openwebtorrent-tracker VS libtorrent

openwebtorrent-tracker (C++, single-threaded) reached 9 thousand responses a second

I think you can find some tracker here: https://openwebtorrent.com/

I did some benchmarks with my load testing tool and got very encouraging results: 1.8M tracker responses per second when using 14 threads, compared to a maximum of 117k responses per second for closest competitor wt-tracker. I suspect that the openwebtorrent tracker would be a lot closer in performance to aquatic, but since it only runs over TLS, I couldn’t test it, as the load tester doesn’t support it. More info on the benchmark (PDF).

I had a longer answer, but qBittorrent is great and has a great webui/server model if you're running a separate linux box to run it in. If you're on windows, grab 4.3.9 or 4.6.2 with libtorrent 1.x. RSS stuff is nice too. Libtorrent 2.x has issues on Windows currently (read these threads if you want to know more - https://github.com/arvidn/libtorrent/pull/7013 - https://github.com/arvidn/libtorrent/issues/6667#issuecomment-1804035311 ). Please don't bug Arvid about it though. Libtorrent 2.x gets us Bittorrent v2, which is a good thing.

Latest libtorrent v1.2 branch changelog: https://github.com/arvidn/libtorrent/blob/RC_2_0/ChangeLog

You can support this feature by writing a comment in https://github.com/arvidn/libtorrent/issues/7252

Where users could finally download files that were present in other swarms. This would decentralize network more, swarms with valuable files wouldn't die fast, even when updating a torrent, where info hash differs, say tv series, adding new episode, all the previous seeds of swarm could contribute bandwidth and resistance to new swarms, there were nice proposals at libtorrent project (library which many clients we know use), as f.e. not announcing that a peer has files to not overwhelm network, but announcing that he wants them.

How much of it do you want to do yourself vs how much do you want to use pre-made libraries for? Many popular clients rely on libtorrent which is a C++ torrent library: https://github.com/arvidn/libtorrent

FWIW, it seems like there'll be an update to libtorrent very soon that'll address the memory issue.

Here is a link to an issue I found on the libtorrent repo discussing the issue.

Many!

https://webtorrent.io/faq

Notable ones include Libtorrent[1], Peertube[2], and their own Webtorrent Desktop[3].

1. https://github.com/arvidn/libtorrent/issues/223

2. https://peertube-viewer.com/posts/2021-02-20-peertube-viewer...

3. https://webtorrent.io/desktop/