4 ms·

Yep. Its incredible that for decades we havent developed a decentralized internet of "apps" that is more aligned with how the economy works. In some cases the

Yep. Its incredible that for decades we havent developed a decentralized internet of "apps" that is more aligned with how the economy works. In some cases the current platform design is totally absurd: like linkedin (or github or azure) thus microsoft having insider knowledge about anything that happens in the digital space. This is basically down to society being largely tech illiterate and various people taking enormous advantage of that.

imiric 3y ago
I think this was a failure of the early web. You can see in the original proposal[1] that the building blocks were a "browser" and a "server". So the web client gets to have a user-friendly application for consuming content, but the web server is just... a server. There's no user-friendly counterpart for publishing content. This would be equivalent to the modern web browsers not existing, and users having to "navigate" using cURL. I love cURL, but it's intended for a technical audience, and the web would've never taken off had it been the default client interface. You can also see in the proposal that phase 2 considered a "universal authorship" goal, and has a rough draft of what a user-friendly publishing interface could be, but it was far too vague, and to my knowledge this phase never came to pass. As a result of this, companies like GeoCities and Angelfire popped up to fill the void of what the web lacked, and later on social media platforms served the same purpose. Advertisers noticed early on a very lucrative opportunity, and tech giants were built to support them. ISPs noticed that the demand was only for downloading content, so they never bothered to build the infrastructure needed to also serve content, and most internet connections were asymmetrical. And, so, here we are. To be fair, there have been some attempts to make publishing easier. I think the best attempt was Opera's Unite in 2009. But by then, it was too little, too late, and Opera being Opera, it never gained traction. BitTorrent is a big one, but it's only a protocol, so it should be a building block of whatever the solution is. Now we have all these distributed projects, web3, etc., but they're still unusable by the average internet user. A part of the problem is education, sure, but I reckon that the average internet user has been conditioned to believe that all these mainstream platforms _are_ the internet. So even if friendly publishing tools existed, and users had the knowledge to use them, they would still choose to use the centralized platforms. Partly because, unfortunately, most people don't care about their data, privacy, etc., and just want to consume content. So I don't think any of this will change. Technical users will keep using niche tools, and the mainstream web will keep being centralized and run by advertisers. [1]: https://www.w3.org/Proposal.html https://www.w3.org/Proposal.html
- nologic01 3y ago
  > but they're still unusable by the average internet user yep that asymmetry is deep-seated. but decentralized serving does not need to be extreme. I can imagine e.g. geographically local (e.g. city-level) infrastructure hosting - supported maybe by local taxes like a utility - and people using wordpress enhanced with activitypub. Between each user doing self-publishing and all users being captive in one gigantic data-mining and humiliatingly abusive platform there are literally billions of options. > most people don't care about their data, privacy, etc., and just want to consume content This is something that is repeated a lot (also here on HN) and of course at some very basic level it is true. If the negative impact of current designs was obvious we would not have the nightmare we drifted into. But it is a rather simplistic take on how things work with technology adoption, regulation of business behavior, markets etc. E.g. people didn't not care about ozone layer depletion, they were probably completely unaware of it. That's why we have experts and regulators and institutions to advise the general population about what is safe and in their interest. In the case of the ozone layer, laws were passed. Substances were banned. Substitutes invented. The planet and our skin was saved. We moved on to the next existential risk. The same playbook has repeated countless times. People generally assume that the system works to protect them and that is not a totally wrong assumption. If everybody moved fast and broke things modern society would quickly collapse. Yet notice that nobody of authority ever warned users not to use these platforms. In fact the opposite. In what is unthinkable and unprecedented collusion, most public sector entities actively advertise them, by being in these platforms, providing links to them etc. What has happened in this space is that a major social contract has been violated. Politicians, in a combination of ineptitude ("We run ads, senator") and capture have acquiesced to this drift. This has played out for so long that people who are aware of what is happening are exhausted and defeatist. We should not give up though. This is important stuff.
- mike_hearn 3y ago
  That's not quite how I remember it. The early web was dominated by user authoring tools like FrontPage, Netscape Composer and Dreamweaver. Then you could publish to sites using FTP or WebDAV. Many ISPs provided a bit of web space for people as part of the default offer. And it worked - the early web was full of self hosted websites about all kinds of niche topics with relatively unique designs and which were thematically organized. A few things broke this model. One is that the web page design tools struggled to incorporate any kind of dynamism or collaboration. They were fine for as long as you had a website with a single author. The moment you wanted to have multiple people working on it simultaneously, or you wanted to add comments / ads / hit counters / guest books, the best you could do was stuff for programmers like PHP and version control systems. Another is that the visual design tools weren't very good. Dreamweaver was by far the best, but expensive, and it was still amazingly annoying and hard to get a visual design that didn't look like amateur trash. Yet another is that the web didn't have good tools for thematic design, and figuring out a good site layout was hard, refactoring even harder. Even things as obvious as a consistent navbar were hard because there wasn't good support for static site rendering and HTML doesn't have an include tag. People hacked it up with framesets for a while but that didn't work well either. Finally, connecting your website address to your ISP meant if you wanted to switch to a better internet offer your website would disappear which was unacceptable, so in practice people wanted independent providers, and they needed to use ads to make money, which in turn meant that you hit the limits of the WYSIWYG editors that didn't really understand ads and which iterated, maybe if you were lucky, once every few years. Into this void stepped yes, Geocities, but primarily WordPress. It supported dynamic features, it understood ads, it iterated fast, it had decent looking themes, it said "don't bother with organizing stuff, just keep a diary" and all that massively lowered the barrier to entry for web publishing. Later sites like MySpace and Facebook provided a kind of WordPress-lite and gained success from that, Twitter made it even more minimal and gained even more success, and then finally Instagram took away the obligation to even have anything to say at all and had even more success than that. So the trend here has only partly been about tools. It's clearly not true that people only want to consume content, the entire success of social media and YouTube revolves around the massively huge desire to create and publish. It's done via centralized platforms for the same reason everything else is: they can move fast, they can raise and make money, they can hire the smartest and hardest working people to build them because there's some real chance of reward. Decentralized systems by and large can't do these things.

