Gavin Wood - Polkadot - Ethereum London

you cool so if you don't know who I am my name is Gavin wood I co-founded aetherium with Vitalik and a few other guys i was the CTO of a theory the first 12 or 18 months and i coded a lot of it helped design it and all that sort of thing i have come here to talk to you today about some of the new stuff that we're doing so i started a company called parity it was at the time named f core since changed its name to talk to basically create and continue the development of aetherium and aetherium like stuff so blockchain crypto and pushing out into the other areas of the generally sort of decentralized trust free ecosystem now I'm going to talk about and we have some interesting stuff that were working on now and hopefully you'll go home a little bit more educated on what it is that we're trying to achieve and so yeah quick show of hands who's heard of etherium Hey okay good now on there I got lost who's heard of who's head of poccadot okay few people cool whose coders how many of you can code in like PHP doesn't count cool okay about half so I'm going to try and keep this this talk like non-technical I don't think it's actually going to be that beneficial to go into the technical stuff if you want to know ask me some questions at the end I want to really sort of get across basically the goals and where it fits in to how what my worldview is and how you know we at parity generally see the world going okay so yeah how did things begin well I'm going to cut out most of history I don't say in the beginning of the web so we need to really understand what the web is you know what what what what did it what does it do what's its job was it there for now I'm going to going to make a point that the web is basically access and it's access to information okay so it's an information access invention mechanism and we see some of the stuff on the web and we can sort of understand that it is is there for information so it's like we don't we get media at the news you know we get the websites we get people's home page and get information on them and they're sort of various lives and we have like Wikipedia people come together and create you know a cache of information that's supposedly and pretty accurate but I'm not convinced it really it works that well and I think that we see increasingly these kind of nexuses of information form and these are increasingly becoming not corrupted exactly but biased let's say towards very particular editorial styles or content guidelines so maybe it's something else and I think you know there are other aspects to the web than just pure information retrieval so one of these other aspects would be economic activity so a lot of people now and use the web for economic activity back in like the late 90s when I started using it of course this wasn't the thing at all no one went on the website you know buy groceries or order goods or whatever and these days it's a little different you know or the rise of Amazon and eBay and all the rest of it we do tend to use the web for economic activity but the interesting thing is that the web itself doesn't provide access to economic signals because the web itself is an information museum right actually what the web does is it provides a communication mechanism to intermediaries in society that then tell us what they're doing on our behalf right so the web for example provides access to a banks website and the bank holds our money for us it tells us what's going on maybe some transactions are in progress maybe some are finished and we can we can go to the bank when you say hey what's going on in my account they all say well turns out this is going on your account so thanks very much Mike or we could go to eBay and say oh wow what's on offer and they say well this is on offer oh well I'd like to bid on that please but really all we're doing is we're communicating with eBay we're not doing actual economic actions directly right it's all very indirect and that's a very important thing to note because it means that there are other economic actors that are sitting between us and what we actually want to achieve so in this respect the web becomes it kind of mirrors Society in society we generally have to rely on these trusted intermediaries Thanks eBay Facebook's of the world now the web has the potential to be a very transformative platform but if we rely on standard kind of trust bound mechanisms that we use in society to do economic interaction then it will never become that progressive disruptive platform that it could be so now we're bringing on forward three so what is web three so web see was basically a originally a notion of how aetherium would be used in the by end-users right so not so much by the people that run websites blockchain exploring or not not so much for like I don't know whoever unlike retailers that want some sort of payment mechanism but rather by directly users trying to use aetherium yeah and it sort of grew into more than that with the ax with the idea that maybe there could be other block chains that be used as well maybe maybe a single blockchain isn't quite what we want so web 3 in this sense is really pushing forward into trying to provide an alternative way of imagining how this information and economic activity access platform can work and rather than trying to take on the the trusted intermediaries the current stablishment of information and economic access surely it's better just to create a new way of allowing access to information in the economy that basically obsolete and in this sense what web 3 is doing is it's an allowing people to interact with each other in and of themselves by providing them with very particular and mathematically based guarantees on what they are what the system is providing to them and these guarantees in some sense mean that they don't have to trust anyone in particular so how are we going to how we're going to push forward in that well the idea is that we want to decentralized the web we want to take the I what the web provides at the moment which I would say is a massively multi-application platform and we want to