7 ms·

I'm not wrong. We have moved towards a world of centralized services because that's what people want. To many people it's the centralized services that "are" t

I'm not wrong. We have moved towards a world of centralized services because that's what people want. To many people it's the centralized services that "are" the Internet: YouTube, Facebook, Twitter, Instagram—that's it. YouTube is the lowest common denominator of video. You aren't going to "fix" its problems with a more complex decentralized network. If you want to solve the problems with YouTube you have to do it with another YouTube. Build another centralized service that fixes the problems that YouTube has, because that's what people want. Not a federated network. Creators want do less work for a larger more consistent paycheck. Have we learned nothing from literally every single project the that has tried to "free" us from the evils of the centralized services? Here ya go: https://en.wikipedia.org/wiki/Comparison_of_software_and_protocols_for_distributed_social_networking https://en.wikipedia.org/wiki/Comparison_of_software_and_pro... Slowly they all go from Active to Dead or Stalled. YouTube's audience doesn't care about anything at decentralized network has to offer. It's just entertainment for them. They just want to watch videos and shitpost in the comments.

smnplk 8y ago
Another point of decentralised services is cutting out the middleman and this is what should be appealing to content creators.
- themagician 8y ago
  Not really. Content creators often look for middlemen to represent them. That's why they join networks that works directly with YouTube. Most of the views to people who do YouTube as a job. When you look on that trending list, 90% of those videos are people waking up and going to a job. The last thing they want is more bosses and more logistics.
  - cyphar 8y ago
    > That's why they join networks that works directly with YouTube. They usually join networks more out of necessity than choice (from memory you get access to more "perks" if you're with a network). Many YouTubers bemoan the existence of multi-channel networks because they take a lot of power away from content creators (even more than YouTube already does). > When you look on that trending list [...] These days it's pretty obvious that the trending list is effectively all paid promotions. It's not uncommon that content that has already gone viral is nowhere near the trending list (and the trending list is instead videos advertising new movies). In the "old days" of YouTube, the front page was hand-selected content that someone working at YouTube thought people would enjoy. These days it's all run by opaque algorithms. > The last thing they want is more bosses and more logistics. Given how many YouTubers are complaining about their current bosses being completely awful, I'd think the last thing they want is to continue with the status quo.
  - altfredd 8y ago
    > Content creators often look for middlemen to represent them. Oh boy. Bees are looking for badgers?! Content creators are looking for support and high reliability. "The middleman" can sometimes deliver those things... Until they suddenly boot you out, because their algorithm determined, that your videos don't deliver enough bucks in advertising revenue to keep them afloat. Of course, "the algorithm" is merely poor excuse for carefully orchestrated scenario, which might have as well resulted from manual intervention (and most famous precedents are indeed backed by staff-approved decisions). Google have put great deal of effort into pretending, that their services have no support teams and no living humans behind them, but in practice places like YouTube are censored to hell and back, — and all of that censorship happens at expense of "content creators".
taurath 8y ago
> We have moved towards a world of centralized services because that's what people want. Almost there. People want INTEROPERABLE services. If that occurs because one place has a monopoly, that’s okay. If everyone can participate across any platform they choose that has the features they want (privacy, a close knit community, etc) that’s better.
- charlchi 8y ago
  Ease of user interoperability trumps nearly everything else. Most users don't care about your conceptual and moral motivations as an engineer. All we need to do is change the tech serving the problem underneath without necessarily changing the user-facing solution too much. It's such a simple idea yet people seem unable to grasp it; and end up whining about issues that have minimal relevance. There's a gap in the thinking somewhere. Plus, please carry on arguing about why I need to view 3 advertisements simultaneously (yes, use the YouTube mobile app). Engineers who have been reading the news and seeing the results of our latest technological models for the past 10 years don't want the same broken manure that is good for the bottom line but bad for people. Literally stop thinking about the bottom line for one second and focus on the people and maybe you'll see sometimes it's not about choking every last bit of stink out of the old model but completely changing the game. Apparently these are broken dreams though.
feral 8y ago
I don't know whether you are right, but I don't find your argument persuasive. You could have made a parallel argument, before Youtube, about how people wanted only professional content, why a service for user created content would never work etc, because that's all that had worked, up to that point. Just because all the decentralized projects have failed so far doesn't mean they are doomed. Its really early days yet. Youtube is just over a decade old. HBO is 5 decades old. Clearly decentralized services are harder to build and get right. But that means there's still plenty to figure out, and plenty of potential. > YouTube's audience doesn't care about anything at decentralized network has to offer. It's just entertainment for them. They just want to watch videos and shitpost in the comments. People change, they grow in sophistication. E.g. look at Facebook - at the start it was full of spammy viral activity notifications for crappy games. People got sick of that and it was forced to change. Now its full of other crappy stuff, but which people are getting sick of now too.
