Freedom: A Browser for the Decentralized Web

Transcript

[music] Wow, right on time. Uh, great. Uh, we're starting exactly at 20 past. I'm very happy about that. So, uh, hello freedom lovers. Hello D web lovers. I'm going to talk a little bit about freedom, the browser for the decentralized web. So, decentralization is great. Uh, but we can build all the tech we want. uh if we don't have user participation uh either voluntary or involuntary uh it's not going to work. It's not going to be decentralized. So um if users don't run their own nodes, don't verify locally, there's not going to be any decentralization soon. And um yeah, we see that with the current state of

things, we have uh decentralized protocols that are running uh through centralized gateways. So the gateway goes down, the RPC server goes down or let's say in the case of the D web, some DWEB gateway goes down. Yeah, we don't have a decentralized web and it doesn't need to be that way. And um yeah, by by using decentralized services, we um reintroduce the the control and the the trust that we wanted to give up with the Dweb that actually in underlying layers already works. Um so we just kind of hand it back to uh Amazon to uh cloudflare that runs like uh decentralized infrastructure behind gateways. We don't want that. And I was when I saw this state after looking back

at it after a few years I was like okay this is uh crazy. This is strange. Why do we do this? And um so we need to make users network participants. Currently they are not. They are u yeah essentially trusting the the ways that are given to them. This is like average users right? If we have D web inter interested users, they might tinker around a little bit with things, but in daily use, they will use Brave, Chrome, uh, Firefox and therefore very centralized infrastructure. And again, it doesn't have to be that way. So browsers defining the protocol usage is sort of um yeah, again shifting the power and making the browsers very very powerful. We see Google and other

centralized uh companies using this power and kind of shoving protocols and things down people th people's throats and uh again there needs to be change. So how about a crazy idea make the browser a node and include a note to the peer-to-peer networks directly into the browser. So very simple all the control all the verification uh all the sort of reliance censorship resistance resistance and also sort of fault uh tolerance um if one one gateway goes down it doesn't matter we have many nodes to take data from so we just put that all in the browser and that's how freedom was born simple idea kind of a a

proof of concept first it does exactly that it takes swarm nodes uh like a peer-to-peer file network and IPFS same thing is kind of protocol neutral in that sense and puts it into the browser itself runs local nodes that connect directly to peers out there so um names are also resolved locally partially mostly so that's one of the bottlenecks actually that ENS heavily relies on smart contract verification and that can't really really be done sensibly without running a full node locally yet or even on a mobile phone it's probably impossible So that's really an area to work on, but the actual content and data delivery is already very very

decentralized within Freedom. So if you want to connect your beautiful ETH name or if you want to publish anything on um like an ETH domain that is a website, you go to ENS, you set the content hash and this selects the protocol. the content hash has a protocol switch swarm or IPFS or anything else that comes in the future and it handles that quite gracefully and freedom knows exactly how to read that and then kind of connect to the correct node and forward the content. So yeah, the content is fully fetched peer-to-peer. So, it's kind of a dream and it's it's a beautiful experience and you will see it in a second to just type in a uh IPFS content ID or a swarm hash into the URL bar and boom, seconds later the website loads

from all the peers that you're connected to. So, yeah, let's get to the demo. And uh this is freedom. Uh beautiful. Let's me let me put it full screen. And as I said, so we can just start with retali. So we just type the ETH domain in here as any normal domain and it's unavailable because we are not connected to the networks yet. So it's a normal browser. Looks like a normal browser. It's kind of modeled after Chrome. Uh so we start connecting to IPFS and then to Swarm and you can see the nodes kind of counting up here. Uh IPFS is kind of leading that race at the moment. Uh but Swarm is getting there. So let's get

back to Swarm in a second. Uh once it has a first node, then it kind of propagates further and kind of gets other nodes information. I hope it usually does that. Come on, Swarm, you can do it. Um but yeah, let's continue with uh Swarm really doesn't do it. Interesting. Uh so we can try Vitalik again. And yeah, there we go. We have Vitalik's website and uh that's essentially already what uh what the browser is about and swarm is still not there. Interesting demo effect. Um right. Uh we any ideas?

>> Yeah. >> Switch back to light mode. >> Ultra light. Okay. Let's try that. Oh, maybe it does come. It does not come. uh in the settings okay and then we relaunch the note so this is essentially controls the node in the background what I'm what I just what we did is like we bundled the existing node implementations and we're shipping it with the browser so and then reconnecting and now swarm Nice. It worked all the time, of course, as it always does in the in the in the

demos. And oh yes, we got the first be. So, um yeah, essentially what we can do now is we can also uh uh switch switch to swarm sites. So, we have uh of course uh this beautiful website mine.eth teeth which is uh resolving to a swarm hash and it is essentially a 3D game that runs in the browser and it's uh yeah it's it's served on swarm and you can see here on this uh globe roughly what the swarm network looks like you have like big peaks in in a data center in Germany and then one in in uh Finland I think and then you have like the nodes everywhere and that's kind of where the data is loaded from so that's pretty cool and um I think this the publishing