reuse as much as we can from the current web because you know reinventing the wheel really isn't the point of this we want to be as inclusive inclusive as possible with as many different groups who want to push forward into this direction and we want to enable users to transition from what they're used to to what they can have in as easy a way as trivial away as possible so I mentioned before the guarantees and this is really what web 3 is all about about giving guarantees so the web gives you relatively few guarantees if I go to the HSBC site there are no guarantees that that it has for example being hacked and that the transitions are the transactions it's showing me or actually are actually valid and moving a little bit further it could be the case that you know with HTTP I mean I can be guaranteed is actually a chess piece you're at least guaranteed to some degree it's been HSBC now of course there are instances where certificates don't actually work because issuing agencies behind them are themselves a bit corrupt or have been hacked but in general this at least pushes into the pushes forward the problem of not being able to trust who you're talking to but still the only thing we actually use them for is to be able to trust that the intermediary that we think we're dealing with is indeed that intermediary it doesn't help with the wider problem of the incoming to trusty intermediate that they're doing their job we don't get that guarantee we don't get the peer level guarantee so what we're coming into now is what guarantees do we need in order to be able to create applications that do much the same thing as what we expect applications to be able to do on the web do today but to do so in a way that can kind of basically cut out all of the economic deadweight the intermediaries live on I've identified three so the first is identity so identity is basically the ability for me to prove who I am and to know absolutely that who I'm talking to is who they are and I've seen this in a a n HTTPS sort of there is a registry we've got some certificates the the you know the root level entity as said that this is in fact a company who's registered as X Y Zed Co I mean in the peer level at the peer level so the ability for me to know that I was talking to this this this bloke and this bloke is I'm now talking to them again yeah so maybe I know that this bloke actually has multiple identities that sometimes they use in one instance sometimes they use in another instance maybe one time they use in one application another time in another application and I know that they're connected so it's an identity platform to first second is it's an accounting platform which is to say basically that the transactions what I expect to happen when transactions go through actually do happen so a trivial example of this is in Facebook now facebook supposedly counts likes likes from various different people on its network and they all get collated into a single button and I can if I click the button it goes up by one and I feel yes I put my vote in that great now what's to stop Facebook from increasing that taking a bit of a backhander and increasing that so that you know it has a thousand more likes than expected nothing yeah I mean we trust Facebook that they don't do that but in principle they marriage do that now maybe Facebook is is beyond suspicion not sure about that but certainly there are many other platforms that are not so accountings really about ensuring that that transactions are executed as we expect yeah you may even call it expectation management and the third is data access so data access is a very interesting one let me get only a fourth yeah okay okay so we're going to manage this expectation management so in this so in this sense we're managing expectations through the blockchain yeah what pain what pain what pain what chain gives us a means of determining that transaction will be executed validly going forward so if someone says I will pay you X Y Z I can be sure that that that payment will actually go through someone says I've done my KYC yes I've done my K would see I am blessed I am and that has gone on on a blockchain then I can now go to to see that transaction and see that in fact yes the KYC was went through from out from an entity that I recognize and indeed that person had is indeed a citizen of this particular country and of that particular name so we get to the point now where we have to talk about given that we have blockchain to be able to manage this kind of expectation management what are the what are the risks that we find in the future and I want to just kind of draw attention to the idea that in the early 90s it was assumed that the internet the web would basically be centralized society like wow we've got this new technology it allows people's computers to talk to each other this is amazing this is going to change everything I mean you know it did change a lot and to be fair but it didn't change things in exactly the way people expected there was still centralization largely for the reasons I outlined and so what stops blockchain from falling into that kind of same trap or some corresponding trap and the interesting thing I see I see here is a kind of self centralization happening and I kind of liken it to maximalism which is sort of like nativism or nationalism within the peer-to-peer sort of technology and groups rights I mean who's read rbtc or the yeas are Bitcoin obviously and are are a theory and even even to some degree is there anyone here would self describe themselves as a maximalist okay one what one yeah one guy very even have two plays on this yeah I mean there is there is a you know there is a desire to pick a team right and run with the team and that's that's kind of an innate sort of human part of human nature but it's problematic when it happens on to larger scale and the reason it's problematic is because innovation works best when there are low barriers to entry is basically why markets work if the barriers to entry are large the market becomes very it is dysfunctional yeah when you have these kinds of sort of barriers to entry you