DeversiFi: User experience challenges on L2with Daniel Yanev

hi hackers welcome back to xerox hack uh with me is daniel yanef he's one of the co-founders of diversify uh looking after the product and um today he's gonna tell us more about challenges in layer two ux uh so yeah welcome daniel the stage is yours awesome uh thank you antec for the uh introduction yeah so as you mentioned my name is dan i'm one of the co-founders of diversify and kind of one of the few things i do is look after our product um diversify is actually one of the first exchanges to launch on on layer two uh we made that leap almost a year ago now in june of 2020 so we've we've had a lot of time to learn a lot about kind of how users interact with layer 2 and what the pain points are there and of course what the benefits are but also we were you know as as being some of the earliest in the space we also experienced a lot of the um you know improvements and i guess kind of the the quirks of the of the very early days so um i'm here to talk to you about kind of some of the latest uh uh challenges around layer two and most importantly what we can do as a community uh to help improve those and uh and bring more people uh more people into our ecosystem so um starting off with um the promise of ethereum so um i guess the whole point um behind layer two is scaling ethereum so let's start talking about that um generally you know the the whole idea behind it and what gets a lot of people excited is this concept that uh we can have this uh financial system that's global that's distributed it's permissionless everyone around the world can use it without intermediaries um and that sounds um that sounds amazing and in kind of the middle of last year uh with the proliferation of d5 we had a kind of glimpse of that uh you know we were all uh swapping token sending tokens getting uh great uh great yields you know taking those funds out and putting them into other extravagant products you know taking uh taking out flash loans and and doing um all that exciting stuff so we kind of saw the glimpse um of that um of that financial freedom um and all those things we can do but now you know that has that has kind of changed a bit since then um and also i guess a lot of the people that maybe maybe are here and are full of the people that participate in the so-called defy summer um are very much like hardcore uh ethereum fans uh and hardcore crypto fans so if we're going to be talking about the mass adoption or even getting to kind of the early adoption of of crypto we need to think about the user as you know someone who is not full-time into crypto someone who uh you know doesn't necessarily have the kind of patience and um you know this and adventurelessness that we do so think about you know maybe some of your friends uh that kind of use apps like etoro and revolute or you know your parents or whoever so kind of how do we make uh ethereum accessible to those people so you know uh last year uh you know ethereum was great and you know our user is happy however let's uh let's think about what now actually the user experience on ethereum looks like for that user so so first off if you try and you know if you try and use d5 actually a lot of the time what you're trying to do doesn't happen so about one in one out of five times uh if you're trying to make a swap on unit swap or or other products i'm not singling them out it's kind of the same across the board that will fail so you pay a bit of money in gas um your traction will fail your market would have moved and and you're out so for kind of people coming from using traditional products that's kind of very strange um and not a great user experience then with the price of ethereum uh raising and more people using the network which is kind of a sign of success um it also has the he the drawback that that you know increases congestion increases the price to use the network so um in may this month the average cost uh to make a trade uh on kind of any of the device four platforms is about 80 which really prices out a lot of a lot of possible trades from the market and you know if the market's volatile and you have a lot of money uh to move around it's feasible but again you know thinking about our uh mass user adopting this not really not really the case and you know that's that's not really um something that they can sustain and we do have some kind of great lending and borrowing rates uh but you know if you were to open a 5 000 deposit 5 000 deposit on compound now you'd have to wait for two months to gather interest and rewards uh just to pay back uh the gas you paid to open it and of course two more months if you want to actually uh unwind that position and on top of that um you know there's generally the quirks of the blockchain if uh you know if you send your funds straight to the smart contract instead of uh via the deposit function all your money is lost forever you could be um in a product that gets rug pulled and again um you lose all your money and so on so for our kind of normal person user that doesn't you know that doesn't really combined to a user experience that you know people would want to use and it's worth using compared to what's out there now but so our user is now kind of unhappy they're not they're not putting up with this um but you know over the last year layer two has emerged um so generally um layer twos are you know it's a combined term for multiple scaling technologies that allow um ethereum to scale and have a lot of benefits from it so layer twos as i mentioned uh can be much lower cost or gas free um they can have instant transactions so none of that kind of waiting around for two three minutes for your transaction to maybe succeed which again is is much closer to what people expect to see it inherent the security of of ethereum which is one of the key reasons why we're talking about layer twos and not just going to a different chain which is an obvious solution to this problem but there are multiple reasons why ethereum is great as a security layer and why a lot of us believe in it but that's a different topic and layer twos can be private and they also can be evm compatible now this is kind of a a menu of the things