- themagician 8y ago
  We grow down in sophistication in tech. Everything goes to the lowest common denominator, then it falls further. The web used to be sophisticated and it was only for nerds.
  - feral 8y ago
    I think your argument is mostly running into selection issues. If its the middle ages and the only people who have learned to read are the very most educated part of the the population, then the average printed material is going to be intellectual. As the reading population expands over the next 1000 years, and the average book hence becomes less scholarly, you might naively conclude that everything is becoming 'lowest common denominator'. But overall, education standards + sophistication are rising hugely. Sure, when only nerds could access the web, it was super sophisticated. And it had to get easier to go mainstream. But that doesn't mean the mainstream trend is to always dumb down forever, as you are arguing. Apple mouse used to have only one button, but now look at the power of the multitouch UX almost everyone uses on their smartphone daily. Look at Gmail compared to Hotmail v1, Youtube compared to cable TV etc. Its easy to forget how powerful and sophisticated they are. People have learned to do really sophisticated things, create, communicate etc. Sure, things dumbed down since only CS grad students used the Internet. But now that growth is slowing, I bet tastes are going to increase in sophistication again.
thevardanian 8y ago
You listing out a bunch of decentralized, protocol based solutions means that there's competition in the market, which is evidence of a new market that is GROWING. The opposite of a dead market. Youtube is also a the result of such a competitive market. There used to be dozens of video hosting platforms. The same applies to almost every other company. Facebook, Google, Amazon, Ebay, etc. all grew up in very competitive markets. Competitive markets means that there's demand. Decentralization does one very important thing. It makes it more profitable, and cheaper to make content accessible to your audience, while giving you the freedom of making money. That is to say decentralization amplifies the ability of computers to make work what Youtube and Google require billions to do. It DISTRIBUTES that workload among many independent workers, and lets the market decide how valuable their input is. Secondly centralized solutions can exist in a decentralized environment. Content distribution should always have been decentralized, making profit on content is the centralization problem. You're getting content distribution, and profitability mixed up.
- stale2002 8y ago
  Umm, I don't think I agree with you on the idea that decentralization decreases costs for contract distribution. Distributing content to millions of people requires massive investment. The total cost would likely INCREASE massively, as my home computer is much worse at acting as a CDN than Google. Sure, the costs get DISTRIBUTED, to places where maybe you aren't the one paying for it. But SOMEONE is still paying that money. And they are doing much less efficient than a centralized service with thousands of engineers.
  - thevardanian 8y ago
    You don't run your website, necessarily, on a local home computer, why would you assume the same for this?
    - stale2002 8y ago
      So it's not decentralized then? I thought that was the entire point. But even if it is not literal home PCs, and is instead some sort of middle road between massive company and home PC, my point still stands. Your 10 person company hosting content isn't going to be able to stream it live to 20 thousand people. That takes lots and lots of money and engineering talent to do right.
tombh 8y ago
Is this not the archetypal battle of modern software? It makes me think of Eric Raymond's classic book "The Cathedral and the Bazaar". At which point the other side rolls out the Linux precedent - the decentralised success. I believe the world very much does want decentralised services, it's just that such services have quite different measures of success.
progval 8y ago
> Build another centralized service that fixes the problems that YouTube has, Then just create a huge PeerTube instance, and you'll get the benefits of both worlds. Plus, you'll pay for less bandwidth than Youtube.
api 8y ago
You're making an assumption-- namely that federated or decentralized systems must be harder to use than centralized ones. Present decentralized alternatives usually are. There are numerous reasons. Some have to do with unsolved or inadequately solved problems in distributed systems engineering, but most are I think more pragmatic. The two largest reasons for distributed/decentralized/federated services sucking are (IMHO): (1) Most of these things are developed by geeks who lack an appreciation for the critical importance of user experience and lack expertise in areas like GUI and UX design. The result is products by geeks for geeks and that fail to get much traction outside small geek-centric markets. (2) There is no good economic model, which means that there's no incentive for developers to do the parts of software development that are not fun. Solving grand problems is fun, but fixing a trillion trillion edge case UI/UX and platform bugs is not. Supporting users definitely is not. But those are the things that take something from a niche or nerds-only product to something regular people will use. Centralized platforms have two economic models: charging directly for use or monetizing the user through surveillance capitalism. (The latter is more common.)
_jal 8y ago