imiric 3y ago

I think this was a failure of the early web. You can see in the original proposal[1] that the building blocks were a "browser" and a "server". So the web client gets to have a user-friendly application for consuming content, but the web server is just... a server. There's no user-friendly counterpart for publishing content. This would be equivalent to the modern web browsers not existing, and users having to "navigate" using cURL. I love cURL, but it's intended for a technical audience, and the web would've never taken off had it been the default client interface. You can also see in the proposal that phase 2 considered a "universal authorship" goal, and has a rough draft of what a user-friendly publishing interface could be, but it was far too vague, and to my knowledge this phase never came to pass. As a result of this, companies like GeoCities and Angelfire popped up to fill the void of what the web lacked, and later on social media platforms served the same purpose. Advertisers noticed early on a very lucrative opportunity, and tech giants were built to support them. ISPs noticed that the demand was only for downloading content, so they never bothered to build the infrastructure needed to also serve content, and most internet connections were asymmetrical. And, so, here we are. To be fair, there have been some attempts to make publishing easier. I think the best attempt was Opera's Unite in 2009. But by then, it was too little, too late, and Opera being Opera, it never gained traction. BitTorrent is a big one, but it's only a protocol, so it should be a building block of whatever the solution is. Now we have all these distributed projects, web3, etc., but they're still unusable by the average internet user. A part of the problem is education, sure, but I reckon that the average internet user has been conditioned to believe that all these mainstream platforms _are_ the internet. So even if friendly publishing tools existed, and users had the knowledge to use them, they would still choose to use the centralized platforms. Partly because, unfortunately, most people don't care about their data, privacy, etc., and just want to consume content. So I don't think any of this will change. Technical users will keep using niche tools, and the mainstream web will keep being centralized and run by advertisers. [1]: https://www.w3.org/Proposal.html https://www.w3.org/Proposal.html