that layer two can provide um no layer two as far as i know provides you all of these um we're not no we're not quite there yet but these are kind of like you know uh a subset of these things would be um included in kind of some of the layer twos that are live today so great our our user is happy um all of the problems that we um outlined earlier are solved um unfortunately it's not that the story is not it's not complete um as many of you know maybe you've tried using some products uh and we're not we're not quite there yet so first off let's look at why that is and we said like you know layer two things are great but how does one get to the layer two so for now the most common route is you would move your assets from ethereum onto a layer two because you know most acids are uh on the on the main chain so that presents itself a couple of problems so moving from ethereum to the layer two and back um there are a couple of gripes uh that we can we can talk about today and are general things that are kind of beyond the scope of one company or one group of people but more of a kind of community effort to educate and and create the tool ligand infrastructure so first off we start with wallet support um so layer twos work in uh different ways um and if we kind of take the general cases of rollups and um and sidechains the interactions with the wallets are actually quite different than layer one for example on on rollups zq roll ups um usually there's a different um signing curve than is on the is on the main chain which means that integrating wallets um is slightly more complicated and usually would require work on behalf of the wallet provider so that means that the actually the choice of wallets that you end up having on a layer to product usually is uh much less so it's about kind of like one to five types of wallets are supported on layer two products some popular ones you might have just one just metamask whereas in on a layer one product it's you know much easier to do that so you know if you go to something like one inch they support like 13 14 wallets and so on so for a user you know they come to to use one of these products and maybe their favorite wallet isn't supported so you know not not great for them um which excludes a number of people from the uh from from using that then even if even if wallets are supported they usually are not able to be supported natively so there's kind of some extra steps for the user to go through for example if we look at metamask which is one of the most popular wallets they don't really support these different signing curves for um uh zq roll ups so what products end up doing for example we do that at diversify uh i believe loop ring uh zk swap and stark and dydx just use darkware do similar things where basically a different key is generated for the user that is compatible with the rollup and then that key is encrypted with their ethereum wallet and that basically means that the user would see a message like this when they're first on boarding that's asking them to complete an action that they don't really know that they've never seen before because on layer two it's kind of a combination between a wallet interaction like you you're doing on layer one but then uh you know operating a normal product like a centralized exchange um and that's kind of like a weird combination of the two that's not that's not very familiar yet and uh putting these extra steps to the user which are unknown is scary so for example you can see these kind of different types of messages that you would get to kind of perform this process and that does care actually a lot of uses away um at diversify we see about a 12 drop-off uh in the user flow at this stage and when we ask people why they basically say i don't understand i don't understand what the step is uh and it's because it's to do with your wallet and signing things you know people are afraid and in user interviews people literally told me i'm scared i don't know what this is i don't want to sign it um and you know there are kind of ways to experiment with this and you know try and word it in better ways but still scary and for the products which are using a basically different chain like a side chain plasma or another layer one that uses the same addresses like bsc um they basically users will have to change networks so for example if you look at what it looks like for matic if you want to use matec you're basically presented with this scary message from metamask that asks you to add a network you know there's a lot of a lot of stuff that you don't understand a warning at the bottom with a red exclamation mark which again not great for the user um you know our kind of general normal person will not will not want to uh deal with that and and that's something that um you know again this comes down to wallet providers and it's not like for a single company to solve so that's something that you know first as a community we need to push uh to get um to get to that adoption some wallet providers um are quite cooperative about this for example we've managed to do a direct integration with ledger so ledger supports our um the keyword natively and the user experience there is kind of much smoother and you wouldn't know that it's kind of any different than using a normal dap cool so the the second thing is yeah now now our user is upset again um they're confused uh then the second thing is the onboarding cost so moving from uh the main chain to layer two um actually incurs quite a bit of quite a bit of cost so this month uh on average we're looking about fifty dollars per token to move to um a zk roll up um i don't know what that is for um the plasma and side chains but it's generally the cost of a smart contract interaction so kind of anything you want to do would be um around there unless it's very simple um now again if you're moving loads of funds around that's fine but for your kind of general normal user paying a 50 fee to just move one token around is a bit ridiculous um and kind of puts us again back into the you know unable to cater for those for those wider users um now what are the what are the ways of of catering for this now this is this is a pretty uh pretty tricky challenge