Macrobius
Megaphoron
thoughts on the future architecture of the Crust
These are just some preliminary rambling thoughts on where Web 3.0 is going. I think 'web 3.0' won't exist -- it won't, really, be 'the web' at all. At least not like web 1.0 or 2.0, though it may use some descendant of the HTTP protocol and a variant of web app tech.
Let's start with that thorn in the side of Tim Berners-Lee's original vision -- DNS. In the early 90s didn't sufficiently realise that Zooko's Triangle[1] would operate at global scale -- that if you attempted to mix 'decentralised' 'human readable' and 'secure' you would lose decentralisation. What was supposed to be a Liberation ended up being a Con Job.
[1]: https://en.wikipedia.org/wiki/Zooko's_triangle
Economics tells us that 'network effects' are important -- and indeed, we ended up with one or two basic 'maps of the web' rather than letting 1000 flowers bloom (as Joseph Wang, author of tkWWW, the first 'graphical browser' put it). So sad.
We ended up in a world where the humourist Andrew Anglin could be run off 'the web' by denying him access to the single way of finding his site (for 99% of the population -- the 'human readable' part). Where Napster could be *crushed* -- and this on a network originally designed to withstand a Soviet-American Nuclear Holocaust.
tfw, we've lost POE (Purity of Essence). Repeat after me: NO PREVERSIONS!
About the Preversions: the original idea behind DNS (and in general TBL's 'Web' was to have semantic tagging of texts (using a sort of bastardised SGML [an IBM monstronsity] called 'HTML') that respected Knoweledge Engineering and organised what was then called 'Gopherspace' into a hierachical, conceptul, semantic sort of space.
Remember the six original protocols please: WAIS (a search thingy -- in practice, Archie/Veronica/FTP worked better for that), GOPHER, HTTP, FTP, TELNET (now by necessity SSH) and NEWS (Usenet News, like BBS an' shit). Later FILE.
The whole shebang bootstrapped off of NEWS and GOPHER, mostly, primarily because of 1/ a provision for graphics ('images') and 2/ incorporation of the Vannevar Bush / Alan Key concept (hyperlinkage with an online database) - later reinvented as the 'wiki-wiki' in terms of the 'the simplest online database that might work'
So, the first thing we know about Web 3.0 is that DNS won't be a very important part of it. Pretty much *any alternative* has to solve this problem.
Single Page Applications (SPA) and Web 2.0 leads the way -- web 2.0 'takes the web back' partly by taking over the 'history' mechanism -- when you are browsing, where do you go and what does 'going back' mean? We fix that in JavaScript and XML (originally). It's a BLOODY HACK, but in the end, it replaces the notion of 'a website' (provided by DNS) as 'an app'... which can be provided by anything. It can even be a thing on your device.
So... if an app is a site, and we can have a single page application, that implies we can have a single page site. And then again, we can have a *single page web* (SPW concept I will call it).
DNS can give you one page, and your can take it from there -- how. Imagine, instead of a browser, you have an application that brings up a single 'webview' thingy in say a mobile app. What do you put on that page? Well, you put a SEARCH ENGINE... a textbox or some kind of input where the user selects what they want. Some normies use Google (or Bing if they are 'captive') this way. You do something and get that text box, and then say 'fb' or 'ebay' or 'google' or whatever -- the most common queries on the web in fact.
So, DNS is not *needed* (screw that Jew owned scam now). Nor, really, are 'bluehost' and 'hostgator'. Or rather, rather than having *one* DNS pointer you could have 1000 or a million. Notice what you stopped using -- the browser.
Now that's a bit harsh. That means you forced someone to put a antithesis browser on your machine. Ah, but you can make the 'antithesis browser app' you want everyone to install be PART OF YOUR BROWSER. Way out.
This is where WebAssembly (WASM) fits in. What if you make a web app in the cloud, and it delivers WASM to the browser eh? (Microsoft's Blazor is one way to do this -- so many others).
Now you have a browser that ones ONE CLIENT, ONE CLOUD, ONE VOLK. All the web it reduced to a SINGLE PAGE.
What do you put on that one page? Well, since we have the JAMStack, you put an AI editor -- what does that look like? A chatbox with an AI talking to you. No shit.
So you go to the one website that is all the web (SPW)... you see a text box that talks to an API... you get a static website (JAMStack tech)... and here we are in GeoCities circa 1997, and everything is White again.
