Summary of Duncan Coutt's Video Update on Testnets

summary

#1

Duncan Coutts, Director of Engineering at IOHK, explains the release of the first Cardano testnet and the types and process of forthcoming testnets.

Summary of the Video Update:

  • There are 2 different kinds of testnets:
    • There are testnets that are about a new feature like a tech preview
    • And there is also the ongoing testnet that are run forever and that show what will be in the next version
  • It is important to distinguish these two types as they have different purposes

Ongoing Testnets

  • This is the testnet that should be available at anytime, for anyone that wants to do system integration or testing, or exchanges that want to use and work with the protocol
  • It is also used by integration partners and end-users to see what will be in the next version
  • There is a process of rolling development and release cycles: internal QA > public testnet > deploy into mainnet
  • The purpose here is for end users to see what’s going on, try out new things, and see if anything has accidentally broken that was missed by the internal QA processes

“Tech Preview” Testnets

  • These are about showing the technology is working and trying to get feedback on these new features
  • For example, the Shelley project and decentralization, wallet features, Goguen
  • Want to get early feedback from developer and users on how its working and whether its meeting their needs

Upcoming Testnets

  • As part of the Goguen project, there will be two testnets released
  • First will be KEVM and the next will be IELE
  • Both are Ethereum / EVM-related features
  • Cardano will support EVM-compatibility as well as an “improved” EVM-like system, which is called IELE
  • The first testnet will be deployed as part of Ethereum Classic client
  • The purpose will be to get feedback on how well these systems work and to receive other suggestions
  • The goal being that down the line, when they are integrated, everything will work smoothly
  • These two testnets will be the ‘tech preview’ style

Additional Testnets

  • There will be a whole series of testnets over the next 6 months
  • There will be a testnet corresponding to the Cardano mainnet – available publicly in the weeks before the mainnet update as part of the rolling release cycle
  • In this, there will be an Ada faucet available
  • Partners and users will be able to try it out
  • There will also be a testnet for decentralization and staking, most likely at the backend of Q3
  • This will test the ability to run stake pools and the ability to delegate your stake using the Daedalus interface
  • There will be further testnets of the “tech feature” variety on the Goguen project throughout the year
  • Features like sidechains and multiple currencies
  • These will be rolled out as and when they are ready

“Different testnets are aimed at different people for different purposes”

  • Smart contract platform testnets are aimed at developers, people who want to write smart contracts, people who might be using Ethereum and writing Solidity programs
  • KEVM and IELE testnets are aimed at people who are familiar with Solidity or writing applications on Ethereum
  • But later in the year, there will be smart contract testnets that will demonstrate the other languages: Plutus and Marlow
  • These use a different style of languages, which many software developers will hopefully find interesting
  • Delegation and Decentralization testnets are aimed at people who are interested in running stake pools, people who are interested in seeing how delegation will work

Stake Pool Registration

  • Stake pool registration process ended in May
  • The actual testnet won’t be ready for several months, probably backend of Q3
  • IOHK will try to keep giving updates on this process
  • Approximately 2000 people have applied
  • Important to note that everyone will have the ability to be involved, but there will be different levels of participation
  • In part, they are testing the social dynamics of stake pools
  • They have the game theory of how this will work, but the testnet will allow them to see how it functions in real life
  • Some people will be operating stake pools but they will also need the end users who choose between stake pools
  • IOHK will provide particular assistance to the selected subgroup of people who they hope can run a stake pool in the testnet
  • This will be in the form of additional documentation and communication
  • During the testnet, they expect to provide those people who are stake pool operators a bit more ‘test Ada’ upfront. This is because part of the game theory requires stake pools to have a slightly larger amount of stake
  • Everyone else will be able to take part and go to the ADA Faucet and get ‘test Ada’ of which they can delegate and choose between stake pools on the testnet
  • It will be a long process and detailed updates will be provided throughout