> YouTube's audience doesn't care about anything at decentralized You're speaking for quite a huge number of people about their preferences vs. something that does not exist. "Content creators" didn't care about the internet, until they did. For that matter, apparently I'm not a member of the "audience", either. Maybe you're right. I'd need to see some argument other than "because I think so", and look at these failures[1], though. As far as bald assertions, here's mine: Video consumers on average don't know what the word "federated" means in the context of networked services. Nor do they care. They don't know that email is federated. They know they hit "send", and their words pop up in grandma's inbox. Users don't care about the tech under the hood. They want the payloff. If the economics of federated content networks work, they will take off once someone figures out the proper feature set, placement and happens into a bit of luck. If not, we're doomed to sharecropping on FaceGoogazon and this whole conversation is pointless. [1] I have here a list of social network failures that prove nobody wants to use social networks. QED.
EvangelicalPig 8y ago
> Have we learned nothing from literally every single project the that has tried to "free" us from the evils of the centralized services? Here ya go: https://en.wikipedia.org/wiki/Comparison_of_software_and_pro.. https://en.wikipedia.org/wiki/Comparison_of_software_and_pro.... Slowly they all go from Active to Dead or Stalled. I think the issue is none of them have commercial backing and are some "fringe nerd project. The problem is unless it's a nonprofit (good luck with that) there is no commercial interest in producing a decentralized platform. (See Google and Facebook killing off XMPP for a small example of this)
lucideer 8y ago
> We have moved towards a world of centralized services because that's what people want. We have moved towards a world of centralised services because decentralised ones are—by definition—difficult to monetise for the platform/software-maintainer(s). What people want is irrelevant when all potential options have to be initially well-funded: the choice is too restrictively limited to perform an honest comparison. > To many people it's the centralized services that "are" the Internet That's because they dominate, because they're profitable for the central host. > If you want to solve the problems with YouTube you have to do it with another YouTube. Build another centralized service that fixes the problems that YouTube has The problems with YoutTube are caused by it being centralised, so this proposal is paradoxical. > they all go from Active to Dead or Stalled It's a difficult challenge. It's David v Goliath. Many fail. > YouTube's audience doesn't care about anything at decentralized network has to offer. It's just entertainment for them. They just want to watch videos and shitpost in the comments. YouTube's audience want to be able to find entertaining/useful/educational/whatever content according to what they've come to Youtube looking for, and promptly leave after having been entertained/educated/generally fulfilled. Youtube is not incentivised to give them this; in fact it's strongly incentivised to make what they're looking for as difficult to find as possible so that they'll spend hours and hours binging vacuous addictive content looking for something but never quite finding it and getting no satisfaction from the experience. This is not something that's unique to YouTube, it's a property of any model whereby the profits of the host/platform maintainer are tightly linked to users spending as much time navigating the platform as possible (see also Facebook/Instagram/Pinterest infinite scroll psychology).
- scoggs 8y ago
  >in fact it's strongly incentivised to make what they're looking for as difficult to find as possible so that they'll spend hours and hours binging vacuous addictive content looking for something but never quite finding it and getting no satisfaction from the experience. Reddit and Original (as in when it was college email addresses only) Facebook are / were really good at this. You may go to either one with some idea of what you want to check out but hours later realize you went down a rabbit hole that never ended. I'm not too sure Youtube is anywhere near as good as Reddit is or Facebook used to be but some people are right in saying that some of the changes Youtube has made makes it easier for a user to snap back to reality and realize it's time to get off of Youtube because it's not serving you anything you want to watch -- because like you said it's hard to find by design.
karmakaze 8y ago
These centralized services just happen to be the current iteration. There will be others. It's hard to do make decentralized software well and up to now hasn't been the main thing so we have centralized software. Before TCP/IP there were many competing networking 'standards' put out by various vendors, some were local nets but others also did routing, but TCP/IP has superceded them and many application protocols are created just for the purpose of interoperability and decentralization.
danShumway 8y ago
> We have moved towards a world of centralized services because that's what people want. No, nobody cares about centralization. Centralization is a mechanism that provides some of the values people do care about; convenience, indexing, free hosting, simple interfaces. But it's a mistake to conflate centralization with the root cause that makes many of these services fail. If centralization was actually what people cared about, Google Plus would have beaten Facebook. After all, it's way more centralized than Facebook is. But it's not more convenient, better indexed, or simpler. It's also a mistake to look at the list of failed efforts and say, "well, decentralized just doesn't work." Bear in mind that the list of failed centralized social networks is actually a great deal longer, and the chance of any one of them succeeding is just about as grim. The reality is that most social networks fail, period. There are some compelling reasons why from an architectural standpoint centralization might make it easier to succeed, but the average person on the street can't even tell you what these words mean. They don't care. And if a decentralized service comes around that's more convenient than the alternatives, they'll switch. Because they don't care.
