4 ms·

I understand this, but you seem to have ignored my point in that it is trivially easy to make something similar for email and yet no one did, and that it is unc

I understand this, but you seem to have ignored my point in that it is trivially easy to make something similar for email and yet no one did, and that it is unclear what’s to stop the same situation from happening again when 90% of users have a lemmy.org or whatever address. Or like happened with the original decentralized web. there was once a thing called a “blog” that you could host on your own server from your home, but these days everyone is “blogging” through fb twitter and instagram (or your choice of a few platforms). Why is this time different?

sangnoir 2y ago
I failed to respond simply because I didn't realize you were making that point. To me, the existence of large players in a decentralized system is not something in need of addressing or fixing; IMO, decentralization is not the right tool to achieve egalitarianism - but it's a great way to avoid a monoculture. The fediverse, like email and most other human systems that benefit from economies of scale will have a Pareto distribution with very large free providers > ISP / work and community instances > individual instances. Banking is decentralized and interoperable, this doesn't prevent banking giants from existing the world over - I expect them to exist, I also am in favor of smaller regional and local banks and credit unions.
- shwouchk 2y ago
  decentralization is not some magical binary property. Real world systems can be more, or less, decentralized. To start with, yes banking giants exists but everyone finds this undesirable to a certain extent and regulators would object if suddenly eg elon musk decided to buy out goldman and chase. banks are highly regulated and enjoy completely different economies than email. Banks of the same nation are often legally required to be interoperable with others. Yes, there are still big players and you can eg get sanctioned by the US - which is why BRIC countries are working on alternative banking networks. with banks, most jurisdictions would disallow a merger of two large banks specifically from decentralization considerations - guess what would happen if two email providers merge? exactly nothing. With email, if gmail decide to ban you, you are unable to communicate with 90% of people with no recourse. A support engineer can decide to ban you on a whim and besides crying on social media, you can’t do much. With banks, you need a large and powerful nation to tell other banks not to play with you, and you can contest this decision legally. Most importantly, with banking money is the product and every transaction brings revenue, so there’s a strong incentive to cooperate. With email - and AP - cooperation incurs a cost on all parties instead and benefits are distributed and vague in monetary terms. Finally, banks are required to deal with you as a customer - provided you meet certain requirements. An email provider can ban you for no reason and your only recourse is arbitration. Compare to bitcoin. Will you still be calling it decentralized when one pool controls 90% of compute (i don’t think this can really happen tbh as there are disincentives to go 51% without abusing the network)? Don’t get me wrong AP sure feels in the right direction compared to eg FB. Just on the principle of being able to run it completely in your own. However it would be nice if whatever new alternative that grew popular treated users, rather than providers, democratically.
  - sangnoir 2y ago
    > With email, if gmail decide to ban you, you are unable to communicate with 90% of people with no recourse To reiterate my point (and perhaps what you said in your first sentence) this is completely unrelated to the decentralized nature of email, and is proof that decentralization is ill-suited to avoid that problem: legislation and oversight are what works for banking. "Decentralized" has never meant "there are no dominant players" Edit: when it comes to tech platforms or protocols, the ability to decentralize is binary: it's present or absent. There is one Twitter/Instagram namespace with no other instances, there are countless email MTA and MDA and standards for interoperability defined in RFCs. That GMail is dominant is not a fault of the protocol - it happened to be a superior product for reasons not included in the protocol like generous mailbox storage quotas, and superior spam identification.
    - shwouchk 2y ago
      I understand your claim though dont agree on several counts - 1. To start, if decentralization does not "solve the problem" then what do all AP platforms bring to the table? There was already a decentralized facebook before mastodon - it used to be called a "web site". You could get a feed with just what you wanted by subscribing to rss feeds, which were also decentralized. You were not tied to 140 characters per post either, and nothing stopped someone from making an rss client that clipped posts at 140 characters in the the feed or whatever. What problem does AP solve for me? Indexing all that content? The hosting? If the whole thing is about getting banned from the domain twitter.com, open your own domain and publish your content - problem solved. 2. If we're solving a solved problem, why not bring something new to the table? eg anonymity of nosr. 3. If "decentralization is not the solution", how come other protocols - eg mentioned above - claim (convincingly imho) to offer an actual improvement if not solution? 4. You claim "decentralization is not the solution" then define some ad hoc line in the sand regarding what is "decentralized" and then make some bold claims based on that. Decentralized could mean many things: could be like ssh keys, could be like gpg keys with directories and web of trust, could be like dns or certificate authorities. Could be like email. Like torrents or like i2p, or kad. All of these are slightly different. Heck, could even mean like fb vs twitter - you are free to move to a different platform. Takeout your data, import contacts - done. Where you claim that decentralization is not the answer i will say maybe insufficient decentralization is not the answer - with email, AP, and many other protocols, the platform ("domain") is the first class citizen so you get strong platforms and weak users. Regulation will help here just like it helped with banking - ie wont. What makes banks cooperate is mutual desire for profit. Banks are truly fungible for the most part - make users first class and you will get the same for platforms