won't work now deploying right >> well we're sorry so um that's all good uh but we also want to write to the network so far we have only been reading so and then from uh writing to the network we switch on the wallet integration also to kind of give identity to us so we have essentially like a private public key pair in the browser that behaves very similar to what maybe meta metam mask would do but this is a first class citizen in uh in the browser and it comes as a sidebar and from there we can kind of uh connect to websites authenticate with websites and also upload content. So if we go to freedom

uh freedom and the publish website. So here you can see that uh we can publish uh a text. For example, let's say uh hello. >> I can do it now. >> So I can publish now. >> Oh, okay. Interesting. [laughter] Floorian did this integration. So, thank you very much as you can tell. [laughter] Um okay. So, we can publish uh text. Uh, hello ETHC. And what it does now it uh connect.

>> Okay. Okay. Okay. Yeah, I think I don't know why the light is not working here. It was working down in the hacker's den just five minutes ago. [laughter] >> Have you tried turning it on and over again? Anyway, so you get the gist. Um we can publish things from here and uh we'll give it a little bit more time and uh the uh the idea is that you are not just a reading node of a network but you can directly participate using the same software. You can dump your files, you can dump text, you can dump whole websites and it will then be available within seconds as a swarm hash or IPFS hash and which you then can send anywhere to anyone in the world. We're not connected again. Connecting again. I think it's

No, we're light mode. So, uh the same text again we can just publish just to show you that uh hello Florian and we just publish that. So you see how fast that is. So first of course it publishes to the local node which then propagates to the network and uh when we open that in new tab so we have a BZ hash here which we then can just uh yeah send to anyone and it will uh easily uh yeah just kind of it's censorship resistant essentially it's censorship resistant information. So um there used to be this art project uh people don't want you to read this project. So this is exactly that you can just um yeah this is will sort of live as long as this part of the storage lives and will

be accessible and uh what we also can do is now we can link this to a um to an ETH domain and for that you go to your ETH or ENS domain service and you can just replace in the in the records for that domain uh the BZ hash we just paste it Here we say save and it's inverted. Yeah, it does. It needs this. And then we can save it. It takes a while. Open wallet.

Good old Metam Mask. Um, yeah. So, ENS doesn't fully work in Freedom yet. It should work. Uh, I mean, this is something on the ENS websites uh site. So, that's why I'm using the ETH limo site. They have a decentralized website which doesn't really uh function yet. sensation translation this was transmitted exactly so we see this uh host name here now if we paste this then back into freedom browser then uh we get the exact same thing so it's now connected to an ENS domain and this is all in real time we can send this ENS domain to someone and it fully uh works uh in a decentralized way um so putting this all together we can then also go to uh a website that is

there we go. Uh so this is a decentralized Reddit and when I say decentralized it's fully decentralized. So like none of this is using any centralized elements. Uh you can go here and post content. Let's say uh to to those to this board. Let's go to the random board. We say submit a new link uh hello K. And then we just say oh we can just take a say a text uh this is swarm amazing and just post this and what it does then is a whole complex unlocked uh with touch ID. So yeah we have touch ID wallet in this as well and

then it's sending it. What happens now is like a series of updating of static files on the swarm network which then is being put together by a JavaScript application into a Reddit like interface. So it's fully it works without database. It's a trick or like a very sort of sophisticated uh chain and tree of JavaScript files or JSON files which then kind of replicates sort of the uh Reddit experience. And uh we'll give this a moment. Uh going back to the to the here we are going back to this. So that's what we saw and uh yeah freedom obviously is open. So you can build it you can run it

you can download the the source you can contribute to the source and more uh more importantly so it's a it's a a chicken and egg problem. Uh so if we don't have many ETH websites or I know not many decentralized websites, we don't really have a use for a browser. But if we don't have a browser, we don't really have a use for these websites either. So there are really just a handful of working um decentralized Diw websites there. And when I researched this whole thing, when I went down into the D web rabbit hole like about half a year ago, I thought, okay, there must be like a browser, a software or something.

It did not exist. That's why freedom came into existence. I'm very happy if there are other browsers coming following the same example. This is really like a let's say Mark Andreasen Mosaik 1993 moment where like a new internet or a new web is born. So really feel free to contribute convince other people if you know anyone at ens.eth to fix their website to work in a decentralized browser that would be amazing. So that's the kind of uh we need to get the word out that this now exists. it's possible and in the west we don't really have so much need for that for sort of I mean more and more actually uh for censorship resistant content but there is places war zones and so on so forth that are really really like relying on decentralized

