Ethereum Swarm - Swarm Team AMA

i'd like to invite all team members to um open up their cameras uh because we are at the end of our event where we traditionally can you say that after a couple of times traditionally we want to answer all your questions so please can you all enable your cameras but please stay unmuted until uh until i um say your name because otherwise it's gonna be a little bit um chaotic um so let's start with the questions that we received in the zoom call um oh we have an anonymous attendee i always like that because it gets my fantasy uh going so the anonymous attendee is is confessing something he's confessing or she or he um it is confessing that uh um she or he hasn't read the entire 300 plus page book of swarm but i've heard of single owner chunks mentioned twice so far today actually it has been mentioned a lot more so can somebody give a short and good description of what single owner chunks are go raise your hands here rinke rico wants to enable his camera i can imagine this because it's a really oh this is funny okay so we're going to the swarm hq our centralized point of attention um and so please link it can you and your panel answer this question what is a single owner chunk okay basically a scheme to distribute part of the address space to to personal use which means that you own if you own a private key then you entitled to write to a particular part of the address space and what you write into that address is signed by your private key so validation of the address is is done by basically doing a validation on the signature and and and and and and the address that you you can write to is is is is a is a deterministically derived from your from okay so we have a problem with the audio coming from the swarm hq um i don't know if there's somebody there who can fix this uh we have i think the audio from rinky's computer who seems to be very far from the table um so if you can fix this for the next question that would be cool did you hear what i said yes i i i heard it because but i have any errors that are like really loud so i i don't know if for you know for the audience this was clear i can maybe uh recap it it's basically a space on swarm that only you can write in with your keeper yes yeah basically if you want to keep you can write to some some part of the of the address space and and thereby like create create locations where based on business conventional or some sort of consensus someone can look up content so this can serve as as the basis for for a communication channel like basically you can please step around the viewpoints where where you find the next message you know in in a chat or you can write feeds like for example temporarily sequenced related content like for example your your twitter updates your facebook updates you can write into a feed where the weather feeds updates addresses are derived from from a basic convention that uses uses your public key and some sort of sequence id okay so the person who was asking this question um well i don't know but so the person was anonymous it was asking this question and then suddenly we see someone popping up in the chat saying i didn't want to be anonymous so i don't know if it's actually true that this person was the anonymous person but someone is claiming the anonymous persona right now but i don't see any hashes or anything so i don't know if you can trust it anyway let's go to the next question maybe this is one for gregor um gregor what is this amazing swarm hq background music is it even coming from swarm hq i don't think so i mean i don't know it's an intimate pirate radio station uh called blur.bus so it's also i mean inspired pirated just by attitude because it's uh playing free music anyway so in the sense of just you know just being less known being more underground so player dot buzz blur dot bus so everybody go to blur the bus for an um free music okay so let's go to um other question um here we go so oh this is i like this question so how does uh i think i hope victor silence is good otherwise people go to janus um how does be better compared to ipfs do you have feature clarity now or is swarm hq able to answer this question with good sound so this sounds better can you say something well can we answer here yeah it's still very silent so i would i would suggest that the ammo should answer this question jonas can you unmute please yeah sorry so the question is uh do we have with swarm be better do we have feature parity with ipfs uh no we don't do not have we do not have that so how many people in this group would say we have it and how many people would say we don't have it sorry it's impossible to do a poll like that um so swarm hq i hope your science is better now because it's like impossible to hear what you're saying let's try it again you know if you're not me maybe i can comment on what i think is similar different so feature parity is is not the right word because swamp has there's a lot of features that ipfs don't and ipfs probably has some features that this one doesn't but if you really restrict your your attention to to fight stories and persistence then then one feature that that came with beta is is kind of uh making up for some for some of the advantage that that ipfs had namely that at least your own node restored stored the data and therefore the last resort if you if you're you're storing uh was was registered in the main register was online then your contents can always be retrieved we will grow by pinning basically we we resolve the same um so at least your if you if you operate voluntary this recovery mechanism this protocol that that after failing to retrieve your chunk from from itself canonical location in the network it falls back to to advertise speeders and recover the content then the content will also be repaired in the in the in the proper locality so that's in that sense we we achieved similar measures to achieve achieve persistence but i have to have to mention that these are really good for ipfs and and for these they're not really meant to be much more than stop stop get measures or really emergency fallbacks in case of or spheres and and most both in case of five when and that's one it's it's long-term storage incentives that is determined to solve the problem of long-term storage that uploaders can choose to to to calibrate how certainty how much certainty without uncertainty and what what period thanks they want their fires to