- StreamBright 8y ago
  >>> No, nobody cares about centralization. Great so it is not an important aspect of the problem. You just proved to yourself why PeerTube does not matter.
- hardwaresofton 8y ago
  Chiming in to say I agree with this viewpoint -- I think conflating centralization as the only way to provide the convenience, indexing, free hosting, etc is definitely what happens when people say that decentralized solutions can't work. Two big blockers that I rarely see discussed for the decentralized future coming to pass: - (D)DNS for the 1+ persistent devices per/user (IPV6 support/rollout basically) - Lowering the barrier to entry for running a server on a users' behalf If we really want the internet to be a web, we need to make it easier to run servers that act on behalf of users/hold their data, that they control. If you'll indulge me, I've had in the back of my mind a project for a long time, a <$30 home-server (I'd call it "the tube" or "pipe") that has a cheaper but good-enough set of functionality that you get from the internet giants today: - Document management and storage - Backups (connect your local "pipe" to a backup storage provider) - Chat/Messaging, SIP + Video conferencing - Local networked/mesh services spanning from blogging, social networking, ecommerce, whatever. I'm very much of the I-just-want-to-write-software so I preoccupy myself thinking of what tech I'd use to make it happen, but I think this is the kind of project that people need to really embrace the internet as a sparse group of interconnected nodes. I also think that humans are ready to accept this kind of structure -- we already have it! People can understand that if you want a storefront to be up all the time, you need to have something run it all the time, the power to run it just isn't in their hands yet. Of course, startups are hard, hardware startups are harder, and consumer product hardware+software startups are pretty hard, so it's much easier said than done. Decentralized systems absolutely can work, but the amount of work, fore-thought, and careful construction with consideration for generating network effects and making it easier for people to run and use -- basically good product design -- is just missing in the open source world, because it's hard.
hitekker 8y ago
You are correct. The outpouring against your position is simply defense of the premise that "centralized is 99% bad, decentralized is 99% good". If that premise isn't true, if being decentralized doesn't always lead to a promised land, what then, does that mean for the companies or projects that promote decentralization as their main reason for existence? As always, people need to defend their jobs and more importantly, their source of income.
resters 8y ago
You are making the argument that because centralized systems have been proven to be able to solve certain kinds of problems in an acceptable way, that it follows that centralization was the magic bullet that made the systems successful. This argument is analogous to saying that because all wars that have been won to date have been fought using swords, that all future victories will be won with swords. In the myopia of any era, such statements seem trivially true, yet they are shown to be false again and again. Let's break down what Youtube actually consists of: It's a hosting platform for videos, a security model, a content recommendation/discovery platform, a piracy/abuse prevention moderation system, an advertising platform, and an affiliate platform (content creators make some of the ad revenue). All of these can be created as federated systems. There are some significant inefficiencies that give a federated system a bit of a disadvantage, but building it in a decentralized way also offers tremendous advantages. Such a system will certainly replace Youtube at some point in the future as long as each of the mechanisms Youtube offers is available and performs with similar quality to what Youtube offers. We've seen some fairly remarkable examples of distributed cooperation (blockchain systems) that many would have thought impossible a few years ago. Critics of such systems still point to centralized control as the inevitable way of humanity, but get proven wrong every day as decentralized systems gain credibility. Credibility is the key. You are right that the average person who wants to flip around and watch funny content doesn't care how the system was built, but average users do care about the massively annoying ads, the freemium model and the myriad dark patterns that try to squeeze every last penny out of the ecosystem. Analogous to how Microsoft took a big nose dive soon after reaching "peak annoying" with its tiered OS offerings, $1000 office suite and many other gimmicks that signal impending disruption, Youtube is getting very close to this pinnacle. The little gimmicks that gradually autoplay you toward the most viral content and ads are pernicious dark patterns that create psychological pain for users and even though they are viewed as successful by the "engagement and profitability metrics" approach, they are creating a bigger and bigger void for some other service to fill. The problem is, Youtube has created a category leading centralized product, which is now very easy to replicate. Just as Bing was able to easily steal much of Google's search market simply by copying many years of Google R&D in a few months, playing trending videos with interruption marketing is now a known product and any missteps can result in massive loss of market share. Google knows this and will not take risks to expand the platform's features. So all it takes for a federated model to succeed is some insight into the things that the market wants that Google is too blinded by dopamine-fueled ad watching to worry about building.

