Do you guys think IPFS is going to replace bittorrent

1

What can IPFS can do that bittorent (and its implementations, including webtorrent in browser) can or can’t

2

Technically

They are really similar however IPFS has a way faster time to first byte than current torrent implementations.
However current implementations of IPFS have really low throught put.

Or in other words where IPFS can start downloading in less than 250ms with good connection, torrent would take 10s to reach that point, however once downloading torrent can easly do 10+Gig/s IPFS is just not smart enough to download at thoses speeds yet.

If you only read that narrative it’s easy to think that IPFS is better for smaller interactive files (example navgating the web) and torrent for bigger files (like HD video, linux isos).

What really matters

However that forgetting that thoses are implementations details, we could take what make IPFS’s time to first byte fast and doing the same for torrent and we could take what make torrent be able to reach ludicrous speeds and apply it to IPFS (and there are plans to do so).

I still think IPFS will get bigger than BitTorrent for other reasons, mainly IPFS is customisable AF, the network stack (libp2p) is used by ETH2.0 for example, IPLD can be used to represent git repos, files, message histories, …

Torrent is kinda limited to sharing files p2p. IPFS allows dev using to be way more creative.
I don’t think that on the file sharing buisness IPFS is gonna get better than torrent (it’s gonna catchup to his speed, I know I work on that), however you will see IPFS and / or IPFS’s stack being used as building for web3 in the future (like we already see).

3

Yea. I feel like IPLD is basically a virtual filesystem program, it brings me back to the unix philosophy that everything is a file (DAG / file-folder structure) because it’s a representation of datas

4

I wonder if after I write a virtual file system myself, which is just a bunch nodes organized in DAG structure, I can put magnet link (lol) instead of CID inside them and use it like I would use IPLD with IPFS’s CID but now I am using bittorrent in an emulated file system that I can delete / add and access these objects / datas in browser or any where to host dynamic website or to represent a git repository.
Maybe if there is improvement to any of the two technology, we can actually have dynamic websites running in d-internet but definitely ipfs is the hot potato that’s moving faster than bitorrent I feel like or maybe I haven’t been reading about improvement in bittorent