Decentralization is not resource-efficient and its cost per unit will be always higher than any centralized alternative. That is a given. It is the cost of having all that redundancy and flexibility. You can not reach economies of scale easily.
With that said, I believe prices will go down for B2C. A few reasons why I believe so:
- Some of these services will consolidate into larger offerings and make a more featureful service. Pixelfed is already experimenting with a different frontend where users can share more than just photos (UI-wise, it looks a lot like early/cool Tumblr)
- A larger number of professional providers that will run instances in the same way that people were running email providers. With Communick I can run Matrix, Mastodon and XMPP and I am charging $5/10 users/3 months during the "soft launch/beta" phase, which is basically $0.17/user/month and will let me hit break even point with less than a hundred paid accounts. I can increase this 10-fold and it still wouldn't be as expensive as any of the current major providers.
- Media storage costs completely dominate the rest of other infra expenses, and IPFS can save the day here. It's not hard to imagine an architecture where servers provide only a cache layer for media serving and the less-accessed data lives on users' clients.
- The real money is in hosting for business/enterprise/white-labeling anyway. As much as I want Communick to be successful and I hope I start getting more people to "get it", I am using it as lab to make sure that I can offer a reliable media presence / communication service for companies, and then "Communick B2C" would be mostly a loss-leader.
I absolutely agree. It’s just a messaging protocol. There’s no reason for all of these apps to be so niche.
And I agree that there’s no inherent reason that all of the current servers are terribad. Although it is strange that everyone interested in decentralization seems to also be interested in bad technology. Though of course this is slowly changing as reality sets in.
Communick sounds interesting. If you can really bundle all those apps together for that cheap? It makes me wonder why Element Matrix hosting is such a ripoff. Or why every other host is such a ripoff.
I agree about IPFS too. Specifically though I’d like to see Sia used as storage because it’s actually working today as opposed to Filecoin and all that which is not close to functioning yet.
With that said, I believe prices will go down for B2C. A few reasons why I believe so:
- Some of these services will consolidate into larger offerings and make a more featureful service. Pixelfed is already experimenting with a different frontend where users can share more than just photos (UI-wise, it looks a lot like early/cool Tumblr)
- A larger number of professional providers that will run instances in the same way that people were running email providers. With Communick I can run Matrix, Mastodon and XMPP and I am charging $5/10 users/3 months during the "soft launch/beta" phase, which is basically $0.17/user/month and will let me hit break even point with less than a hundred paid accounts. I can increase this 10-fold and it still wouldn't be as expensive as any of the current major providers.
- Media storage costs completely dominate the rest of other infra expenses, and IPFS can save the day here. It's not hard to imagine an architecture where servers provide only a cache layer for media serving and the less-accessed data lives on users' clients.
- The real money is in hosting for business/enterprise/white-labeling anyway. As much as I want Communick to be successful and I hope I start getting more people to "get it", I am using it as lab to make sure that I can offer a reliable media presence / communication service for companies, and then "Communick B2C" would be mostly a loss-leader.