DAPPCON 2018: StatBeyond Hypertext - Reflections on the Post-Web System - Jarrad Hope (Status)

DappCon 2018 · July 2018 · 20:44

Jarrad Hope of Status offers reflections on the post-web system at DappCon 2018, walking through the messaging and dApp browser project's vision for a hypertext layer built on decentralized protocols rather than centralized web servers. The talk covers how Ethereum smart contracts and peer-to-peer messaging can underpin a browsing and communication experience less dependent on corporate platforms.

Speakers

Topics

From DappCon 2018, hosted by Dappcon Berlin.

Transcript

so yeah hi I'm Jared from status we work in a hybrid mobile browser and that browser and instant messenger it's built entirely on top of the ethereum stack essentially so using whisper and doc doc messaging and the idea is that you're running a full note on your on your mobile device and you can access the apps through this but today I'm kinda here to talk a bit about post web systems and basically the web today and then what hypertext is and where we go from from here you know I know it's late and leave in the afternoon so we're all

pretty tired don't worry this is pretty pretty like lightweight just lots of nice colors and images and that sort of thing so it should be all good click next slide please cool so yeah so basically this is mosaic this is one of the sort of precursor to Netscape and it's kind of like a first web browser and it's quite it's quite interesting because pretty much the web

browser back then 30 years ago pretty much looks exactly the same as our browsers today at least on desktop cool the difference being is now we're doing it on a mobile devices on the the far right left sorry image is basically a webview and you you have the back and forward buttons you can get an address bar and that sort of thing so ultimately the the user experience and the UX hasn't particularly changed that much and it's it's kind of interesting to think about because it's basically we're doing all this kind of crazy interesting things on

the blockchain side you know like we're creating things like swarm or ipfs creating a decentralized file storage we're trying to move quote server-side logic away and put it into this weird smart contract thing but then on top of that we're just kind of using WebKit or just the the existing version of the web and it's a little unclear to me is what that looks like you know in the next 30 years if we're moving to personal computing devices which are our smartphones is that where it ends and is that where we're just going to continue how using I'm not sure if that is the

case here's a picture of leader oceans project North Star this is a mixed reality device or augmented reality device and the construction of it is actually open source as well which is kind of cool but this is something that exists like today so if you can imagine in like maybe 10 15 years time with the current trends in terms of like battery advancements the trends in mobile bandwidth and this thing getting a whole lot smaller hopefully then we might actually start wearing our phones on our faces and in that cat that point in time we have to start revisiting how we actually do our user experience and what does the web actually look like I don't know if

anyone's had much chance to work with with virtual reality yet but bits of paper or like these sort of planes don't really work all that well and this is just matter a ours concept image kind of cool when it there's also another interesting divide which is like currently the web is contained purely within our devices we're essentially looking onto the digital world through a window but if we start having these mixed reality devices then the web is and will the line between the digital world and the real world is going to start blurring quite a lot and again like maybe maybe windows and pages are

not necessarily the the right way to go about this perhaps we start seeing a lot more augmented objects or you start seeing other because of completely virtual objects living in this same 3-dimensional space that we're in or both of these pictures are from like this little a crowdfunded clip called hyper reality definitely worth like watching because it's not necessarily dystopian but it's very predictive of how things would currently go giving our current trends in tech so the question here is do pages really make sense and and really like why is it called

hypertext anyway now because most of what we use for the web is essentially an application sandbox and application delivery and that got me thinking about like these pages and wondering well what is the web in any way and how do we actually get here and so I kind of want to quickly go over a quick sort of history of the web and the reason for this is because I won't go into too much detail with of the individual characters but they all essentially some kind of contribution which I'll try and summarize towards the end in terms of an unrealized vision of the that does not exist in in our current web and with most things it being begins with God well his Bible it is basically the first

sort of large text or document that's been severely annotate annotated within itself and self referential so you can start to think of these things as like hyperlinks and then fast-forward to 1737 we had our boy Andrew Ramsey basically it was a freemason and set an objective to create this Universal encyclopedia basically describing the the Internet in some kind of form and that seemed like a really good idea to pull out Lee who ended up actually creating a completely analog version of the Internet he had his UBR which essentially was a series of index cards all filed away

