4 ms·

> A distributed web with many small nodes would still be hard to control and police effectively. I would guess that today's distributed web (IPFS, etc.) has mo

> A distributed web with many small nodes would still be hard to control and police effectively. I would guess that today's distributed web (IPFS, etc.) has more content, and users, than the "wild web" of 1992 or whatever era you fondly recollect. Yes, centralized players created tools that produced exponentially more content (and control). But it's lazy to blame that peripheral phenomenon for a perceived "loss". The wild web still exists, and has the potential to grow (cf. Urbit). What's stopping it?

cookiecaper 9y ago
IPFS is stopped by its insistence on things that are considered anti-features by non-cryptopunks, like immutability. There's a lot of promise behind the "distributed web" or "P2P web" prototypes, but they haven't been effectively commoditized, and their value proposition over the top-level "spyweb" isn't strong enough for normal users who don't understand the inner workings. In fact, they tend to offer less end-user benefit, since a P2P web is necessarily slower as peers negotiate connections, deal with vaporizing peers, etc. Consumers care only about getting their desires fulfilled as quickly and easily as possible. If you want a new web to dominate, you must make something that gives the consumers and/or the companies upstream from them an incentive to exert the effort to change. Personally, I believe the web is already sufficiently "decentralized" in the necessary sense. We just have legal restrictions that prevent true competition with the tech conglomerates, as they're allowed to effectively claim exclusive domain over all user-generated content. There's no reason I shouldn't be able to go to NewGoog or NewFace and multiplex in the available data streams hosted by the web nodes at google.com and facebook.com, except that this violates the CFAA and the Copyright Act, and Google and Facebook can and will sue you to make you stop. This is the primary constraint that has led to the recentralization of the web. What do you expect when you make it a crime to send any type of undesired communication to a server? The CFAA doesn't just prevent harmful or disruptive access, but all unauthorized access. In practice, this is almost always used to stymie small innovators who are putting the data from a data source like Facebook to better use. This is fundamentally anti-competitive, and again, it's no wonder that the web is coalescing into the hands of a few central controllers. It has never been more apparent that the user is the product being sold and peddled. We need to reorient our perspective of the internet with users, not services, as the central entity, and give them full control over the data that they generate. Then, the user will be free again, without the need for complex P2P solutions.
- StavrosK 9y ago
  I don't know, I think IPFS has at least a good few use cases. For example, it's great for posting censorship-resistant content, and I wrote something to do just that: http://ipfessay.stavros.io/ http://ipfessay.stavros.io/ There are also things like OpenBazaar and a whole host of other applications running on IPFS that are quite useful. Sure, it may not be mainstream, but nothing starts out being mainstream.
  - didip 9y ago
    I have always wondered about IPFS and its relationship with DNS. What's stopping authoritarian entity from poisoning/redirecting ipfs.io to static page?
    - detaro 9y ago
      Nothing, but that would hopefully primarily encourage people to set up IPFS clients themselves instead of relying on the public bridge. (Which is IMHO one of the weaknesses of IPFS, that they haven't made that more prominent/easier, through browser plugins or a dedicated IPFS-browser. Compare to e.g. the publicity boost Dat got from Beaker Browser)
      - StavrosK 9y ago
        This is a very nice Chrome extension: https://chrome.google.com/webstore/detail/ipfs-station/kckhgoigikkadogfdiojcblegfhdnjei https://chrome.google.com/webstore/detail/ipfs-station/kckhg... It'll redirect all IPFS URLs to your local daemon.
        detaro 9y ago
        Ohh, interesting, that seems to have most of the things I was looking for. Not sure why I missed it when I went looking for one recently. Thanks!
        9y ago

cookiecaper 9y ago

IPFS is stopped by its insistence on things that are considered anti-features by non-cryptopunks, like immutability. There's a lot of promise behind the "distributed web" or "P2P web" prototypes, but they haven't been effectively commoditized, and their value proposition over the top-level "spyweb" isn't strong enough for normal users who don't understand the inner workings. In fact, they tend to offer less end-user benefit, since a P2P web is necessarily slower as peers negotiate connections, deal with vaporizing peers, etc. Consumers care only about getting their desires fulfilled as quickly and easily as possible. If you want a new web to dominate, you must make something that gives the consumers and/or the companies upstream from them an incentive to exert the effort to change. Personally, I believe the web is already sufficiently "decentralized" in the necessary sense. We just have legal restrictions that prevent true competition with the tech conglomerates, as they're allowed to effectively claim exclusive domain over all user-generated content. There's no reason I shouldn't be able to go to NewGoog or NewFace and multiplex in the available data streams hosted by the web nodes at google.com and facebook.com, except that this violates the CFAA and the Copyright Act, and Google and Facebook can and will sue you to make you stop. This is the primary constraint that has led to the recentralization of the web. What do you expect when you make it a crime to send any type of undesired communication to a server? The CFAA doesn't just prevent harmful or disruptive access, but all unauthorized access. In practice, this is almost always used to stymie small innovators who are putting the data from a data source like Facebook to better use. This is fundamentally anti-competitive, and again, it's no wonder that the web is coalescing into the hands of a few central controllers. It has never been more apparent that the user is the product being sold and peddled. We need to reorient our perspective of the internet with users, not services, as the central entity, and give them full control over the data that they generate. Then, the user will be free again, without the need for complex P2P solutions.

StavrosK 9y ago
I don't know, I think IPFS has at least a good few use cases. For example, it's great for posting censorship-resistant content, and I wrote something to do just that: http://ipfessay.stavros.io/ http://ipfessay.stavros.io/ There are also things like OpenBazaar and a whole host of other applications running on IPFS that are quite useful. Sure, it may not be mainstream, but nothing starts out being mainstream.
- didip 9y ago
  I have always wondered about IPFS and its relationship with DNS. What's stopping authoritarian entity from poisoning/redirecting ipfs.io to static page?
  - detaro 9y ago
    Nothing, but that would hopefully primarily encourage people to set up IPFS clients themselves instead of relying on the public bridge. (Which is IMHO one of the weaknesses of IPFS, that they haven't made that more prominent/easier, through browser plugins or a dedicated IPFS-browser. Compare to e.g. the publicity boost Dat got from Beaker Browser)
    - StavrosK 9y ago
      This is a very nice Chrome extension: https://chrome.google.com/webstore/detail/ipfs-station/kckhgoigikkadogfdiojcblegfhdnjei https://chrome.google.com/webstore/detail/ipfs-station/kckhg... It'll redirect all IPFS URLs to your local daemon.
      - detaro 9y ago
        Ohh, interesting, that seems to have most of the things I was looking for. Not sure why I missed it when I went looking for one recently. Thanks!
        9y ago