smnplk 8y ago

Another point of decentralised services is cutting out the middleman and this is what should be appealing to content creators.

themagician 8y ago
Not really. Content creators often look for middlemen to represent them. That's why they join networks that works directly with YouTube. Most of the views to people who do YouTube as a job. When you look on that trending list, 90% of those videos are people waking up and going to a job. The last thing they want is more bosses and more logistics.
- cyphar 8y ago
  > That's why they join networks that works directly with YouTube. They usually join networks more out of necessity than choice (from memory you get access to more "perks" if you're with a network). Many YouTubers bemoan the existence of multi-channel networks because they take a lot of power away from content creators (even more than YouTube already does). > When you look on that trending list [...] These days it's pretty obvious that the trending list is effectively all paid promotions. It's not uncommon that content that has already gone viral is nowhere near the trending list (and the trending list is instead videos advertising new movies). In the "old days" of YouTube, the front page was hand-selected content that someone working at YouTube thought people would enjoy. These days it's all run by opaque algorithms. > The last thing they want is more bosses and more logistics. Given how many YouTubers are complaining about their current bosses being completely awful, I'd think the last thing they want is to continue with the status quo.
- altfredd 8y ago
  > Content creators often look for middlemen to represent them. Oh boy. Bees are looking for badgers?! Content creators are looking for support and high reliability. "The middleman" can sometimes deliver those things... Until they suddenly boot you out, because their algorithm determined, that your videos don't deliver enough bucks in advertising revenue to keep them afloat. Of course, "the algorithm" is merely poor excuse for carefully orchestrated scenario, which might have as well resulted from manual intervention (and most famous precedents are indeed backed by staff-approved decisions). Google have put great deal of effort into pretending, that their services have no support teams and no living humans behind them, but in practice places like YouTube are censored to hell and back, — and all of that censorship happens at expense of "content creators".

taurath 8y ago

> We have moved towards a world of centralized services because that's what people want. Almost there. People want INTEROPERABLE services. If that occurs because one place has a monopoly, that’s okay. If everyone can participate across any platform they choose that has the features they want (privacy, a close knit community, etc) that’s better.

charlchi 8y ago
Ease of user interoperability trumps nearly everything else. Most users don't care about your conceptual and moral motivations as an engineer. All we need to do is change the tech serving the problem underneath without necessarily changing the user-facing solution too much. It's such a simple idea yet people seem unable to grasp it; and end up whining about issues that have minimal relevance. There's a gap in the thinking somewhere. Plus, please carry on arguing about why I need to view 3 advertisements simultaneously (yes, use the YouTube mobile app). Engineers who have been reading the news and seeing the results of our latest technological models for the past 10 years don't want the same broken manure that is good for the bottom line but bad for people. Literally stop thinking about the bottom line for one second and focus on the people and maybe you'll see sometimes it's not about choking every last bit of stink out of the old model but completely changing the game. Apparently these are broken dreams though.

feral 8y ago

I don't know whether you are right, but I don't find your argument persuasive. You could have made a parallel argument, before Youtube, about how people wanted only professional content, why a service for user created content would never work etc, because that's all that had worked, up to that point. Just because all the decentralized projects have failed so far doesn't mean they are doomed. Its really early days yet. Youtube is just over a decade old. HBO is 5 decades old. Clearly decentralized services are harder to build and get right. But that means there's still plenty to figure out, and plenty of potential. > YouTube's audience doesn't care about anything at decentralized network has to offer. It's just entertainment for them. They just want to watch videos and shitpost in the comments. People change, they grow in sophistication. E.g. look at Facebook - at the start it was full of spammy viral activity notifications for crappy games. People got sick of that and it was forced to change. Now its full of other crappy stuff, but which people are getting sick of now too.

