On April 22 the Idena team hosted its first AMA (Ask-Me-Anything) in the Idena’s official Telegram chat. In case you missed it, here is the AMA transcript. Answers to some questions that were asked but not addressed during the AMA have been added.
Project & team history
A project such as Idena can’t come up overnight. Could you explain to us how the idea germinated? How the team was created? Why the name “Idena”?
Our team is in crypto since the end of 2015. We were actively involved in tackling the Bitcoin blockchain scalability issue with 2nd layer solutions like micropayment channels /lightning network. However, as we faced the limitations of these approaches because of the Layer 1 constraints we decided to go deeper.
We believe that decentralization is the most critical part in the blockchain scalability challenge. Only an exaggerated decentralization on layer 1 can enable sharding with the true horizontal scalability.
We realized that personhood is the only alternative to money that is used as the basis for both PoS and PoW consensus mechanisms, so we started considering the formalization of cryptoidentity.
The person who created the Idena’s logo saw Idena as the goddess of fairness and identity. Her name is an allusion to the words ‘Identity’, ‘Idea’, ‘DNA’.
Why is the team anonymous?
From the very beginning we wanted to decentralize the Idena protocol as much as possible and didn’t want to personally expose ourselves not to become a point of failure. We envision that in the long run Idena protocol should be able to sustain its own development without reliance on the current core team.
Can you kindly give us a rough idea of the size of the development team?
There are 5 developers in the core team. However in fact there are many more of them today from the community side who develop Idena-based apps and services and contribute to the Idena client development, to mention a few: @endogen, @realRioda, @menxit, @bingbinglee1, @Angainor, @ProjXDev and many others. Thank you for your great contribution!
Does Idena have anything to do with Ethereum?
No, it doesn’t. However, we think it might be useful to bridge the Idena blockchain with Ethereum smart contracts. The Idena address generation is compatible with Ethereum. Theoretically it is possible to export the list of identities into a smart contract without having trusted oracles but with some availability assumptions.
Use cases
In some ways Idena feels similar to Bitcoin/Ethereum in its early days, solving problems yet to exist. What first “killer use case” do you see coming from Idena’s protocol?
Bitcoin gave us censorship-resistant cryptocurrency, its killer use case is p2p payments. Ethereum’s crucial innovation was smart contracts and its killer feature is dapps, like DeFi, DAO, etc.
The key building block provided by Idena is a unique human cryptoidentity. We think that the killer use cases for cryptoidentity can be related to governance in digital democracies, public good and community project funding, and other useful tools for people worldwide. We would want to rely on our community to help us build such applications on top of Idena.
Does the team view Crypto UBI as the ultimate goal of Idena? Was this the primary motivation behind the project?
The primary driver for creating cryptoidentity for us was the attempt to resolve the blockchain scalability trilemma to have a safe and decentralized consensus with a throughput sufficient for worldwide applications. Basic income from the protocol reward is rather a nice side effect associated with the formalization of cryptoidentity.
In the landscape of identities, how do you see yourself competing with identity solutions such as uPort and Sovrin?
We do not consider these and other SSI projects to be Idena’s competitors as they do not set the task of ensuring the uniqueness/semi-uniqueness of human identity.
There’s a lot of test models to validate humanity. In terms of being easy for humans and hard for AI, are you 100% convinced the flip system is the most optimal?
So far the Flip Challenge was our best guess for a Turing test that could be language agnostic and wouldn’t require personally identifiable information. We are open to other approaches.
As Idena is decentralized and doesn’t collect personal info, the marketing use case is confusing: Advertisers want to see PII such as age, sex, location, interests etc., of which Idena shows nothing. How will you combat this?
Advertisers should reveal the information about those who may be interested in a particular advertisement. And users, seeing all advertisements, select those which are relevant to them, without revealing their identity properties such as gender, age, or location. We call it self-targeting.
I’m tired of having to create a new identity everywhere a new reputation on Amazon, eBay, Facebook, etc. Will Idena allow me to bring my one identity everywhere?
The use of the Idena cryptoidentity largely depends on the decision of these services. From our side, we will provide convenient tools, so that the cryptoidentity could be integrated into any websites and Internet services.
Seems to be that Idena could be a great use case for crypto projects specifically via distribution, etc. Will Idena partner up with other projects in the future? Any plans?
Fair value distribution and crypto governance are very relevant use cases. Idena intends to build bridges with other crypto projects. For example, we will support a Hackathon soon to be launched by Radical Exchange and Gitcoin. Idena will grant DNAs for the Hackathon challenges related to governance in electronic democracies. Everyone is welcome to join and contribute. Some more information is available here.
Flips
Idena says you can have a human status whether you make a good flip or not. If the same bad flip is used to decide someone’s status, how are these not correlated?
