Curate an ipfs directory (discoverability)

[rht]

Qmb8tKfz3RFiasE5yoaaU3d4CF7SaCEh8h2z3g6ZoYGvw6 ipfs.io QmQv4YQNmRPuTTHs4AgBhKEFDdN7eQYeTbSmr8JVWVfury sintel.mp4 QmWom8ULsNCbtMgNMk82aJT7G6gU55mSGxVaC2i1fuyX3s pitcairn island QmWWLNQER7Mbx9NumMzgxeDhFpDErZUTU64LhbHmNRtY3b lots of jpg pics QmYv5fL23nQGft73zVV2coWXBvrcWbqGm8pNaoexMPrBBw doublec's blog ...

More can be scraped from #ipfs logs. Though with no context of what the hashes refer to.

The global tree structure of the merkledag nodes can be drawn. However, the links between html files inside the nodes & their contents can't be drawn directly just from the protocol. We need a global (indexer + search-engine) or ("locate"/"find" + "grep") for this.

[nonchip]

i personally don't think you want to do this automatically, because there might be much private (or at least irrellevant to the public) data in ipfs. but maybe a search engine where people can add their hashes and small searchable description field, like e.g. TPB&co did for torrent hashes. explicitely publishing seems better imho than implicitely. also: improve ipns and use that for "public stuff", because it generates much more stable links

[rht]

@nonchip, for encryption / private content, I redirect you to https://github.com/ipfs/notes/issues/63. For distributed indexer / search engine, see https://github.com/ipfs/archives/issues/8.

[the8472]

i personally don't think you want to do this automatically, because there might be much private (or at least irrellevant to the public) data in ipfs.

That's a reason for indexing. Otherwise it will create an expectation of privacy that can be easily violated by some 3rd party that does implement indexing.

[normandmickey]

I recently registered www.ipfs.directory and am looking for people interested in helping create a directory of sites using IPFS. Initially I will adding static links but eventually I would like to automate the process.