end up with stagnation within the the group that would sort of lead you end up with a bit of an echo chamber going up and because you don't have people sort of moving to and fro in ideas moving to and fro you end up with a group that tends to tends to consider itself and everyone else considers to be their sort of leaders and they don't really they move away from a meritocracy and much more into a sort of hereditary sort of a notion of how they should be changed and you end up with decisions being made largely down to how to keep this particular group in its current position in its current role yeah so there's no notion that we should do think we should make a decision because of it it's technically the better thing sometimes that's the case but other times we make a decision because this way we will retain whatever ground that is that we have right now and that's unfortunate it slows down innovation in the field as a whole and it can eventually over time cause large amounts of disruption when the leader eventually breaks and a second runner-up must must take the wheel and I mean I think any Bitcoin max must I think we kind of you know maybe maybe not who knows I think we're kind of being a little bit maybe to that point with you know a Bitcoin and maybe there's some of the other block chains just beneath where it's really nice to stick with the leader it's like oh well let you know they've got the inertia they were first move all over but actually hmm maybe there's something wrong in inside the the governance inside the community and it's yeah maybe it's time to sort of rethink how decisions should be made maybe it's time to sort of involve maybe time to push towards another another group that seemed to be able to move be making decisions better and that causes disruption now if you don't have these barriers is less disruption you can move back and forth much more easily and that ultimately allows the field to push forward faster so the one chain model I would argue is really not inclusive in this sense I don't think it's that sensible to have a one chain model and the 1k model is simply the model of the world that says eventually there will be launching just one chain to process all the world's transactions so we've become under polka dot so the point of polka dot from a sort of semi optical point of view is to try to break down those barriers to try to enable multiple chains to co-exist and to move at their own rate without the need for without having the network effect and the maximalism or at least mitigating the network effects and maximalism that we would otherwise see in this multi horse race to do this we allow them to talk to each other so the goals are polka dot are fairly simple on the one hand we we make invention that much easier on the other hand we make connection possible and thirdly we make security easy now coming up with a new blockchain idea I mean maybe it's not easy but if you if you think about lattes if you're in the space it's not that hard you know it's like it's maybe harder than coming up with a new DAC idea but it's still not that hard I da ting isn't too difficult we see quite a few ideas sort of floating around in the ecosystem one or two white papers and and that's cool you know it's great people are people over sort of experimenting at least in that now developing one from scratch all the way to the point that you can actually deploy it for a production system and securing it which is to say finding enough buy-in from people that want to reload they're either trustworthy or enough of the public is really difficult it's really not to be underestimated polka dot is there to make it easy but just to say it provides the infrastructure and hopefully ultimately the tools as well to be able to create new blockchains new ideas behind them new parameterizations new state transition functions new ways of processing transactions that we don't really know about now because you know it's it's early date and the sister there's only there's only sort of been so many minds thinking about this stuff so far and it's really expanding I mean I presume that how many answered how many people actually considered you know writing a blockchain or or thought about you know how clock change might evolve the underlying technology okay so like yeah probably about a tenth maybe 15% of people here I mean ups and we got like what couple hundred people okay and that's just like one meetup in one city so you know there's there is there is a potential yeah and we're going to see an increasing amount of of this kind of innovation over time I'm sure so that actually is to try and connect trains so we've made them easy to to create a build to deploy and actually is to try and try and tailor now with with polka dot really the goal is really to be able to do this for any kind of chain so the model what the change in the fit umber is accordingly general now I'm going to that model in this talk but I mean you can you can read about it online if you want the point is that chains we don't want to restrict this to things like tokens because at least in my opinion tokens are just the first use case of block chains they're a very handy use case they could they they have a lot of potential uses but they're still just the first use case and there's an awful lot more out there for block chains to be doing than just recording balances of tokens so what polka dot does is as fine then we won't bother with tokens or there's other things other projects can sort of do that very particular use case polka dot can do the use case but we will abstract it further we will generalize further and rather than just talking about tokens will talk about messages messages are really general bits of data going to and from an end point within a block chain very very general indeed we can we can imagine middleware that would then take this messaging infrastructure and be able to provide token transfers because check across chains for sure relatively easy but we can also imagine an awful lot of other stuff as well and obviously there's plenty of stuff that we can't imagine furthermore want to limit