The Idena identity status reflects how good the person is in solving flips. The person who creates bad flips should be penalized. We plan to improve the protocol in order to address the flip quality issue.
Can you please low report barrier for a bad flip from current 66% to 33%? Otherwise the whole committee based flips curation doesn’t work. What do you think of that?
Yes, we think it is a good idea to lower the report barrier to 33%. This would increase a fraction of disqualified flips roughly by 10%.
Will we primarily rely on reporting for consensus to weed out bad flips? How will you prevent malicious entities who create simple flips (i.e. 1–2–3–4 sequences) from gaining a majority of the network, given nobody bothers reporting?
It can be regulated with the help of criteria by which a flip should be reported.
We also believe having AI embedded in the Idena app is a mid-term task in order to prevent users from creating flips that are too easy for AI.
Why are reported flips not excluded from consensus?
It will be implemented in the next hard fork release.
Is there a reason that flips are distributed in such a way that you get multiple flips from the same identity (sometimes all of their flips)?
At a short session, participants receive flips from various authors. It is the long session where flips from the same authors are usually qualified. This is done intentionally in order to optimize the network load during the Flip lottery.
Unfortunately bad actors create multiple (if not all) bad flips so you often get lots of bad flips from one address. Will there be a possibility to distribute flips over more addresses with sharding?
Sharding won’t change it. We’d rather plan to disqualify reported flips.
Can biometrics be used to replace flips in an anonymised and protected way?
Biometrics is not a good idea, it relies on specific sensors and algorithms which can be faked. The problem is also that it doesn’t have plausible deniability.
Do you think Idena flips can be used instead of Google Recaptcha?
It is not possible because every single flip can be used only once.
Time between validations
As you have seen, the network grows between 15–20% epoch over the epoch, and so does the time between validations. Usage of Idena Identity as a login on third party services won’t be comfortable with 91 days between validations, will it?
The Idena’s identity should be validated every epoch. Third-party services may choose to log in users with any Idena address but to give/not to give them certain privileges, depending on their current identity status. Services can also give privileges to identities with the Candidate status and not wait for validation if for example other validated members of the network have vouched for these identities.
In addition, there was a proposal suggesting to reduce the maximum length of an epoch to 1 month — this is a subject for a community voting, but we are ready to support this proposal.
What about the timeshift problem with the validation? If you work you can’t just stop and do the validation or you can get fired.
Sorry guys, it cannot be done differently. Read more here.
Voting/governance
What are your thoughts on on-chain governance mechanisms?
We suggested principles for it in the Community report. You are welcome to read and comment.
Are there any plans for a voting system for verified identities so they could vote on suggestions and propose their own? Timeline?
In the short term, we plan to give a protocol for signing transactions with the Idena app to 3rd party services. This will allow to organize voting sites outside the application. We’ll be happy to provide co-funding to such projects.
AI
Have you considered the use of AI to prevent people from uploading numbered flips/progressive object count flips? Also NSFW flips?
It makes no sense to use AI for the numbered flips. It can be regulated with the help of criteria by which a flip should be reported. However, NSFW flips are to be filtered with built-in image recognition algorithms both at flip submission and validation sessions.
Is it possible to easily hard fork to a different kind of human validation if somehow bots learn to solve picture flips successfully?
Technically any binary test (with two answers) can be a replacement for flips without a blockchain hard fork. It is enough that this new test is supported by the Idena app that participants use during validation sessions.
Cultural diversity
As people from many cultures join Idena, how do we approach issues with different cultures creating flips only they understand?
We believe that the diversity of the network should provide sufficient economic incentives to create flips that are understandable for everyone regardless of cultural background. Culturally specific flips are penalized if the committee is sufficiently diverse.
Bad flips very often are just bad flips regardless of cultural or regional biases.
Technical questions
Community is often wondering about scalability. When will sharding be implemented? Will it be region-based?
We are working on this currently. First we plan to implement sharding for the validation session. The second step is the blockchain sharding. Every identity will be randomly allocated into a specific shard based on their public address.
Are there any alternatives to sharding that you considered? So far each and every sharding implementation we saw turned out to be centralized and prone to attacks.
Proof of person enables sharding without sacrificing decentralization. At the same time, we believe that only sharding allows horizontal scalability.
What is the exact implementation used for the BFT protocol underpinning Idena’s consensus (pBT, Paxos, etc.)?
Idena’s Byzantine agreement was inspired by the VRF-based approach first proposed by Silvio Micali. It’s a leader-based consensus with fast finality for every single block. Randomly selected participants generate block proposals and broadcast them into the network. Random committees are selected to reach a consensus about whether to include a candidate block into the blockchain.
