we get them?

Reply
user

Author Eugene Wang

Posted at 9:20 pm May 9, 2015.

I’m having trouble running ethereum on Ubuntu Vivid Vervet. Help?

Reply
    user

    Author Stephan Tual

    Posted at 4:44 pm May 10, 2015.

    14.04 is supported, anything else is going to be trouble at this stage. If you find a bug in Olympic however, please report it to github.com/ethereum/go-ethereum/issues. Thank you!

    Reply
      user

      Author Eugene Wang

      Posted at 12:36 am May 16, 2015.

      Ok, I’ll see if I can work with docker.

      Reply
user

Author Joseph Lubin

Posted at 3:02 am May 10, 2015.

In preparation for takeoff, please ensure that your seat backs and tray tables are in their full upright and stowed positions, your seat belts are securely fastened and your space helmet is securely fastened to your zero-G suit.

Reply
user

Author Alexey

Posted at 12:19 am June 1, 2015.

Looking at the state of things so far, I noticed that there are so few uncles included in the mined blocks. When I read white/yellow papers, the expectation was that there would be lots of them, because of the insentives. I looked at the source code of ethminer (assuming that most of the Test ethereum is mined in this way) and confirmed my suspicion that the miner throws away its work on the current block as soon it receives notification about an alternative block. That makes perfect sense in Bitcoin mining, but might not – in Ethereum mining. I think we need to experiment (via an extra option in ethminer) with another strategy: instead of throwing away the work and try to join the race after the new declared block, finish the current block and propagate it, it will very likely be included as uncle.

Currrently, from my point of view, inclusion of uncles only happens due to propagation latencies, i.e. when one miner thinks it was the first to have completed the block, but it wasn’t. That would explain such a low number of uncles. And I believe the idea was to enhance the network by throwing away much less work than in Bitcoin PoW mining. what do you think?

Reply
    user

    Author Alexey

    Posted at 1:14 am June 1, 2015.

    I was a bit naive 🙂 after I modified the ethminer to follow the strategy of not throwing away the work, I realised that the main limitation for my miner is not competition with others, but the actual difficulty of the task – most of the time it cannot (with 4 MH/s) solve the problem within 84 second (12 – average block time * 7 number of generation where uncles could be included). So now I am experimenting with strategy somewhere in the middle – I would try to mine my block until either I find the solution or 50 seconds have elapsed, whichever comes first. Will see how it performs 🙂

    Reply
      user

      Author Alexey

      Posted at 8:27 am June 1, 2015.

      That strategy (50 seconds) did not perform very well – it did not mine anything. I am trying to understand why

      Reply
user

Author Alexey

Posted at 9:37 pm June 5, 2015.

I now suspect that the mining performance is not really determined by the hash rate, but by something else, for example, quality of entropy in the random number generator.
Over last few days, I consistently got surprisingly good mining output.
For example, today I left my miner running when I left for work at 8am. When I came back at 8pm, it mined almost 600 ether, which is approximately 5% of all ether totally mined during these 12 hours (maybe my calculations are wrong). But I only possess about 0.2% of the network hash rate (if the status monitor is right). So today I was 25 times more lucky than I should have been 🙂
My suspicion (I will try to check that if I have enough time) that the current PoW could be more effectively performed by a stochastic algorithm (or more stochastic than it is currently)

Reply
    user

    Author Alexey

    Posted at 9:45 pm June 5, 2015.

    Here you go, as I was typing it, another 60 ether mined 🙂

    Reply
      user

      Author Alexey

      Posted at 10:34 pm June 5, 2015.

      Actually, it looks like all my peers are refusing synchronisation/download requests, so I am mining my own fork :). It is strange though, how come such “blocade” would come about 🙁

      Reply

Leave a 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 (16) ...
Navigation