be preserved these these features storage incentives will come in the next three days yeah and maybe also has one example of something that um swarm has but ipfs doesn't currently have uh so ipfs is not really you're not really uploading something to the ipfs network what you're doing you're announcing to the dht from ipfs that's people can find certain resources in your location so there's not really something as uploads and disappears because you're not uploading to the ipfs network doesn't exist um so in swarm really you're uploading something to a network so what you can do in swarm and not in ipfs is upload something and go offline that that's that's an example but i think really this year two different bees people think about them in one go because it's both decentralized storage but they're really solving different things and doing it in different ways and that's why feature parity on series is maybe not the right question and i should add that maybe in the future the two projects will develop apis by which they can complement each other so actually we are looking forward to a time when swarm content will be available through an ipfs namespace and conversely where ipfs addresses will be it will be possible to use swarm as a back-end storage for ipfs content i think that is not of the table so these two projects are in some ways complementary to each other okay that's amazing to hear it seems like the world is not enough for swarm um so one question i think maybe the same uh group of people can answer what will swarm 1.0 have that beta still doesn't have and i think you can maybe address that yeah of course so um what i mentioned in my presentation um really the big thing of swarm 1 0 will be the storage incentives and apart from that we will iterate on the features which we already released with vida to make them more useful and yes sorry to interrupt you but this is i think this is important to address so are you talking so you're so right now we are talking about uh a bus token that's on girly network right it's g bus um so what will be in 1.0 then so the 1.0 release uh will have bcc tokens on the swarm sorry on the ethereum mainnet really yes cool yep [Laughter] i believe you i believe you it's just like it's that's really cool um so um what does this mean does this mean that everything that people do now to help us out or to run notes that it's basically worthless until we have 1.

0 yes and no um like tokens on on the gurley ethereum network are are tokens so it's it's on a test network but it's still on on the blockchain which is uh maintained by people so definitely if you're going to run your your swarm notes and uh earn girly bcc tokens i i cannot promise anything here but internally we have been thinking as well about what we can do with that because definitely we are really valuing uh people who are already starting up their busy notes or sorry the swarm those right now and with the girly test network right now we have a way how we can actually validate and quantify uh that you have been contributing to the network so okay thank you thank you thank you for that i want to go quickly to gregor granger are you there yes so yeah gregor i mistook your background for ones and zeros which is not true right this is something else oh can i quickly point out igor igor kraft is here he's the guy who who is doing all this amazing artwork and and um visuals and art he could just wait for a second i don't you know it's it's i'm not going to ask you to to start talking about all this because it's you you should have your own uh event i think but gregor what what rinky just said is this legal even it's because he's basically saying like we're gonna have test net tokens that are sort of valuable like i'm not saying that i think you just did but anyway we'll we'll do a replay later so gregor but you know it's like i i need to say like please whatever i say don't consider it as a financial advice don't consider this is a legal advice and i cannot deny or confirm it that's awesome i just love it when you say that it's it's like you push a button and then the sound clip from gregor [Laughter] anyhow so let's go back to the ama um yes this is a simple one like like um for danny can you please share a link to the description of the dream protocol i would love to read it as well so i think it's a very valid question so if you can just share this somewhere in some orange language description of the dream protocol will be part of the book of swarm as you know the book of swarm is a living document that we keep extending and maintaining so the ultimate source for everything swarm is and will always remain the book of swarm so that's where the description of the dream protocol will be there will be soon a draft proposal for it online uh there's i'm not sure that there's already a link which we're ready to share but there will be really soon and also going back to the previous question maybe just maybe the dream protocol will also be part of form 1.0 okay this is amazing you put three brilliant guys together in a room and you get a lot of announcements that's that's really cool thank you for that um i'm not gonna do this even um so uh the other question is uh this is a somewhat more serious note do i build on top of the higher level for instance bcc api or should i build on top of the lower level chunks api and i also and what is the difference between those and i would also want to add a little bit of a personal uh question and that is um i heard rinker talk about i think a directory endpoint or something like that so i would like to know more about that so who wants to answer that question yeah i mean i can i can answer it um so as you know we are with the b note we are releasing releasing apis on different levels um we have the apis which give the features and everything but we know that the story mute to drink it i mute it no i i can hear lincoln no it's not a problem i can hear it so we know that's that we as the swarm team only have limited know-how of what people can make with one like swarm is so so new and so big that there are probably a lot of things which we cannot even think about so that's why we also have the lower level apis and a good example of that is the chunks api where you can directly upload and download chunks from from your keynote and we really want people to have the possibility basically to even remake our dears api with some sort of second level