See how that works?
These are just some preliminary rambling thoughts on where Web 3.0 is going. I think 'web 3.0' won't exist -- it won't, really, be 'the web' at all. At least not like web 1.0 or 2.0, though it may use some descendant of the HTTP protocol and a variant of web app tech.
Let's start with that thorn in the side of Tim Berners-Lee's original vision -- DNS. In the early 90s didn't sufficiently realise that Zooko's Triangle[1] would operate at global scale -- that if you attempted to mix 'decentralised' 'human readable' and 'secure' you would lose decentralisation. What was supposed to be a Liberation ended up being a Con Job.
[1]: https://en.wikipedia.org/wiki/Zooko's_triangle
Economics tells us that 'network effects' are important -- and indeed, we ended up with one or two basic 'maps of the web' rather than letting 1000 flowers bloom (as Joseph Wang, author of tkWWW, the first 'graphical browser' put it). So sad.
We ended up in a world where the humourist Andrew Anglin could be run off 'the web' by denying him access to the single way of finding his site (for 99% of the population -- the 'human readable' part). Where Napster could be *crushed* -- and this on a network originally designed to withstand a Soviet-American Nuclear Holocaust.
tfw, we've lost POE (Purity of Essence). Repeat after me: NO PREVERSIONS!
About the Preversions: the original idea behind DNS (and in general TBL's 'Web' was to have semantic tagging of texts (using a sort of bastardised SGML [an IBM monstronsity] called 'HTML') that respected Knoweledge Engineering and organised what was then called 'Gopherspace' into a hierachical, conceptul, semantic sort of space.
Remember the six original protocols please: WAIS (a search thingy -- in practice, Archie/Veronica/FTP worked better for that), GOPHER, HTTP, FTP, TELNET (now by necessity SSH) and NEWS (Usenet News, like BBS an' shit). Later FILE.
The whole shebang bootstrapped off of NEWS and GOPHER, mostly, primarily because of 1/ a provision for graphics ('images') and 2/ incorporation of the Vannevar Bush / Alan Key concept (hyperlinkage with an online database) - later reinvented as the 'wiki-wiki' in terms of the 'the simplest online database that might work'
So, the first thing we know about Web 3.0 is that DNS won't be a very important part of it. Pretty much *any alternative* has to solve this problem.
Single Page Applications (SPA) and Web 2.0 leads the way -- web 2.0 'takes the web back' partly by taking over the 'history' mechanism -- when you are browsing, where do you go and what does 'going back' mean? We fix that in JavaScript and XML (originally). It's a BLOODY HACK, but in the end, it replaces the notion of 'a website' (provided by DNS) as 'an app'... which can be provided by anything. It can even be a thing on your device.
So... if an app is a site, and we can have a single page application, that implies we can have a single page site. And then again, we can have a *single page web* (SPW concept I will call it).
DNS can give you one page, and your can take it from there -- how. Imagine, instead of a browser, you have an application that brings up a single 'webview' thingy in say a mobile app. What do you put on that page? Well, you put a SEARCH ENGINE... a textbox or some kind of input where the user selects what they want. Some normies use Google (or Bing if they are 'captive') this way. You do something and get that text box, and then say 'fb' or 'ebay' or 'google' or whatever -- the most common queries on the web in fact.
So, DNS is not *needed* (screw that Jew owned scam now). Nor, really, are 'bluehost' and 'hostgator'. Or rather, rather than having *one* DNS pointer you could have 1000 or a million. Notice what you stopped using -- the browser.
Now that's a bit harsh. That means you forced someone to put a antithesis browser on your machine. Ah, but you can make the 'antithesis browser app' you want everyone to install be PART OF YOUR BROWSER. Way out.
This is where WebAssembly (WASM) fits in. What if you make a web app in the cloud, and it delivers WASM to the browser eh? (Microsoft's Blazor is one way to do this -- so many others).
Now you have a browser that ones ONE CLIENT, ONE CLOUD, ONE VOLK. All the web it reduced to a SINGLE PAGE.
What do you put on that one page? Well, since we have the JAMStack, you put an AI editor -- what does that look like? A chatbox with an AI talking to you. No shit.
So you go to the one website that is all the web (SPW)... you see a text box that talks to an API... you get a static website (JAMStack tech)... and here we are in GeoCities circa 1997, and everything is White again.
See how that works?