to solve because you know that is gas that's the price to operate on the network and all the benefits of layer 2 is that you you know you know you no longer need to pay gas but that only applies once you're on it so then you know kind of having to pay to get on it is kind of a you know not a great start of your of your experience and you know there are ways to make contracts more efficient and there are ways to reimburse people but that's not really a long-term solution to this problem and with ethereum becoming more successful and uh more expensive although not today um that um that problem's only gonna get worse so the other way to think about it is um okay instead of you know just moving from ethereum to layer two um we could be moving from one layer two to another layer two and obviously uh back and forth between them so that basically means that once you onboard onto one layer two then if it's integrated with uh a few others and uh has bridges with a few others you only kind of experience that pain once uh and then you're and then you're kind of uh free to operate um in this nicer world now obviously a lot of funds are not on layer two yet so that kind of is a limited um a limited source of uh onboarding to start with um but most people actually stealing crypto uh keep their funds on a centralized exchange i know in d5 um we kind of don't think about that world that much but actually you know most funds are on uh people you people keep on centralized exchange um so that would be another source of onboarding people uh to layer two and of course in the kind of general wider economy most people keep their funds in fiat so as we were welcoming more people into the crypto ecosystem they will be looking to onboard by fiat so the idea is that you don't need to onboard from ethereum we can be onboarding users from all these different sources um now this sounds um you know good in good in theory uh in practice actually integrating layer twos with each other um and other places is quite tricky uh there is still not really uh an emergent standard of how to do that um and not really many examples of how it's how it's happened so um again this is something that um you know is something that we should solve as a community in order to basically make sure that we have a viable solution for people to be using ethereum rather than kind of moving to uh to other chains or compromising their security or compromising um on their uh on their centralization and um you know we here at diversify are working on some of these and kind of doing our part but when it comes to these bridges and these connections um it really shouldn't be something that one player does or is completely controlled by one party because that creates creates you know centralization and lack of redundancy in the system so hopefully in the near future we'll see a lot of bridges with a lot of operators and a lot of different models between these um and a lot more interoperability um so that we can kind of create this um system of uh you know of affordable and usable ethereum products um yeah as i mentioned here at diversifier we're working on a couple of these so stay tuned in the next month for um for potentially trying some of these out and that's uh yeah all i had for you today uh yeah thanks for taking the time to listen and if you have any questions i believe there's a chat here and people can raise hands so yeah please ask if there's anything you're curious about okay um thanks for sharing your presentation uh daniel uh yeah uh for any links that you would like to share how we can uh try to find you and reach you um yeah yes sure so um probably the best oh we have a question um so the best way to reach us um would be you know our discord um i will drop a link to that in a minute uh when i when i pull it up um where uh yeah our team hangs out there so if you have any questions about how rollup works um how you could uh build an mvp of a bridge with us or anything like that um do let us know and uh um and we can help out with that other than that you know for general news we're on twitter we announced some uh pretty cool news yesterday yeah congratulations on that thank you uh yeah so we have uh we have a question from the audience uh i'll let you read it out sure okay so we have a question which is why would someone use various layer twos uh or do you think that there will be an emerging winner platform um yeah so this is a good question so basically why do why why do you think there'll be bridges between layer twos instead of there just being one winner and there's been a lot of speculation about you know some similarities it might i need over um and i guess basically the answer to that is that you know if we go back to uh kind of the slide that i showed with the different benefits a layer two can have um no layer 2 has all of those benefits yet so um there's basically um layer 2 which is better for certain things than others so for example um you could have a layer that's very evm compatible and hence you can have kind of uh composability in interesting products but that doesn't tend to be as cost effective so you know that would be only kind of let's say 10 times cheaper than ethereum whereas something more dedicated ethereum so if you're not using fancy products maybe you want to uh you don't want to you don't want to be paying that uh some layer twos like ours are private uh some are not uh which have we heard their benefits so um basically i think the the application that you use will work better on some layers than others um so it still makes sense to uh not have that centralization on one you know absolute winner layer yet yeah thanks for answering the question i think uh do you have any other questions if so feel free to share them in the chat um yeah next up uh we're gonna have yustena from uh ifworks teach us how to build ups like a pro using used up and that's our framework but uh that we enjoy using a lot uh yeah daniel thanks a lot for your time i appreciate this uh and yeah let's uh let's keep in touch on twitter and uh discord awesome yeah uh yeah thank you for having me and yeah thanks for putting the conference together looks really good so far thanks daniel see you bye