themagician 8y ago
We grow down in sophistication in tech. Everything goes to the lowest common denominator, then it falls further. The web used to be sophisticated and it was only for nerds.
- feral 8y ago
  I think your argument is mostly running into selection issues. If its the middle ages and the only people who have learned to read are the very most educated part of the the population, then the average printed material is going to be intellectual. As the reading population expands over the next 1000 years, and the average book hence becomes less scholarly, you might naively conclude that everything is becoming 'lowest common denominator'. But overall, education standards + sophistication are rising hugely. Sure, when only nerds could access the web, it was super sophisticated. And it had to get easier to go mainstream. But that doesn't mean the mainstream trend is to always dumb down forever, as you are arguing. Apple mouse used to have only one button, but now look at the power of the multitouch UX almost everyone uses on their smartphone daily. Look at Gmail compared to Hotmail v1, Youtube compared to cable TV etc. Its easy to forget how powerful and sophisticated they are. People have learned to do really sophisticated things, create, communicate etc. Sure, things dumbed down since only CS grad students used the Internet. But now that growth is slowing, I bet tastes are going to increase in sophistication again.

thevardanian 8y ago

You listing out a bunch of decentralized, protocol based solutions means that there's competition in the market, which is evidence of a new market that is GROWING. The opposite of a dead market. Youtube is also a the result of such a competitive market. There used to be dozens of video hosting platforms. The same applies to almost every other company. Facebook, Google, Amazon, Ebay, etc. all grew up in very competitive markets. Competitive markets means that there's demand. Decentralization does one very important thing. It makes it more profitable, and cheaper to make content accessible to your audience, while giving you the freedom of making money. That is to say decentralization amplifies the ability of computers to make work what Youtube and Google require billions to do. It DISTRIBUTES that workload among many independent workers, and lets the market decide how valuable their input is. Secondly centralized solutions can exist in a decentralized environment. Content distribution should always have been decentralized, making profit on content is the centralization problem. You're getting content distribution, and profitability mixed up.

stale2002 8y ago
Umm, I don't think I agree with you on the idea that decentralization decreases costs for contract distribution. Distributing content to millions of people requires massive investment. The total cost would likely INCREASE massively, as my home computer is much worse at acting as a CDN than Google. Sure, the costs get DISTRIBUTED, to places where maybe you aren't the one paying for it. But SOMEONE is still paying that money. And they are doing much less efficient than a centralized service with thousands of engineers.
- thevardanian 8y ago
  You don't run your website, necessarily, on a local home computer, why would you assume the same for this?
  - stale2002 8y ago
    So it's not decentralized then? I thought that was the entire point. But even if it is not literal home PCs, and is instead some sort of middle road between massive company and home PC, my point still stands. Your 10 person company hosting content isn't going to be able to stream it live to 20 thousand people. That takes lots and lots of money and engineering talent to do right.

tombh 8y ago

Is this not the archetypal battle of modern software? It makes me think of Eric Raymond's classic book "The Cathedral and the Bazaar". At which point the other side rolls out the Linux precedent - the decentralised success. I believe the world very much does want decentralised services, it's just that such services have quite different measures of success.

progval 8y ago

> Build another centralized service that fixes the problems that YouTube has, Then just create a huge PeerTube instance, and you'll get the benefits of both worlds. Plus, you'll pay for less bandwidth than Youtube.

api 8y ago

You're making an assumption-- namely that federated or decentralized systems must be harder to use than centralized ones. Present decentralized alternatives usually are. There are numerous reasons. Some have to do with unsolved or inadequately solved problems in distributed systems engineering, but most are I think more pragmatic. The two largest reasons for distributed/decentralized/federated services sucking are (IMHO): (1) Most of these things are developed by geeks who lack an appreciation for the critical importance of user experience and lack expertise in areas like GUI and UX design. The result is products by geeks for geeks and that fail to get much traction outside small geek-centric markets. (2) There is no good economic model, which means that there's no incentive for developers to do the parts of software development that are not fun. Solving grand problems is fun, but fixing a trillion trillion edge case UI/UX and platform bugs is not. Supporting users definitely is not. But those are the things that take something from a niche or nerds-only product to something regular people will use. Centralized platforms have two economic models: charging directly for use or monetizing the user through surveillance capitalism. (The latter is more common.)

