I recently came across a torrent that seems to be an archive of Reddit. It got me thinking if it would be possible to make it locally browsable. However, I also considered the possibility that someone might have already addressed this by creating a public Lemmy instance, enabling the content to be accessible from any federated instance.
I can actually see some merit to a lemmy API accessible reddit corpus. it would be interesting to reference old reddit info in a lemmy compatible way with zero reference to reddit itself.
doing so for the entire corpus properly (link fixups, etc) would be… challenging, but doable.
Letting the LLMs source from their for free, completely invalidating the proposed licensing model at Reddit.
lemmit.online ?
They only copy posts afaict. I subscribed to some of their communities but unsubscribed rather quickly as the post feed without comments usually has little value.
I’m at a loss as to why anybody wouldnwant this in the first place.
Am I really going to buy a 2TB drive to hold all of reddit…
Actually, I’m pretty surprised that it’s only 2TB.
Edit: and it looks like it’s only captured data up until about six months ago.
It would be helpful if there were an instance that migrated all of this to Lemmy so that we could access it from any other instance, instead of having to download it for local browsing.
I haven’t downloaded it. Looks like a collection of compressed files, but I don’t know exactly what’s inside of them. Do you know what format they’re in?