ourselves to change that already exists or change that only would exist within polka dot or change that only are coming into being now we don't want to connect the chains that are being developed now only we want to change connect change in the future and change from the past and similarly we don't know limit ourselves to just public chains or just private chains or just change that in perform part of a consortium or sovereignty change that can only exist on their own we want to take all kinds of chains connect them together so polka dots designed not to care about what the chain is or how it works it's just designed to be able to move messages between chains in a tryst free fashion so finally security aspects and this is really where oh do I think the batteries run out should I stop talking securing chains yes it's this is an interesting problem it's probably one the most difficult bits of polka dots to provide so it boils down to scalability now by and large chains compete with each other over their security resources so how does Bitcoin secure itself how does the theorem secure itself it secures itself by paying miners from inflating its token mix so it incentivizes miners who whose job it is basically to check transactions not to let any through that are they're invalid in any way and then also to finalize those transactions under a proof-of-work which is very difficult to sort of undo by creating moments that this supposedly harder to degrade so these resources are in competition yeah if you have a certain amount of money you can choose to use that money to either mine Bitcoin and secure the Bitcoin network or to mine either and secure the ether Network you can't use the money for doing both doesn't work there is a notion called merge mining but it mostly economic economically speaking doesn't work that well and we see that because no one really knows right now wouldn't it be good if you could take many different block chains and have them all contribute to each other's security so rather than adding a block chain and effectively taking the finite amount of resources and splitting off a bit from the other block chains and therefore all what cane security goes lower when you add a block cane you actually just effectively share in the same overall security apparatus well that's what polkadot provides and in this sense polkadot scales now so what degree it scales is a is a harder question to answer and that's something that we'll be looking into over the next few month khoka dot in this sense scales when you add a new block chain you don't reduce the security of the other block chains so we come to the diagram we're going to redo this diagram but it's not redone yet so you stick with this one but this is is the slide where I'm going to sort of try and explain roughly Finn in broad strokes and how polkadot works so there are in principle there are two key classes of participant the who helped maintain polka dot one is the validators and you can see the little triangles in the middle right there validators and they kind of they kind of swarm around in the middle and they check they take turns checking the transactions that are going on in the chains towards the edge so the little random polygons towards the edge that power chain so there's this sort of chains that are managed by polka dot part of the polka dot community and they get their state transitions checked so when a new block comes along it gets validated and it's the validators in the middle rabbit Valerie now you don't want to have a sort of carrying on validating the same chain why not well because validators tend to be able to they're generally good guys but if they spend a long time pollinating the same chain then you have a potential structural weakness because if they end up becoming bad guys or paid off they could eventually corrupt the chain they could introduce invalid state transitions and it's not clear or at least it's it's more likely that those invalid state transitions would go by and reported or unnoticed so what we do is we say well let's keep mixing them up yeah so one block we pick you know a few validators one and five and nine and another block we take validators two three eight okay the second kind of role are their Co laters now the Co laters don't exist in the middle the collectives exist on one of the peripheral chains on one of the Power chains right so they manage very particular chain and what they do is they both do more or less what miners do at the moment they fish around for transactions collect them all together bundle up into a block and then provide a sort of proof that normally - provide proof of work if you're on a proof of stake system and you provide a prove mistake in this case the Co laters and provide a proof of validity now proof of validity doesn't sort of kind of change it mean that one block will be accepted over another block rather it simply means that a validator who doesn't maintain that network like validators don't maintain any particular pirate chain they just sit in the middle and floods around it's so that a validator regardless of that and what state they're in can can check that this block is valid now we can do this by using a light client so we have a light client proof and this is indeed one of the one of the prerequisites to a pair of chain joining the polka dot community other than that the validator is basically swarm around they stay they provide proofs that a particular power chain had a particular output Q so para trains can send messages between each other that was the connection bit communication and they also provide the proof that that given it was in particular para chains output Q it should be in this other pair of chains input Q and X blocks so they route messages they did a routing thing now we see at the bottom there there's power chain bridge and this is basically the idea that we're going to bridge it to a sovereign chain so chain is already going it's got its own consensus mechanism if they're in would be the example they're in principle Bitcoin could be could be done now the tricky thing with Bitcoin is that it doesn't