_jal 8y ago

> YouTube's audience doesn't care about anything at decentralized You're speaking for quite a huge number of people about their preferences vs. something that does not exist. "Content creators" didn't care about the internet, until they did. For that matter, apparently I'm not a member of the "audience", either. Maybe you're right. I'd need to see some argument other than "because I think so", and look at these failures[1], though. As far as bald assertions, here's mine: Video consumers on average don't know what the word "federated" means in the context of networked services. Nor do they care. They don't know that email is federated. They know they hit "send", and their words pop up in grandma's inbox. Users don't care about the tech under the hood. They want the payloff. If the economics of federated content networks work, they will take off once someone figures out the proper feature set, placement and happens into a bit of luck. If not, we're doomed to sharecropping on FaceGoogazon and this whole conversation is pointless. [1] I have here a list of social network failures that prove nobody wants to use social networks. QED.

EvangelicalPig 8y ago

> Have we learned nothing from literally every single project the that has tried to "free" us from the evils of the centralized services? Here ya go: https://en.wikipedia.org/wiki/Comparison_of_software_and_pro.. https://en.wikipedia.org/wiki/Comparison_of_software_and_pro.... Slowly they all go from Active to Dead or Stalled. I think the issue is none of them have commercial backing and are some "fringe nerd project. The problem is unless it's a nonprofit (good luck with that) there is no commercial interest in producing a decentralized platform. (See Google and Facebook killing off XMPP for a small example of this)

lucideer 8y ago

> We have moved towards a world of centralized services because that's what people want. We have moved towards a world of centralised services because decentralised ones are—by definition—difficult to monetise for the platform/software-maintainer(s). What people want is irrelevant when all potential options have to be initially well-funded: the choice is too restrictively limited to perform an honest comparison. > To many people it's the centralized services that "are" the Internet That's because they dominate, because they're profitable for the central host. > If you want to solve the problems with YouTube you have to do it with another YouTube. Build another centralized service that fixes the problems that YouTube has The problems with YoutTube are caused by it being centralised, so this proposal is paradoxical. > they all go from Active to Dead or Stalled It's a difficult challenge. It's David v Goliath. Many fail. > YouTube's audience doesn't care about anything at decentralized network has to offer. It's just entertainment for them. They just want to watch videos and shitpost in the comments. YouTube's audience want to be able to find entertaining/useful/educational/whatever content according to what they've come to Youtube looking for, and promptly leave after having been entertained/educated/generally fulfilled. Youtube is not incentivised to give them this; in fact it's strongly incentivised to make what they're looking for as difficult to find as possible so that they'll spend hours and hours binging vacuous addictive content looking for something but never quite finding it and getting no satisfaction from the experience. This is not something that's unique to YouTube, it's a property of any model whereby the profits of the host/platform maintainer are tightly linked to users spending as much time navigating the platform as possible (see also Facebook/Instagram/Pinterest infinite scroll psychology).

scoggs 8y ago
>in fact it's strongly incentivised to make what they're looking for as difficult to find as possible so that they'll spend hours and hours binging vacuous addictive content looking for something but never quite finding it and getting no satisfaction from the experience. Reddit and Original (as in when it was college email addresses only) Facebook are / were really good at this. You may go to either one with some idea of what you want to check out but hours later realize you went down a rabbit hole that never ended. I'm not too sure Youtube is anywhere near as good as Reddit is or Facebook used to be but some people are right in saying that some of the changes Youtube has made makes it easier for a user to snap back to reality and realize it's time to get off of Youtube because it's not serving you anything you want to watch -- because like you said it's hard to find by design.

karmakaze 8y ago

These centralized services just happen to be the current iteration. There will be others. It's hard to do make decentralized software well and up to now hasn't been the main thing so we have centralized software. Before TCP/IP there were many competing networking 'standards' put out by various vendors, some were local nets but others also did routing, but TCP/IP has superceded them and many application protocols are created just for the purpose of interoperability and decentralization.

danShumway 8y ago

> We have moved towards a world of centralized services because that's what people want. No, nobody cares about centralization. Centralization is a mechanism that provides some of the values people do care about; convenience, indexing, free hosting, simple interfaces. But it's a mistake to conflate centralization with the root cause that makes many of these services fail. If centralization was actually what people cared about, Google Plus would have beaten Facebook. After all, it's way more centralized than Facebook is. But it's not more convenient, better indexed, or simpler. It's also a mistake to look at the list of failed efforts and say, "well, decentralized just doesn't work." Bear in mind that the list of failed centralized social networks is actually a great deal longer, and the chance of any one of them succeeding is just about as grim. The reality is that most social networks fail, period. There are some compelling reasons why from an architectural standpoint centralization might make it easier to succeed, but the average person on the street can't even tell you what these words mean. They don't care. And if a decentralized service comes around that's more convenient than the alternatives, they'll switch. Because they don't care.