sangnoir 2y ago

I failed to respond simply because I didn't realize you were making that point. To me, the existence of large players in a decentralized system is not something in need of addressing or fixing; IMO, decentralization is not the right tool to achieve egalitarianism - but it's a great way to avoid a monoculture. The fediverse, like email and most other human systems that benefit from economies of scale will have a Pareto distribution with very large free providers > ISP / work and community instances > individual instances. Banking is decentralized and interoperable, this doesn't prevent banking giants from existing the world over - I expect them to exist, I also am in favor of smaller regional and local banks and credit unions.

shwouchk 2y ago
decentralization is not some magical binary property. Real world systems can be more, or less, decentralized. To start with, yes banking giants exists but everyone finds this undesirable to a certain extent and regulators would object if suddenly eg elon musk decided to buy out goldman and chase. banks are highly regulated and enjoy completely different economies than email. Banks of the same nation are often legally required to be interoperable with others. Yes, there are still big players and you can eg get sanctioned by the US - which is why BRIC countries are working on alternative banking networks. with banks, most jurisdictions would disallow a merger of two large banks specifically from decentralization considerations - guess what would happen if two email providers merge? exactly nothing. With email, if gmail decide to ban you, you are unable to communicate with 90% of people with no recourse. A support engineer can decide to ban you on a whim and besides crying on social media, you can’t do much. With banks, you need a large and powerful nation to tell other banks not to play with you, and you can contest this decision legally. Most importantly, with banking money is the product and every transaction brings revenue, so there’s a strong incentive to cooperate. With email - and AP - cooperation incurs a cost on all parties instead and benefits are distributed and vague in monetary terms. Finally, banks are required to deal with you as a customer - provided you meet certain requirements. An email provider can ban you for no reason and your only recourse is arbitration. Compare to bitcoin. Will you still be calling it decentralized when one pool controls 90% of compute (i don’t think this can really happen tbh as there are disincentives to go 51% without abusing the network)? Don’t get me wrong AP sure feels in the right direction compared to eg FB. Just on the principle of being able to run it completely in your own. However it would be nice if whatever new alternative that grew popular treated users, rather than providers, democratically.
- sangnoir 2y ago
  > With email, if gmail decide to ban you, you are unable to communicate with 90% of people with no recourse To reiterate my point (and perhaps what you said in your first sentence) this is completely unrelated to the decentralized nature of email, and is proof that decentralization is ill-suited to avoid that problem: legislation and oversight are what works for banking. "Decentralized" has never meant "there are no dominant players" Edit: when it comes to tech platforms or protocols, the ability to decentralize is binary: it's present or absent. There is one Twitter/Instagram namespace with no other instances, there are countless email MTA and MDA and standards for interoperability defined in RFCs. That GMail is dominant is not a fault of the protocol - it happened to be a superior product for reasons not included in the protocol like generous mailbox storage quotas, and superior spam identification.
  - shwouchk 2y ago
    I understand your claim though dont agree on several counts - 1. To start, if decentralization does not "solve the problem" then what do all AP platforms bring to the table? There was already a decentralized facebook before mastodon - it used to be called a "web site". You could get a feed with just what you wanted by subscribing to rss feeds, which were also decentralized. You were not tied to 140 characters per post either, and nothing stopped someone from making an rss client that clipped posts at 140 characters in the the feed or whatever. What problem does AP solve for me? Indexing all that content? The hosting? If the whole thing is about getting banned from the domain twitter.com, open your own domain and publish your content - problem solved. 2. If we're solving a solved problem, why not bring something new to the table? eg anonymity of nosr. 3. If "decentralization is not the solution", how come other protocols - eg mentioned above - claim (convincingly imho) to offer an actual improvement if not solution? 4. You claim "decentralization is not the solution" then define some ad hoc line in the sand regarding what is "decentralized" and then make some bold claims based on that. Decentralized could mean many things: could be like ssh keys, could be like gpg keys with directories and web of trust, could be like dns or certificate authorities. Could be like email. Like torrents or like i2p, or kad. All of these are slightly different. Heck, could even mean like fb vs twitter - you are free to move to a different platform. Takeout your data, import contacts - done. Where you claim that decentralization is not the answer i will say maybe insufficient decentralization is not the answer - with email, AP, and many other protocols, the platform ("domain") is the first class citizen so you get strong platforms and weak users. Regulation will help here just like it helped with banking - ie wont. What makes banks cooperate is mutual desire for profit. Banks are truly fungible for the most part - make users first class and you will get the same for platforms