recognize n of n signatures so basically Bitcoin has no way of recognizing the authority of polka dots and the polka dot validators who run the polka dot networks yeah so what we need to do is introduce a means a mechanism by which Bitcoin would be able to recognize their sovereignty over funds held on a Bitcoin chain now this is non-trivial it requires an alteration in the Bitcoin protocol or more than 50% of the miners basically you do a soft pork if you have more than 50% of the miners if you can also a Bitcoin protocol then you just introduce something like no signatures which would allow what we call em event signatures or threshold signatures basically allows you to do quite efficient very large t6o multi SIG's away you have need to have some particular number maybe two signatures of a set of three right so as your standard sort of multisig in the Bitcoin world I want to allow these funds to be spent if two of these identities or accounts agree to the funds actually being spent it doesn't matter which two of the three but has to be two or three now in in polka dots case there will be very many validators perhaps several hundred maybe even a thousand so we need to be able to have very large signature very nice pools of potential signatures that could allow the spending of the polka dots funds that are held on a Bitcoin chain and that's why we need a protocol alteration but in principle it's possible now I want to give you a bit of a feeling for where we are when to polkadot so so far it's being sort of prototyped and developed within parity but there is a an ongoing effort to move polka dot and the governance of polka dot over into an entirely independent nonprofit organization so where are we well with busy fruits by cooking here are lots of parts now I'm going to read through these parts to let you know where we are with them but basically this is all of the stuff that needs to be coded up to make polka dot a real real thing so the first is the real a chain core now the real a chain was that bit of a bit of a diagram in the middle right with the validators sort of connected to the into the peripheral chain so the proof of authority consensus is is largely done or at least this prototype now these are earlier algorithms that were working with in reality the polka dot will probably have more efficient algorithms but we have nonetheless implemented a pbft derivative which will be sufficient for doing most of the early prototyping work and Peter can can can stand up take a bow bow for that he's been working on ago is very good he's been working on that next thing is the parallelized multi chain proposal aggregator now this is basically the part of the system that says right we have all of these different para chains all these different chains working parallel and we need to all of the co laterz that are in these particular chains are providing proposals for blocks right so there's many different blocks coming from each different chains we need to choose one block from each chain we can't have many different blocks or all of the chains getting validated I only want one now the question then becomes well which one do we choose from a set of potential candidates and we don't really want to start like naming any validators that sort of get to pick because that a introduces weaknesses B introduces the the possibility of like centralization so these guys that they went down for an offline we won't be able to process the blocks so we have a parallelized in decentralized aggregator so all of the different validators when they're looking out for which blocks they should actually be validating they can agree they can make a reasonably and succinct decision that they can agree on which block from each of the Power chains they will actually end up putting into the final block of the relay chain yeah that's been prophytes in JavaScript I mean and it's available online to look at next is the proof of stake interface contract now this takes the proof of authority and consensus mechanism and turned it into a proof of State mechanism so approval Authority basically works by having a number of named authorities that sit in a contract and a contract states which which authorities these authorities are and the Mehcad the actual algorithm looks after taking those authorities and actually turning them into decision making elements of a blockchain so this block comes up and then this in this block now the proof of stake interface contract is the logic that decides who those authorities are based upon who was staked what tokens and and when now with polka dot it becomes a little more interesting because want to attach parameters to staking offering so it's not simply choose the tops takers but actually we want to talk about things like I which takers will take the least rewards for their work and in that sense it's it's a little more complex but nonetheless that's ongoing and we have we're sort of working on that at the moment that's maybe 30 or 40 percent done the next is the weather meta protocol so why some if you're not familiar with the term is short for webassembly and this is a a new language server for the web basically is supported by the big web firms Google and Mozilla and it's basically I mean you could call it a compound it's an alternative virtual machine that can that you can basically compile software into that can then be recompiled to run on the target machine within your web browser basically means that you can have very high performance programs that distributed over the web and will actually work in any browser on a platform it's very handy now if you take out a couple of the instructions it also becomes consensus sensitive which is to say we can it's absolutely deterministic and we can use it within these kinds of consensus operators so we can use it as a means of stating what the polka-dot protocol is and what the pair attains what we call validation functions are basically how the power chains are meant to work we can also state that in web assembly and we can to a very high degree of certainty that they will work the same way on everyone's computer regardless