javascript uh tooling and in the end this is also how we envision uh db project even though right now we we offer higher level apis we actually envision the b nodes in the future to do less than we are doing right now so right now we took a shortcut and having a richer apis just to get this uh this you know this wheel to start flying and people to start building on top of uh v um but i personally um would like b to be in in one and a half slash two years from now to to have it only the lower level apis and do it much less than it does right now and have a very rich source of developers and people who build the second layer tools that can do what leaders right now but can also do much more yeah okay thank you yeah totally do you think in fact in fact the real world is to have a dslr which can which can kind of reformulate all these higher level structures with the help of lower primitives and that that the node would run and it's kind of a i think it's a good plan in the future because reducing the scope of the critical component is always a good goal to have and that they're thereby offering plurality of the high level tuning but what i what i also need to need to say to answer your question which one you should be doing that although i did this that's true that you can imagine various very multiple different hybrid apis and high level ways of for example represent files do encryption represent directories however a lot of research and trial and error we've been through and to to converge on the current solution and i think it's far from trivial how to do things and therefore we offer this this know-how and this is this particular way of representing files and directories and and if you if you want to do only that i would definitely recommend to use the high level api because why not that's why we're offering it and then offer the door everyone even well if you want to experiment with different this structure representation of files or or or or use user we have a use case where we don't even talk about eyes for example some sort of back a front-end to a block device or something so this is what the rule of three is meant to be thank you so much victor for that uh sorry so i was going to say that i would go even one step further and say that try to implement your application using the high level api and use the low level api even if you find some deficiency in the higher level api and even then try to build a layer in between your application and the low level api and maybe that layer that you build will eventually become part of the standard swarm apis so because eventually swarm will also store all the libraries that are used for swarm content and the nice way to contribute to the growth of swarm is if you find the deficiency in the higher level apis then you just build a library of your own and share it with the rest of the community i think that's the way forward yes absolutely and it and it's and it perfectly um also adds to the message of of the unity or like build this together come together um this is a question um so in the next question please raise your hand if you want to answer this question who is the ceo of swarm let me answer that question there's the answer yes so gregor please can you can you shine your light on who is the ceo sir sword doesn't have a ceo because ceo is basically a symbol of exactly the uh of the world that we don't want to build of the alternative that we want to create so um we do have in this regard i could i guess uh we have project managers managing directors and stuff like that uh so uh there's an executive team that deals with uh strategy but the goal of swarm is actually uh to really democratize this uh to extend uh all the parts into the environment and to grow to basically create a swarm so um in innovation management or business models science uh i guess it could be shortly described that we are aiming to be a so-called steel organization and what a so-called teal organization this is a part of la luz's deluxe classification uh basically essentially saying you know it's like ceo is the symbol of centralization so how would swollen rockets stop with a ceo yeah and this is not just a dream from your regular regular like i can speak from my personal experience that we have actually many leaders in this one project we have fault leaders we have leaders i don't know for what we have leaders for a lot of different areas and then i think there is room and space uh if you are a member of the swarm project and even if you're not a member of the swarm project you can become a leader and and hence we have a lot of leaders and sometimes difficult but i think it works really nicely yeah yeah so is it is it a little bit like uh like this what is this called this mythological beast with multiple heads and if you cut off one head it grows too new yes you know yes yes very good yeah i mean you know it's like when you mention ceo so the next question is what's the change in the shareholder value and again you know it's like what kind of question is this in the world of swan uh okay um well you know you're absolutely i mean no i totally agree you're preaching to the choir here i mean um this is an interesting question by jean emmanuel baliba uh are we still planning to start the ethereum blockchain on swarm i think this is a fun question that's that could have many answers and and let's make it not too long the answer let's try to do this in a couple of minutes maybe absolutely yes no that's too short a little bit you get a little bit more please yeah so so there was a there was a kickoff pilot of this project which we've been doing with with zahoor and it was it was very successful we worked together with with the uh trinity team like the python client team with with piper and basically got as far as uh having having the having the the header chain basically they had the chain served from small last one two two two trinity nodes and so that was the cross protocol uh scheme now we already have laid out the the actual design for how how the how the blockchain and the state would be stored and retrieved and then they see it so it's all done but we have to concentrate on the fundamentals now i mean this this this would have come in very handy but at the time when when it was taken off taking off we didn't have the funding to to support it now probably we will this this will continue with as time allows us to to set up new vestiges but the core team is