it's pretty crazy it started off I think in 1895 with roughly 400,000 index cards and towards its end at 1934 he had accumulated over 15 million of these cards and had like a whole bunch of workers on there and he also had this other concept of a knowledge city which I think was going to be for the League of Nations I'm not sure but was called the mundaneum and it's basically you could go to these terminals and you could like look up this information and people would find it for you then we come to the ear era or the coming decades of global brains we had peer talking about the news new sphere in which humanity was creating a thinking

membrane that was going to cover the entire planet and that was roughly around 1920 you had HG Wells now starting to talk about the world brain a really interesting read and yes it's same sort of ideas and then you had Charles Carter towards the the end of his little life saying you know man would no longer need documentation if he was assimilated into a being that become omnis isn't in the manner of God himself it's also kind of interesting because back in 1883 he was basically saying yo those future people they're totally going to have desktops with broadband pretty pretty profound thinking for back then fast-forwarding a

little bit more we start to start to get to things that actually look closer to the web we have our mispronouncing this probably but Vannevar Bush had a conception basically got frustrated with the existing technology today so he gave up and trying to build anything instead he just tried to think of what a perfect system was and he created an essay called as we may think absolutely one of the most profound sort of documents that influenced a lot of the web highly recommend reading it and he had this conceptual conceptual machinery called mimics and it has a lot of these ideas of associative linking you know hyper linking all these sort of networked

texts and that sort of thing he also imagined that you know like you could have a camera on there and you'd be documenting what you did and oh well I can't come to crazy stuff you can also see some like initial ideas of like tablets which is kind of cool then we got Douglas Engelbart who was was heavily influenced by bush and he started to think about how we can actually use computers to solve our collective just collective problems and these these more and more complex problems that we wouldn't be able to solve with our current technology and so he created a framework called augmenting

human intellects and then he basically gave birth to like modern computing he created it as and that's in the form of nls being demoed in the mother of all demos and it's pretty much where everything came from he even invented the mouse you know incredibly incredibly intelligent person and really interesting conceptual thinker roughly around the same time you had Ted Nelson on the scene who who actually coined the term hypertext and he has the quote longest-running vaporware which is product Xanadu Xanadu actually has some really really interesting ideas that have never been fully realized Ted

actually worked with with Andy in creating Hess the hypertext editing system and this is basically the first it was actually the first word processor but it also had hyper linking capabilities within it as well and they were completely unaware of NLS at a time then Douglas Engelbart kind of perhaps the torchon sister to Alan Kay in terms of Xerox PARC and he pioneered this whole windowing graphical user interface design okay he came up with object orientated programming Dinah books were you know precursor to laptops and and tablets and small talk which is a really

interesting way of programming where the the code is is and the UI can all be kind of modified on the fly and it's a concept that you don't really really see that much of except for maybe in lists to a smaller extent and then we got our golden boy Tim berners-lee he basically invented the the modern world wide web as we have today you know he credited the URLs hp's and you know HTML but as an HTML derivative and yeah but there are some problems with with the current state of the web and a lot of it has to

do with ephemera morality of data in particular I don't know how many times you've got such the web and you've come across like a dead link for example the same with with data processing sometimes you can find that data but you can no longer view it or access it because the ability to process of the software to process longer in existence which is called bit right in the in the current version the web you also have a pake one way links so links are embedded in the documents and you can't really get access to those or see them except for being on that particular page and I also wanted to talk about this page analogy inside like Windows applications it's also kind of

it's pretty interesting actually how that happens basically in nls or angle BOTS work they had this concept of a page but all of the text was actually networked underneath and then he kind of showed that to Kay and was like yeah this is like Network text and here's all its paper and he's like nah and then Steve Jobs came along and basically was like oh this is great this is like a desktop and paper and Kay is like well no it's not actually you know like all of this is like a completely different UI you can pull it apart it's all networked it's all linked in underneath there bill gates kind of came along and appropriated jobs work and kind of

popularized this is a siloed application and so we never really got to a chance to to spread this particular idea out that we could actually compose software out of different views and out of different modular components its reimbursed in certain certain ways but it's never been like an underlying principle of how we done computing anyway so all of those people have basically contributed to to a large list of wants or desires in what would be the idealized version of the web I mean firstly if we're going to create a new version or new metadata system it needs to subsume the existing web so basically what that means is we all we