nologic01 3y ago
> but they're still unusable by the average internet user yep that asymmetry is deep-seated. but decentralized serving does not need to be extreme. I can imagine e.g. geographically local (e.g. city-level) infrastructure hosting - supported maybe by local taxes like a utility - and people using wordpress enhanced with activitypub. Between each user doing self-publishing and all users being captive in one gigantic data-mining and humiliatingly abusive platform there are literally billions of options. > most people don't care about their data, privacy, etc., and just want to consume content This is something that is repeated a lot (also here on HN) and of course at some very basic level it is true. If the negative impact of current designs was obvious we would not have the nightmare we drifted into. But it is a rather simplistic take on how things work with technology adoption, regulation of business behavior, markets etc. E.g. people didn't not care about ozone layer depletion, they were probably completely unaware of it. That's why we have experts and regulators and institutions to advise the general population about what is safe and in their interest. In the case of the ozone layer, laws were passed. Substances were banned. Substitutes invented. The planet and our skin was saved. We moved on to the next existential risk. The same playbook has repeated countless times. People generally assume that the system works to protect them and that is not a totally wrong assumption. If everybody moved fast and broke things modern society would quickly collapse. Yet notice that nobody of authority ever warned users not to use these platforms. In fact the opposite. In what is unthinkable and unprecedented collusion, most public sector entities actively advertise them, by being in these platforms, providing links to them etc. What has happened in this space is that a major social contract has been violated. Politicians, in a combination of ineptitude ("We run ads, senator") and capture have acquiesced to this drift. This has played out for so long that people who are aware of what is happening are exhausted and defeatist. We should not give up though. This is important stuff.
mike_hearn 3y ago
That's not quite how I remember it. The early web was dominated by user authoring tools like FrontPage, Netscape Composer and Dreamweaver. Then you could publish to sites using FTP or WebDAV. Many ISPs provided a bit of web space for people as part of the default offer. And it worked - the early web was full of self hosted websites about all kinds of niche topics with relatively unique designs and which were thematically organized. A few things broke this model. One is that the web page design tools struggled to incorporate any kind of dynamism or collaboration. They were fine for as long as you had a website with a single author. The moment you wanted to have multiple people working on it simultaneously, or you wanted to add comments / ads / hit counters / guest books, the best you could do was stuff for programmers like PHP and version control systems. Another is that the visual design tools weren't very good. Dreamweaver was by far the best, but expensive, and it was still amazingly annoying and hard to get a visual design that didn't look like amateur trash. Yet another is that the web didn't have good tools for thematic design, and figuring out a good site layout was hard, refactoring even harder. Even things as obvious as a consistent navbar were hard because there wasn't good support for static site rendering and HTML doesn't have an include tag. People hacked it up with framesets for a while but that didn't work well either. Finally, connecting your website address to your ISP meant if you wanted to switch to a better internet offer your website would disappear which was unacceptable, so in practice people wanted independent providers, and they needed to use ads to make money, which in turn meant that you hit the limits of the WYSIWYG editors that didn't really understand ads and which iterated, maybe if you were lucky, once every few years. Into this void stepped yes, Geocities, but primarily WordPress. It supported dynamic features, it understood ads, it iterated fast, it had decent looking themes, it said "don't bother with organizing stuff, just keep a diary" and all that massively lowered the barrier to entry for web publishing. Later sites like MySpace and Facebook provided a kind of WordPress-lite and gained success from that, Twitter made it even more minimal and gained even more success, and then finally Instagram took away the obligation to even have anything to say at all and had even more success than that. So the trend here has only partly been about tools. It's clearly not true that people only want to consume content, the entire success of social media and YouTube revolves around the massively huge desire to create and publish. It's done via centralized platforms for the same reason everything else is: they can move fast, they can raise and make money, they can hire the smartest and hardest working people to build them because there's some real chance of reward. Decentralized systems by and large can't do these things.