AMA Transcript: Extended Edition
Hello everyone! I’m Dmitry Gerasimov, CEO of Demlabs. Welcome back to our AMA stream. We are here to discuss the most pressing questions of the day:
👉 BSC staking — what’s happening with it now, when it will be launched, and what to expect in the future.
👉 Token migration — I will give some details about it, what it will look like.
👉 And, of course, Wen Mainnet? News about the state of Cellframe development today.
One of the topics that many of you are interested in is, of course, BSC Staking. What’s happening with it now, what is the current status? I will tell you.
First off, I want to say that when we launched single-token staking, we were in a rush just to show that it is possible. It was important from a technical point of view, because it showed that we could create such functionality on our native system. BSC staking, along with ETH staking, work on BSC and Ethereum networks. There, we’re not demonstrating anything major, just our ability to connect to the other networks. Rewards will always be given on our native network.
So when we launched single-token staking, it ended up being a little undercooked. Fundamentally it worked, but there were some issues. So all this time we’ve been working to clear up any possible issues with BSC staking before we launch it. What I can say about the launch of BSC staking: just an hour ago I got a message that BSC staking passed QA. The team went through all the stages successfully, including withdrawing rewards. So receiving rewards should soon start working on the single-token staking as well, and BSC staking development is in the final stretch.
As far as release date estimates, I’ll talk about that later, and tell you about how I will make these estimates. But I can say that today BSC staking is complete, and a test version is working. We don’t want to rush, because we want to launch it in a polished form.
Many of you know we are doing a migration for BEP20 tokens. People have asked how it will work, so I will give a basic explanation. Like with BSC staking, we are already in testing, but it’s not fully approved by the QA team yet.
Token migration is closely linked to BSC staking. So, we had to wait for BSC staking to be complete, and only now we were able to start the final process of finishing token migration.
That’s because there are 2 ways of migration — LP pair or directly. So if you have an LP pair on Pancake, you should hold it, and it will migrate automatically inside the BSC staking mechanism. The second way, if you don’t have an LP pair, will be to use the migrate.cellframe.net site to migrate directly.
Development of token migration and BSC staking faced similar obstacles, related to the Cellframe node itself. Now that those are fixed, both of them will get to a launch quickly. Then there will remain some more work on ETH staking, which has a major difference in the way tokens are distributed and rewards calculated.
This is the big question, about the main event, when people can finally create master nodes and start to earn.
To start, let’s look at the networks we have already. There are four networks: three test networks, Subzero, Mileena, and Minkowski, and one production network: Backbone or Scorpion. To make it more clear — Backbone is the beta version of the Mainnet.
For now, all networks use the same software version, but soon that will change. Subzero will move to the developer version of Cellframe, and Mileena — to the release candidate version. In about two weeks we plan to have our first release candidate version. A new net will be released for hot fixes. These networks will be used to test various functionalities like setting up a master node after having staked, dApps, etc.
What’s left to do before we can release the mainnet: We need a Security mechanism (like Fisherman in Polkadot — we call it a Sheriff). These are observers with the ability to veto new blocks and give out fines. This is the “stake slashing” mechanism, where a master node’s stake may be levied a fine. So these will also be tested out on the testnet. We also need to finish the mechanism for delegating keys to master nodes at the time of staking.
Before launching our core mainnet, we will launch the KelVPN mainnet, where you won’t be able to launch a master node and verify transactions, but you will be able to provide the service. So we will make a kind of separation between service nodes and master nodes.
These are the stages left to cover before mainnet. There are also a number of small functions related to security, consensus, and services, and we need to go through an audit.
Last week we saw the first attempt to hack our system. We tried to track it as far as we could, and saw that it was a regular Ethereum wallet. It was probably an ethical hacker, and not a malicious attack. So that’s a good sign, we appreciate ethical hackers, and we will create a special fund for rewarding them. But we need to finish some things with the DEX first. By the way, we do have a DEX already, but we’re still tinkering with it, and there’s no market-making support yet. But soon, before mainnet launch, we will finish it, create a new UI for it in Cellframe Dashboard, and set up third-party market making. We also need to make a function to collect commissions — right now our DEX is commissionless. This will contribute to the fund for ethical hackers.
After the mainnet launch, we will split the team into two — one for stabilizing the network and for working with bugs, hackers, and the community, and the other to develop distributed services. We will start to host hackathons, and we welcome all developers who are out there to make yourselves known in our Developer chat on Telegram. When the whole system starts working, there will be a foundation for awarding prizes and supporting development teams.
Progress of Software Development Overall
Because many things we do are being done for the first time, this creates a very large number of unpredictable elements. Estimates that programmers give of time to complete a task usually need to be multiplied by two or three times. In some cases, it can even be 10 or more. In our case, it’s more like 2X when I’m talking to programmers about time estimates.
What I can do is give estimates in terms of order of magnitude, if something will take days, weeks, or months:
👉 BSC staking: days, or maybe weeks. We already have good results with few bugs.
👉 Mainnet: weeks or months. Not years, definitely. For sure, more than one month. Some team leads are saying we can make it by New Year, I’m more conservative, I have my own demands. And QA have their own demands. We’re pushing as fast as we can.
Let’s see how quickly we can get a release candidate, possibly in a couple of weeks we will have one, and it will go to QA to start testing. Our release candidate will have a new architecture, and if it works as I expect it to it will not take long to polish it. Of course something can happen to delay things, but we now have enough experience to give better estimates.
Why are there so many bugs? And what are the risks/rewards of holding CELL assets?
There are bugs because these things are being done for the first time. Most of our bugs involve moments when a user interacts with several layers of programs. Maybe there’s no bug in the base code of nodes and ledger, but when passing data from backend to frontend, an error occurs. Maybe a stake is complete, but the mCell emission doesn’t show up, for example.