of platform regardless of hardl now the point of the weather meta protocol is to facilitate upgrades within polka-dot itself that don't require hard Forks so at the moment if you want to upgrade grader protocol you have to do a hard fork right there's no other way of doing it in hard Forks are quite honestly having been working on a hard fought for theorem for quite a while now time consuming and very very very sensitive to two errors particularly errors in communication so if the if the specific protocol alterations are not specified in an extremely rigorous fashion it's and you often don't know whether it's actually rigorous enough until eventually at fork so it's like oh yeah it turns out that wasn't rigorous enough then you have a you have a problem this having everything on a meta protocol basically takes away that issue when you do a hard fork you state the new logic in terms are already a language that is fully deterministic and it avoids having to re-implement it and therefore suffer the problem if the implementations are different now it does also have an issue with low morale as in the centralization out we're supposed to be isn't reimplementation in many different languages good those are there's definitely a discussion to be had there but now is at the time so the final thing for the real h8 core is the peer steering multi-format kind of funky name whistle this is basically the ability to have changed your peer set depending on which role you currently have and what roles you will have in the future so validators for example will know which power chain they need to be validating two or three or four blocks down the line which allows them to pre connect to likely Colitis who will give them candidate blocks two or three or four blocks down the line this is super important because the latencies are likely to be quite quite small in terms of in terms of little times so if you're coming out if you're getting a block every two three or four seconds and it takes you half a second to negotiate a connection then you really need to be pre connecting to the people to the notes that you're actually going to need information from before the point that you need information so this basically is a peer network manager that will intelligently pre connect to change the peer network connectivity topology in response to the the overall characteristics of the network as is required on a higher level so the next thing is bridging now bridging is is this idea of connecting polka-dot to proprietary or legacy block chains ones that already have their consensus mechanisms already going all via bridging is something that was like pretty excited about I'm pretty excited about it's a few people in the team are very excited about it this is something that we're actively working on for all of this stuff so break in contracts break out contracts are already been completed these are basically the bits that allow aetherium or whatever else to the stand that a today well not to understand but to contain the logic to allow aetherium to forward messages from the from the outside world into contracts within etherium so they exist as the gateway contracts yeah similarly they allow transactions coming or messages coming into these contracts allowed forwarded out into the into the external world hence break in and break our contract next thing light client block proofs this is effective the light client for aetherium so how do we given that we don't know anything about the blockchain state at the moment can we be be guaranteed that a transaction executes in a particular way or that a particular part of the state has a particular a piece of data in it this is largely complete and will be part of the next parity release finally the decentralized secret store and site analysis is basically to to allow the validators of the network to each have a portion of a key that will allow them or allow a set of the validators to coordinate a signing of a transaction on behalf of the full set of validators so the idea is as a sort of protected before and the validators or polka-dot validators sort of operate as trustees of would be able to own funds in another network BSA in aetherium or or bitcoin BTC in bitcoin now for them to be able to own funds they need to actually be able to operate as a whole and thus you need this idea of a multi signature system that would allow them to cooperate and coordinate in order to create a transaction that's spent the funds that they supposedly owned now you don't want to allow one of them to spend all of the funds you could just give them all the same private key but that's really bad because of course one of them only to be bad before they run off with the fund so you want to make it much harder so you want to require like two thirds or four four fifths or whatever serve them to actually coordinate before they can spend the funds and then the assumption is that no one 20% will be found well actually no more than 20% will be bad if they are then you could block spending over them and no more than 80% will ever be bad if they were they can actually spend the funds so the decentralized secret store and signer is effectively a way of allowing authority or validators nodes to be able to own an address without any one of them owning that address yeah so the only address as a whole and that's complete so moving on to the economic side of things we have a few so most of this stuff is still in the ideation phase kind of moving through design we have the incentivization model which is basically how do we ensure that validators actually do their job it basically means where we don't pay them pay them in dot and that's sort of that will be clarified probably over the next month next few months next is the stakeholder modeling which is basically to say why would people what how will people act when they have dots and we give them the offer of rewards how much will they actually want to receive back you know will they want 10% a year 50% a year five percent in year two percent a year one doesn't mean nothing I don't know it's very difficult to model this behavior until you