really really concentrating and focusing on delivering the the the core of the product 1.2 but yes definitely definitely yes and this we have very very interesting features so for example it will it will directly be usable by light hands it will it will support basically it will provide a direct api serving as a decentralized uh explorer so so lots of interesting features nothing just happened perhaps not sync that's that's most important because of syncing lag or the the storage capacity and and time it takes for for ethereum notes to to sync up the blockchain is actually prohibitive at the moment and then really really causes a trouble in the diversity of node operators so let me just clarify this a little bit uh the developer of the light client and light server protocol actually happens to be a sworn veteran he no longer is part of this team he's working full-time on the elect land but and we're talking about joel falfeldi who lives in budapest and is our good friend so he has a very good understanding of swarm and because of that the future of the light protocol is very much interwined with the future of swarm and right now the bottleneck that the light client faces and which everybody who tries to use it will see that this the currently the load on the light servers is just too much and swarm aims to take a lot of that load off the shoulder of flight servers and serving state data directly to lifelines so that's thank you so i i and um so let's do two more questions and then we should wrap it up uh victor i i loved when you say like it it stems from a time where we didn't really have the funding to do it that seems like a distant memory now and which is like totally cool because now we're unstoppable i just want to repeat this because i think it sounds cool um this is maybe another question uh what about a multi-client network that's something that i've been thinking as well so we're now now our client is in the goal line um how do you say this language and so will we have like a multi-um multi-client network whenever you the audience got up your asses and start putting together a client please so yes in a way we we encourage to encourage people to to to to consult the the first dating documentation and specs and try to try to build alternative friends we originally planned it to be to be to come a bit earlier but it doesn't seem to happen i think i think it's very important to have to have a real consolidated network and and and and as ethereum showed it's very important to have a very strict protocol specification and it helps thinking about potential vulnerabilities and can can get four more security reaches easily so i i encourage people to to do this i mean as a as a kind of side more specifically swarming the browser because we ultimately want to achieve is to is to if for any user with a completely clean slate and completely naive user without any crypto just to browse to the gateway download the executable javascript live client and connect to the network directly this this is the dream situation and of course of course the promise of bootstrapping your your hip to holding from running running a swarm node is still on the cards and it's something that we're gonna do it very very 1.2 yes i think this is something we should definitely touch on um next time like how how basically running a note will you know will be a way to get started in crypto um i have to almost wrap it up there's one more question which i think is relevant and that is like you know um i don't know how much a transaction is now on ethereum mainnet but it's it's a lot so just even like sending money to someone it's it's it's like it's like just it's stupid let's let's call it what it is so what with you know i've seen all these checkbooks and accounting stuff etc but what good is it if we don't have like a blockchain that actually can handle this yeah the funny thing is that both with the postage stamp uh redistribution method of the storage incentives that we're planning and as well as the the the actual settlement of cash out of checks both are calibrated or can be calibrated in a way that that that the node can make sure that their transaction cost will only be a particular allowed percentage of the of the gain that they get from the transaction so basically you can maximize your administrative overhead and calibrate you from what's that way of course there's certain consequences so in the checkbook it is the case that there's certain risk zone that that you have to then put up with because because the basically accumulating checks and the anti the caching of the check is more than like i do 20 times your transaction cost then of course you implement principle uh allow yourself to be cheated by by by bounce check and this is this this is supposed to be we're not not the huge amounts and there's other types of disincentives to notes for for against behaving this way basically [Music] unfounded and i should add that the beta release is uh the first release of the b client where you can actually experiment with this stuff so the this is the first release where you can actually get down to you know balancing what happens on chain and what happens off chain and sure inside the team we are going to you know find out ways to present it in a user-friendly fashion that it happens as seamlessly as possible from the user's perspective that you never lose money okay thank you so much for all these answers i want to go to dickson nixon can you unuse these things how are you holding up what time is it in japan so 4 27 in the morning it's 4 22 yeah in the morning i have to go to work again at 9 o'clock i i think you're you're very brave to still be with us um so dixon you'll probably agree with me that we should round it up here no yeah okay last question this was the last question basically yes so i want to have a special shout out to personal need to ralph because you rocked my worlds tonight um to the entire team of course thank you so much for for doing this together um i think we totally have come together this evening and uh i also want to thank the attendees um who witnessed this long but very interesting event i think i can announce the next event which will be swarm live and then there will be again tons of new features presented by this team um from antwerp i want to beat you a very good night or a good morning for you dixon and uh um yes so we'll see each other soon bye-bye bye-bye thank you