magnetico
magnetico copied to clipboard
Peer / seeder counts?
I noticed other dht indexers have this, is this available here? When there are 20 torrents of the same thing, I need to know which one is most seeded.
We need to implement BEP 33 for that. I will definitely look into it, thanks for your feedback!
Much, much needed :D
Good idea
DHT scrapes are too expensive, tracker scrapes should be preferred for bulk scraping.
DHT scrapes are too expensive, tracker scrapes should be preferred for bulk scraping.
You do know that magnetico's goal is to function without trackers?
Using the internet is also too slow, sending hard drives per express courier should be preferred for fast data collection.
Hard drives are too slow, would rather send solid state drives.
In all seriousness, BEP33 could allow this data to be collected with existing queries, and responses stored in the DB.
@boramalper it looks like there's columns in the database, and constants in the API to support seeder and leecher counts and querying, but I don't see any code actually using it anywhere. It appears that querying on these causes a panic, and that none of the DHT exploration code fills in the database values. Could you confirm that's the case? Are there any PRs/existing attempts to do this, or was it scoped out in the boilerplate, but nobody got around to it? Cheers
I've implemented BEP33 upstream, and started trying to integrate it with magneticod
. As far as I can tell, magneticod
inserts complete torrents into the database only when it has the metadata, it has no provision for updating or merging peer/seed counts over time. It looks like some fields where intended for this like updated_on
, but were never wired up. I assume it would require some redesign or an additional service within magneticod
to continuously maintain an up-to-date value for n_seeders
and n_leechers
.
It's not pretty, but it works as a proof of concept: https://github.com/boramalper/magnetico/compare/master...anacrolix:popularity?expand=1.
Any updates on seed/peer counter :)
I built a library that does this a few years ago that I'm using in torrents-csv: https://github.com/dessalines/torrent-tracker-health