actually have a network that requires that has real value on it and that's product something that will be looking into quite soon just to say launching a value bearing test network so we can get empirical results on how people actually behave in these systems next thing is the DAO and so polka-dot includes the notion of a dowel which is there's basically there for long-term management of the polka-dot ecosystem basic idea is it's funded from the proceeds of its funded from inflation of the dot base so require any additional sort of funding it's there it's built into the protocol and there would be in much the same way as there are another and so I think - is quite a good exam or but basically there will be proposals of how the protocol could be improved or maintained and these would be voted on by holders of the by stakeholders within the network next thing is the super eel a guarantor now this is quite of an interesting thing this this is basically the road to infinite scalability within polka-dot so polka-dot at the moment i should go back to the previous thing polka dot at the moment kind of goes one level beat now i did i did suggest that there could be a second order relay chain that's the purple thing on the right there which would be a sort of copy of the main relay chain but it would exist at a lower level right it would exist as a pair of chain at the top with a chain now this is how we would achieve infinite scalability right we would basically build a tree and polka dots yeah and then if you needed to send a message to the other one it was sort of move and then sort of go up and then move around again and sort of go up and move around again and go up and then come down again on the other side yeah and the more you have to go up and down probably would be the modern sort of you had to pay the validators and of each level to do that now the problem is validators if they operate if you have funds that you want to lock up as a validator so you want to basically say I am going to give my funds to be bonded which is to say again I will lose these funds if I misbehave and that's something that you need to be able to do to ensure that people actually operate as you expect then we need to be able to provide that mechanism so that funds that are locked up at the top level can also effectively be used as a promissory bond for lower levels activities lower level activities and that's something that we need investigate as part of the sub relay guarantor system but if we crack that then we've probably got something that we can scale out pretty much indefinitely final part power chain stuff so powertrain stuff if you're not the terminus is all the peripheral stuff so not the relay chain in the middle but the things on the outside first thing is parroting the quiddity provision this is basically like well I have dot tokens which exists maybe on the in the middle bit but I want them to exist within the pair of chains yeah how do i how do they get from the middle bits of the tire chains how do they make their way to the outside and that's that's a portion of the protocol it's a little bit sort of it's mostly fleshed out but it needs to be properly formalized next thing the primary pair of chain protocol so this is we want to release a power Akane that is sort of exactly right for the system it uses all of the best bits of poccadot actually exists within the native polka dot infrastructure rather than etherium which would of course have its own consensus mechanism and therefore be bridged so the primary power chain protocol would would have a bunch of additional sort of bits to it one of those bits would be a web assembly execution environment so the actual language used as part of the smart contracts would be represented that's both of these things are things that we are currently working on can actually see if you look in the parity call requests you'll see and again pull requests coming that that that's from github part of thing is coming that they're actually going to be pertain to this your stuff next thing this protection in chromatic gasps these are two protocol alterations chromatic gasps basically says at the moment in aetherium gasps is of a single type so if you use it for doing chain storage use it for doing memory enlargement or you use it for doing processing like just computation it's the same thing yeah so you do loads and loads of computation or alternate if you could do loads of loads of stuff on a chain like push push loads and loads of data onto the chain in either fire transactions or in the contract storage or you could just add large memory to something massive like you know 40k or 16 Meg's or whatever it is the maximum that's problematic because it means that if we want to do lots and support the potential for lots and lots of computation then similarly we also have to support the potential for a contract that just when it executes dumps loads of spam onto the chain and we saw this being exploited over towards the end of last year in the attacks the China attacks yeah the with the issues of variables try to actually sync chain at the moment it takes ages to do a full sync and it's largely around the blocks that these these kinds of protocol mistakes happened in next thing is this protection which is basically saying that accounts that more or less have no real value in them should be automatically removed and we call accounts that have no real value in this accounts final thing sudbury lays which is basically moving on to the seborrhea later on tours but how do we actually do the the logic to swap these things into place now this this the subarray point hasn't actually been started yet that's likely to be a much later part of development okay that's where we are so quite a bit of stuff sort of in progress and we really hope to have bridges being demoed in the not-too-distant future [Applause] you you

Channel: Stephan Tual

Views: 29,785

Rating: 4.9174314 out of 5

Keywords: ethereum, blockchain, bitcoin, cryptocurrencies, gavin wood, ETHLDN, london, meetup, Polkadot

Id: lIghiCmHz0U

Length: 51min 54sec (3114 seconds)

Published: Mon Jul 10 2017