really need to do is kind of pull links out of documents as we we browse them and separate them one of the key features is that none of these none of these systems had the links in there except for Tim berners-lee but that was actually one of the features that got Webb adopted in the first place is because it was so easy to author so anything that we have to needs to allow to allow for the creation of content the linking of it and recomposing of content into into new contents as easy as possible data should in the current web that this is the case you know you have to connect to a server

you know things like IP FS and swarm I'm making huge leaps in terms of solving this particular problem again data is is a first-class citizen that means that you're not going to have any special rights as an author once you've deployed that data on to the web and everyone a host has the that's using the web has exactly the same rights to that to that data which means you can then like create new versions of its you can link to you can create new links on any kind other kind of data and so you get this really interesting sort of social space of local knowledge that develops around it yeah mentioned that yet I mean links

themselves should also be immutable and they can have multiple in-and-out references it doesn't have to be just a simple one-way link you can create links that actually have tight so after attributes as well so at the moment we just have this one particular link and you click it and you go to the next place this would allow you to do other interesting ways of linking perhaps like you can imagine a link that requires F or you can stake it so it's a signal that that ain't link is more valuable to somebody for example later on you can actually turn these links into views themselves which I'll talk about on the bit yeah I mean so the link is also it should be attributed to user if you

create a link it should be attached to an author an ID links are transactional so while they're also attributed to user they're also attributable to a point in time their link should be actually directed and bi-directional and visible to all so currently right now it's a fake like you can't actually see a link outside of it and if you can't see what's linking to you okay there's if your web browser gives some kind of refer and the server like says okay record that backlink then you get some visibility on there as the the server but as a user you have no idea what's linking to that content that you're currently viewing yeah all new versions

of any kind of data should have a link to its past so you everything is retraced to pull in some step and there should be a layer two system in terms of subscribing to black link lists voluntarily if you don't want to save certain kinds of link types or links and this idea of views which is I mentioned before but is essentially you can think of it kind of like CSS so kind of like grease monkey in terms of being able to overlay other kinds of data onto other data and that's somehow it gives some kind of transformation but you can actually take that like much much further and you can do some really really interesting things with that if you had a chance to play around with small talk you'll get a much better understanding okay so I mean like what

guess what what this gets us basically I think one of the most powerful things here is this the idea of democratizing of search if you're doing decentralized search then you have some degree of proved ability to show that things have been linked in a certain way and that data is open to anybody like right now you have other other entities out there that are holding this and creating monopolies is around around their data and this would make it completely open you also get this idea of local knowledge mapping in certain areas you'd actually be able to have communities who are very knowledgeable in a certain space actually structure information in very unique and interesting ways there's also this sort of social link space

because not only can other people create links on your own on your content on your website you can also imagine sites that are siloed right now such as reddit they could actually live as a link type and you can actually get comments on web pages if you're if you're so desire without having to go off to a separate web site and then back you get self-describing data types gotcha so basically with this like right now you can you can solve bit rot by just having a viewer or a link in terms of pulling down the codec for a video and the video itself and so it automatically executes in a place without you having

to download any of software or have any extra plugins or an extensions in the actual browser itself and you can do think more complex data structures so you can imagine actually taking get objects and deploying them into this train into this structure but not only that you actually get to take along the issues you get to take along the pull requests you get to see in history how the the project's developed and you can actually go and view that in your browser we could even include like Douw like any decision-making that's include included in that including funds and you can see how all that worked and you can navigate it quite work quite effectively and if we do it as a directed hyper graph then we could even describe concert complex networks and perhaps

this gets us close to maybe a conscious global brain and this other idea of the 3d web all this post symbolic communication will actually allow us to encode abstract concepts into objects themselves and this is absolutely transformative for learning you can imagine actually having an intuitive objects that you can play around with that teach you some kind of formal physics or formal mathematical mathematics yeah we could rebuild it on at a time so directed hypergraph that's what we you need to do and there's a possible implementation in terms of that you can actually deploy in smart contracts that would actually do this really well it's basically get rid of hyper edges

and vertices treat them all like atoms have ins and outs for different address types yeah so I mean if you're interested in doing applied research on this please contact me because I'd love to work with you anyway Thanks [Applause]