Better discovery of cyberlinks

Current focus of IPFS project is all about discoverability. But there is the room for significant improvement for agents. We talked a lot with @litvintech about it. In general I can see 3 vectors of improvements:

  1. 1-hop content discovery
  2. client integration of decentralized storage protocols
  3. introduction of social mechanics in cyb to encourage altruistic storage

One hope content discovery

In short: If we (1) we embed ipfs node into cyb with deterministic node id computed from common seed and (2) will be able deterministically compute ipfs node id from cyber address, this enable nearly instant content discovery even if content is pinned inly at one peer. After some research we concluded that it’s likely that this ability will not be delivered any time soon because of dependency on several projects. I decide to start this topic as I think this could be a common question in the future.

Integration of decentralized storage protocols

The following protocols are in development:

  • filecoin
  • sia
  • chia
  • maidsafe
  • solana
  • rchain

I expect to integrate Filecoin into Cyb as the first case

Altruistic storage

I would love to hear some ideas on this

3 Likes

The only cases I know for altruistic storage are all build upon importance and / or feeling of belonging / making a change. In some weird sense social networking are such a case (although technically its not like that, I am talking about the fact users believe its like that). A person feels a sense of belonging and “stores” content on his feed, albums, etc (in reality the centralized server pay for it).

The other case, is a sense of importance and making a difference, bit torrent, napster and the rest.
Now, why the introduction. I believe that if we want to see altruistic storage, we need to give people a sense of awareness and belonging. Due to cybers unique architecture, cyber is able to produce cool social networks and / or social cells / groups based on sharing certain semantic interests (at least).

I believe that if people have some kind of a social cell, where they share reputation, contaracts, thoughts, ideas, values, etc. They will be ready to store altruistically. Another case, is hope for the greater good

1 Like

I think that one of the options could be to force validators to store a certain amount of data in the Knowledge Graph. But I do not think that this is a good option.
I can imagine that some non-commercial organizations can store data altrusitically. Science institutes and libraries can also store and share data.
I also believe that altruistic storage can be of several types. The first is the storage of their files and the second is the storage of other people’s files. Storing your files can also be altruistic. For example, the publication and storage of their content.

1 Like

Anything with the word force is a no-no

1 Like

I am also interested in what your forecasts are about how much memory I need to actively link and store my files. I understand that it depends on what files I will store, but we still have to add a lot of data to our knowledge graph. I want to be the one who will actively upload files and links.
How much memory is needed for our Knowledge Graph?
What do you think could be such that a large amount of memory will be more important than the GPU? I understand that they have different importance and application, but if I want to store many files and load many links, then I need a lot of memory.

1 Like