How to Use Testnets

  • Access and information on the testnet can be found on the testnet website: testnet.iohkdev.io
  • This website will be a portal of all testnets
  • You will be able to see what the testnets are, information on each one (what’s it for, who is it for, detailed documentation)
  • There will also be resources on using them (installers, other tools)

Getting in Touch and Involved

  • It depends on the questions you have
  • If you have a bug that you think you found in the beta tests and you are using Daedalus, then please submit logs to IOHK to be analyzed by developers
  • If it is about how to use testnet or API, use the community forums

Timeframes of Testnets

  • Different testnets will last for different periods of time
  • Rolling release testnets, that correspond with the current mainnet, will ‘never finish’
  • It will be constantly updated and always there
  • Feature or Beta-style testnets will have a limited lifetime but even then, can be updated multiple times (with new features added or fixes incorporated)
  • In particular for smart contract testnets, it is expected that there will be several testnets in parallel and some to run for some time with updates every couple months

[뉴스레터] 카르다노 재단 뉴스 레터 - 2018 년 6 월 15 일
#2

Looks like the Cardano Network does require Stake Pool Operators to hold a significant stake. #Smart #beginners


#3

The question is “how many”?


#4

Enough to hurt if the operator does something funny.


#5

I’d also like to know that too and in addition to how many, if they can release the connectivity and uptime requirements for a node. We can’t expect decent network liveness if nodes are run on someone’s cable/DSL connected HTPC or mobile phone (at one extreme), so I’m gonna take a guess here and say the minimum will be either a dedicated 1U server colo or a reserved VPS instance with guaranteed CPU/RAM performance, both of which would be inside a Tier 4 data center…

And if solo staking is untenable for most, one can always try to pool with others… although in either case, I’d be very surprised if this is some astronomical number of ADA, as this would go against the grain of what they’re trying to achieve (IMHO).

Whatever number it may be, it sure will shed more light on the character of the project…as well as the founders. I too am very curious how they will determine that ‘golden’ number of Ada required for staking…solo or pooled. That number also has implications and direct impact on the network’s performance, scalability, latency, throughput, level of decentralization, etc. Maybe it won’t be a fixed number …


#6

No it does not, technically. It says “This is because part of the game theory requires stake pools to have a slightly larger amount of stake” and the key words here are game theory. So the right term would be “suggests” and not “requires”. It’s just a predictor that most probably pools with slightly larger amount of stake will survive on average better. That’s it. Of course, in my poor understanding.

Original Ouroboros requires some limitation mechanism on the committee size, since PVSS exchange requires quadratic communication overhead. But 1/k rule, imo, helps to mitigate this issue. Praos does not require any limitation, so technically nodes may be created with 1 ADA.


#8

For solo stakers, I don’t think there is a minimum. It doesn’t make sense to have a minimum because the probability of being able to validate transactions depends on how many ADA you have.

There might be an implied minimum beyond which it may be not profitable to run a node. Say, if you get picked once every month on average to validate a block, but your cost of running the node for a month are higher than the reward you get from the network…

Once we have a clear understanding of rewards, we can calculate how much profit we can hope to make. My guess is at lower ADA levels it just won’t make sense to solo-stake.


#9

You are absolutely right. Duncan may just be talking about Pools having 1/k stakes, which, compared to the rest of the network, would likely be higher than average.


#10

For solo stakers, I don’t think there is a minimum. It doesn’t make sense to have a minimum because the probability of being able to validate transactions depends on how many ADA you have

Well said!


#11

so there is a magic number of Ada that is worth having then :slight_smile: . I definitely want to be able to stake and keep the network alive. Just not sure if I’ll be able to … yet.


#12

Economically, there may be. And it will depend on your power costs. I hear electricity is cheap in some countries so leaving a pc on for months isn’t that expensive.BUT if you have a small stake, it’s much better to just delegate it to a pool. It will save you the headache of running your node AND improve your chances of getting more money without spending on electricity.


#15

a little typo


#16

:woman_facepalming:t2: Fixed now. Thanks for flagging!