You are viewing a single comment's thread from:

RE:

in Threespeak - OLD5 years ago (edited)

Application devs/engineers NEED to know more about decentralized solutions or else they'll reach for the centralization solutions they've been trained to reach for instead... We already know why we can't have good things lol... it's simply b/c we keep reaching for the centralization hammer for every problem.

I'd suggest if you wanted faster progress than learning and building a DHT yourself, then you could instead figure out how to leverage existing DHT solutions that are mature and without UX issues in their DHT design.

Webtorrent compatible with IPFS is a great example. There's already an open request!
https://github.com/ipfs/js-ipfs/issues/779

EDIT: Good on the experimentation though! Love to see that. Deep down, I know you devs want to build something and improve upon the existing tech :p

Sort:  
 5 years ago (edited) 

Hm, do you truly believe there is a way to achieve decentralisation? I ask because I hope you put some thought into it before asking, God willing, so that you can recognise what I'm talking about! If not, I hope this inspires you, God willing.

By it's very nature, there's almost too many ways to describe how we can achieve decentralisation lol! God bless :)

The main question is, why can't we use YouTube, WebTorrent, Amason S3, etc as "pinning" services? Ultimately, we're looking for idempotency and the ability to fetch and verify data p2p. If we can achieve this with the various "centralised" solutions, we can basically bootstrap decentralisation and 3rd parties will have to play whack-a-mole to censor our information. IPFS can be a way to merge this strategy (which already technically exists) and wrap it with native p2p technologies that IPFS/Libp2p are blessed to have focused on.

Ironic how decentralisation takes advantage of the various methods of centralisation!

So there are 2 ways I know to achieve using these 3rd party backends. Both are pretty similar. The first provides information WHERE to find a specific hash and how to download it, all the information one needs to either fetch using the DHT or fetch via 3rd party (and then verify). The second instead hides the fact we're fetching via a 3rd party so our peers can fetch data as if it were in the DHT raw.

That's about all I'm going to elaborate on unless you're hella excited and motivated and inspired lol. There's much to do on IPFS towards decentralisation! Haters gonna hate but I obviously don't think you are a hater... But the CIA is all about wasting our time spinning our wheels going nowhere lol so I have to be wise with my time, God willing. Peace!