What is the theoretical limit of nodes Idena can support during validation sessions at the current network state?
The total number of the validated participants is limited by several factors:
- The number of blocks in which users’ flip answers should be mined during the long validation session
- The size of EvidenceTx transactions with which nodes vote for the timely submission of responses by other participants
- The network throughput at peak load during validation
We estimate the limit at about 8–10k nodes. Further network growth will require sharding.
Do you think Idena can suffer from block propagation time when the network becomes large?
Of course, the load on the network and on the node will increase. We are working to ensure that the node stably copes with large blocks.
Why the results on scan.idena.io start from Epoch #007? What happened with Epoch #006 and those before it? Premine?
Unfortunately, the data from the blockchain explorer was lost due to the hack of the indexer database at the early stages. To ensure it won’t happen again, the indexer database backups are to be shared. The latest backup is available here.
Idena uses IPFS to transfer flips, but IPFS mainnet is not launched, do you see any risks to use it?
We consider IPFS to be a promising technology, by using it we not only test IPFS but also contribute to IPFS bug fixes. Also we do not use the latest rc-marked (release candidate) IPFS versions.
When do you think smart contracts go live presumably?
We are still focused on the core part of the protocol. No ETA for this yet.
What is the size of the block and how many transactions (approximately) one block can have?
The maximum block size is 1 MB. This is approximately 8–9K transactions, depending on the size of the transactions. This roughly corresponds to 400–450 transactions-per-second peak throughput.
Could you point us to the part of the code where the VRF-based BFT protocol is implemented? Also, where in the code can we read more about the backbone P2P architecture and the encryption scheme for flips on IPFS?
Copyright
Are you not worried that 99% of images used in flips are copyrighted material that is used by the network for mining profit? What protects the foundation from a lawsuit?
Strictly speaking, there is no way to control the copyright of the images used by the network. However, most of the images are small-res and encrypted. Last but not least, Idena has no entity to be a subject of a lawsuit.
Will our flips be monetized since it is content creation? One day if it were monetized, will participants get extra bonus for early participation?
Flips are used only once during a validation ceremony. They are stored encrypted and can’t be used afterwards.
Wallet
When can we expect a mobile app?
At the moment, we are not developing a mobile version and will be ready to return to it as soon as the desktop client is fully ready.
When can we see Ledger hardware wallet support for Idena?
We are ready to co-fund Ledger support for Idena if the community considers it a necessary component.
Why don’t we still have the option of a cryptographically encrypted wallet so that we can’t kill the node/send coins without a password?
We have this task on our to-do list. However, so far there have been more critical issues we needed to prioritize.
How are flips encrypted on IPFS before the validation session and how are they decrypted?
The flips are encrypted according to the following protocol described on the website.
What if I build my own client to bypass the checks? Flips dismissing should be integrated in consensus for it to be impossible to bypass.
You may try to build your own client and node. The flip qualification protocol is part of consensus.
Do you plan publishing a more detailed whitepaper tackling technical aspects of the project?
We gradually update the Concept paper. The protocol is still in a hot phase of development.
General
Why such a large premine and why has Idena chosen to have an unlimited supply of tokens with inflation instead of a fixed total supply?
You can read more about the premine structure in our Community report.
We benchmarked with the premine size of other blockchains, e.g. Ethereum: The Idena premine is not that large in comparison to others. You should also bear in mind that we did not have any ICO and a vast amount of the coins will be used both for further funding of the project and community-based funding (quadratic funding).
Are we going to see proper experimental mainnet to mainnet transition in Q3–2020?
We are working on it heavily. However, we must admit that we will not manage to do everything that we plan by then. Now the development team is trying to implement the maximum number of protocol changes requiring hard forks. Further growth of the network will make it more difficult.
There are almost no marketing efforts, why?
We believe that the product itself and the benefits that people get are the most effective marketing tools. Given that the product is still under active development, increasing attention may negatively affect the healthy growth of the network.
When are you going to start with listing Idena on bigger exchanges with more volume?
We are in contact with some bigger exchanges that could be interested and are ready to support our project. Any inputs from the community would be helpful here.
Can you tell us the names of exchanges in contact which could possibly list DNA?
Some of them are Tier 1, we cannot disclose the information now.
Do you track the changes to the current price of DNA on qTrade or do you not pay attention to it at all?
Yes, we are aware of what’s going on with the DNA price.
What are your dreams for Idena? In a world where everything works out for you, what does it look like?
We believe in the 3.5% rule: the 3.5% of active citizens that can change the status quo. If the Idena network grows to a certain critical mass then we could see a qualitative transition of the network to a new level and the birth of something new which we have not seen before on our planet. We believe it could be the birth of new social structures governed by direct democracies towards a fairer world.