infrastructure so that's where the cypher punk ethos comes in these people really need uh safe ways to publish and read content so what's next you saw already a preview of the right part and the identity part and obviously All this uh is not much use if it's not also available on mobile. So we will work on mobile implic implementations. The freedom part the browser sort of uh I candy around it is not so much the issue. They're really the node implementations or light implementations that run on mobile. Uh that's really what's what's lagging behind because we are currently not writing these. But there is uh early implementations of swarm and potentially also of IPFS that do run in a light mode on mobile. And there usually the issue is uh that apps

are being closed uh and and processes are being shut down if you leave the app. But we can work around those and I think once we have that mobile it's going to be a super interesting use case. Also local ENS and fully decentralized ENS because we are relying on RPC at this moment and uh the other day RPC for Cloudflare was down and kind of it killed a whole lot of uh DB or not D web decentralized web 3 applications. This doesn't have to be that that way and it shouldn't be. So yeah, uh Freedom Baby, uh that's our domain that is the gateway into the decentralized world. From then on, you can use freedom browser as as your daily news source. And uh yeah, join the Telegram community, uh

Freedom Browser. And that's pretty much it. If there's some questions or remarks, happy to take five minutes of questions. Do we have a microphone? Or maybe just yeah shout it out. I can repeat your question if you please. >> Yeah. So used to also support people. Yeah. >> Unfortunately he removed this feature in 2024 and saying that it's very few uh usage >> less than 0.1% they said. >> Yeah that's from their announcement. So how do you address this uh issue? Like I think um because this is probably really common.

>> Yeah. Uh short answer is we don't need to. So this is open source. There's no real business model at this point. It's not a company like a commercial project which Brave very much is. So they have to look at kind of what makes gives them most bang for the buck. Here it's not the case. And uh yeah it doesn't die with less usage. You see so many tiny tiny like tools open source tools out there that just survive. We don't run the infrastructure. We don't run hard we not we're not running hardly anything. We're just supplying the software to access the existing decentralized networks. So in that case we are resilient to the sort of commercial attacks. Yeah. Next question. Yeah.

>> Yeah. Thanks for the presentation. Nice work actually. >> So you you mentioned that you cannot work with ENS locally because you cannot run a node. Did not it's not true. You could use a stateless client for instance. Did you think about this using status client to really verify everything in your application? >> So um uh Helios I believe is is also an implementation of that. Is that what you mean? >> Yeah. So um we looked into that. It was kind of outside the scope for this because the the the trade-off of like using RPC for now for this phase was good enough. we are exploring this and we exploring other ways and there's also

um like decentralized or like a federated network of RPC that's already a step forward into like federalized land and not completely decentralized land. So yeah the answer is yes we would kind of definitely look into that. Uh we have been focusing on other things UI and and the the writing part of things which was most most important. So the team is quite small it's essentially Florian over there and me at the moment. So uh and it's really a passion project which we believe has yeah as I said great potential to really open up a completely new uh experience for the web.

>> Yeah. Next question. Um do you have any security options or feature for in order to prevent uh a user to access bad actor disinterest website where a bad guy could try to just to make you download a malware or something. Uh so filtering always has like obviously two ways at least right so uh if a filtering is in place it can be used for sensors very very easily and and someone can just say okay you have a filter in place just add this uh anti- whatever government website there um so the short answer is not yet not at this point uh we want to harden and there's a lot of a uh like the security model is at the moment

similar to other sort of browser wallets where it kind of asks you for permissions and Um it's a field of work so we we want to make it as safe as possible and we want to have as little filtering as as possible as well. Maybe some sort of a malware with clear malware filter that would be one thing but it's a slippery slope into sort of censorship land. So that's why it's also not the main priority of what we do now. I can imagine that maybe at some point legal pressure would would come. We will see about that. um saving users. Yes. But yeah, it's it comes at a price as well. So, do we have time for one more question? I think so. One question over there.

>> Last question. >> Uh hi. Does Swarm being built on Gnosis instead of Ethereum mainet add any friction? Cuz I mean you kind of need like two RPC or two nodes to be running for that to work. So that right there is the the only friction and it's not really being built on on Gnosis infrastructure. It's sort of patched onto it or there's sort of one payment gateway or the just this sort of monetary part. The swarm network is completely separate. It runs by itself. It just needs to be um kind of topped up with so-called postage stamps uh to to kind of supply storage. So that's really the only link there. And uh there are as I understand also um yeah ideas about uh integrating other

chains and other currency models. Um but yeah the friction point is really just that the the the one RPC uh extra RPC connection uh which we will have anyway by having multi-chain support which we have already now. So like Nosis and Ethereum are the chains now that that we're currently supporting but everything EVM compatible is also supported but in the future there could be other ones as well. So um it's not really an issue like in the practical sense and swarm is only one of the networks that we support. There's also actually radical support already as an experimental feature. Radical being a decentralized git um git technology uh that's already supported as well similar way. So we have a radical uh client and

and sorry radical note light note in there. Um so we sort we we treat these as plugins right so we plug in a new chain a new network and then then it kind of works for us so friction was minimal. >> Thanks for the presentation. Okay, thank you very much. [applause] >> [music]