StreamBright 8y ago
>>> No, nobody cares about centralization. Great so it is not an important aspect of the problem. You just proved to yourself why PeerTube does not matter.
hardwaresofton 8y ago
Chiming in to say I agree with this viewpoint -- I think conflating centralization as the only way to provide the convenience, indexing, free hosting, etc is definitely what happens when people say that decentralized solutions can't work. Two big blockers that I rarely see discussed for the decentralized future coming to pass: - (D)DNS for the 1+ persistent devices per/user (IPV6 support/rollout basically) - Lowering the barrier to entry for running a server on a users' behalf If we really want the internet to be a web, we need to make it easier to run servers that act on behalf of users/hold their data, that they control. If you'll indulge me, I've had in the back of my mind a project for a long time, a <$30 home-server (I'd call it "the tube" or "pipe") that has a cheaper but good-enough set of functionality that you get from the internet giants today: - Document management and storage - Backups (connect your local "pipe" to a backup storage provider) - Chat/Messaging, SIP + Video conferencing - Local networked/mesh services spanning from blogging, social networking, ecommerce, whatever. I'm very much of the I-just-want-to-write-software so I preoccupy myself thinking of what tech I'd use to make it happen, but I think this is the kind of project that people need to really embrace the internet as a sparse group of interconnected nodes. I also think that humans are ready to accept this kind of structure -- we already have it! People can understand that if you want a storefront to be up all the time, you need to have something run it all the time, the power to run it just isn't in their hands yet. Of course, startups are hard, hardware startups are harder, and consumer product hardware+software startups are pretty hard, so it's much easier said than done. Decentralized systems absolutely can work, but the amount of work, fore-thought, and careful construction with consideration for generating network effects and making it easier for people to run and use -- basically good product design -- is just missing in the open source world, because it's hard.

hitekker 8y ago

You are correct. The outpouring against your position is simply defense of the premise that "centralized is 99% bad, decentralized is 99% good". If that premise isn't true, if being decentralized doesn't always lead to a promised land, what then, does that mean for the companies or projects that promote decentralization as their main reason for existence? As always, people need to defend their jobs and more importantly, their source of income.

resters 8y ago

You are making the argument that because centralized systems have been proven to be able to solve certain kinds of problems in an acceptable way, that it follows that centralization was the magic bullet that made the systems successful. This argument is analogous to saying that because all wars that have been won to date have been fought using swords, that all future victories will be won with swords. In the myopia of any era, such statements seem trivially true, yet they are shown to be false again and again. Let's break down what Youtube actually consists of: It's a hosting platform for videos, a security model, a content recommendation/discovery platform, a piracy/abuse prevention moderation system, an advertising platform, and an affiliate platform (content creators make some of the ad revenue). All of these can be created as federated systems. There are some significant inefficiencies that give a federated system a bit of a disadvantage, but building it in a decentralized way also offers tremendous advantages. Such a system will certainly replace Youtube at some point in the future as long as each of the mechanisms Youtube offers is available and performs with similar quality to what Youtube offers. We've seen some fairly remarkable examples of distributed cooperation (blockchain systems) that many would have thought impossible a few years ago. Critics of such systems still point to centralized control as the inevitable way of humanity, but get proven wrong every day as decentralized systems gain credibility. Credibility is the key. You are right that the average person who wants to flip around and watch funny content doesn't care how the system was built, but average users do care about the massively annoying ads, the freemium model and the myriad dark patterns that try to squeeze every last penny out of the ecosystem. Analogous to how Microsoft took a big nose dive soon after reaching "peak annoying" with its tiered OS offerings, $1000 office suite and many other gimmicks that signal impending disruption, Youtube is getting very close to this pinnacle. The little gimmicks that gradually autoplay you toward the most viral content and ads are pernicious dark patterns that create psychological pain for users and even though they are viewed as successful by the "engagement and profitability metrics" approach, they are creating a bigger and bigger void for some other service to fill. The problem is, Youtube has created a category leading centralized product, which is now very easy to replicate. Just as Bing was able to easily steal much of Google's search market simply by copying many years of Google R&D in a few months, playing trending videos with interruption marketing is now a known product and any missteps can result in massive loss of market share. Google knows this and will not take risks to expand the platform's features. So all it takes for a federated model to succeed is some insight into the things that the market wants that Google is too blinded by dopamine-fueled ad watching to worry about building.