So in terms of your funds’ risk, it is minimal. At this point, it is about the same as the risk of Ethereum being hacked by a quantum attack. Your funds are protected. And of course, if a quantum hack happened on Ethereum that would be good for our token! So the risks are justified, and staking rewards are also available as incentive. The earlier you get in, the more you earn.
What unique features does Cellframe already have?
We have a well-developed API for decentralized services. You can already start making decentralized services, even a cloud service, based on the Cellframe nodes. Lots of things are already working on the node: a VPN service, enterprise services, a web service with a Python backend. Our nodes are even working in the sensors of a high-tech fusion reactor. It’s a very special and useful piece of software.
One aspect is that right now you have to use the C language. There is a Python binding, but it’s not as popular. To take advantage of full SDK functions, you need lower-level languages like C, C++. Python is more for expanding on existing services. We will conduct seminars and hackathons to teach what is unique about Cellframe.
👉 We are expanding our communications internationally: talking about our development progress and about new projects such as ConfCall.
👉 We started a beta testing group.
👉 We are also participating in conferences. Just to give you an idea of how it is: if we went to every conference that we were invited to in November around the world, it would be necessary to make clones of all of us. And we wouldn’t have any time to work on the project.
👉 That’s why we chose one conference — Digital DAO in Almaty, Kazakhstan, on November 11. I will be speaking there, talking about fog computing in the context of blockchain.
Plans and Future Outlook
👉 Decentralize Backbone network.
👉 Decentralize the company. This is already partly done: we have entities in Hong Kong and the USA. We are opening a new office for developers in Kazakhstan.
👉Decentralize development. We are already talking to active communities in Canada and Mexico. So far North America has been the biggest source, but some proposals have also come from Africa and other regions.
👉 The overall outlook is very positive. When the next bull market happens, we will be ready.
How resistant are you to hacking?
The hack happened recently. We defeated it. This was a big concern for us, and we prepared a lot for the attack, constantly doing reviews of the code. It’s working so far. We have already been working with some core aspects of the code for 5 years, and that’s a long time to test the security.
The next few years will be the most critical. We have some aces up our sleeve to get through it. Some aspects of centralization we will keep in place for a few years for added security. I would say that we have an above-average level of stability.
How good is the C language in terms of security? How safe is it compared to other languages?
👉 Security is not in language, it’s in the programmer. This is known.
👉 In the Solana hack, the element that was hacked was written in Rust, which is considered to be a highly secure language. But some problems in Rust were revealed.
👉 It was a supply chain attack: there were too many third-party components. I would say that in Fintech overall, third-party components should be used very carefully. This is a rule which, if broken, can lead to hacks no matter what language you use. There are many such rules that we follow.
👉 Ultimately, C has about the same level of security as other languages. It all depends on the programmer.
👉 Recently, statistical analysis made C more safe. Machine learning is used to automatically track down vulnerabilities.
👉 Overall, I would say that Cellframe has a higher than average level of security.
Can fog computing improve live multiplayer games, with reduced latency, better connection, security, graphics? Does Cellframe have advantages for developing metaverse/P2E versus other networks?
Not for graphics; that’s a question for video card makers. But latency, yes, that’s something Cellframe can definitely improve. And it will be beneficial for the metaverse, since it requires sending and receiving a large volume of data.
Let’s look at the two most popular fog services that exist today: Torrent and Tor. These two services were basically created to meet users’ data transmission needs.
The popularity of torrents seems to be tied to legality. But even if we made piracy legal tomorrow, we would still find that Torrent downloads faster. I was recently downloading a torrent (a legal one, an installer) and I reached a speed of 1GB, even though my internet service was only 0.5GB. Somebody in my building must have been seeding that same installer, so it was downloading from a source very close-by in the network.
This is important for games. What we’re talking about is specifically fog services, not simply decentralized services. A company won’t put a server right in your building, but maybe your neighbor can host something. And it’s not just games — any type of content delivery service can achieve optimal user experience with fog.
The biggest thing holding back fog services today is insufficient motivation. I’ll be talking about this in my speech in Almaty. The biggest source of incentive now is a barter system. Once a better rewards system is in place, they should take off.
There’s also the catch-22 that the quality of fog services depend on the number of users, but the number of users depends on the quality. In part, we are solving this through our staking program. We are incentivizing enough people to join the network so that when we launch, good quality is already available.
Will there be an auction for a KEL slot in CELL? Does KEL get a free slot? Or will they get enough CELL to reserve a slot without crowdfunding?
The KEL auction that began earlier was frozen. Kel did not receive a slot yet. The auction will be restarted, and will be taken to completion as planned. You can participate, and there will be a KEL airdrop tied to it.
This auction is tied up with the development of Staking and Bridge. When the Mainnet is getting ready for launch, we will also relaunch the KEL auction at the same time.
For ConfCall, we have not decided yet if it will have its own token. It’s not ideal for one company to have many tokens, since exchanges don’t like it. So with ConfCall we will first make the product, then launch it on an enterprise level, and try to sell it. At some point after that, with new investors coming on, we can talk about releasing a token.
How anonymous are fog computing operations? For example, if somebody is streaming a UFC event, is it possible to identify the person who uploaded the video?
Any activity on the internet is anonymous only up to a point. How anonymous is Tor? If you’re hiding something small, Tor is anonymous, but if you’re doing terrorism or something they will find you. So fog is more anonymous than the cloud, but still not perfect. Tor is working on it, but it’s not 100%.
Is there any way to launch a master node without staking?
No, there is no way to launch a master node without staking!
Thank you very much for your attention! Until next time!