I see this as a big issue for content going forward. I get that reddit is one single point of failure but I wonder how you can keep knowledge in a static state when most instances lose money and it’s a hobby. What happens when the owner grows tired.
Could also be an issue from a data privacy perspective. If I want to delete comments I left, but the instance I made them on has shut down how can I delete them?
This is so stupidly relevant to strange motherboard problems. I hope that lemmy is able to be scraped by web crawlers to help assist with these sorts of problems in the future.
That part won’t change. If your current instance was federated with an obscure community/instance before that community/instance disappeared, then you will still have the content from back then and will be able to find the discussion and solution
Ok, but in that instance, unless that niche instance that disappeared never federated with anyone, then their content is going to be available on the instances they did federate with before they went away, and those will continue to show in google
That still works with replication. What won’t work is having any discussion going forward be replicated over all replications. But the replication works fine for archival purpouses.
Perhaps each client should keep data. Something like a blockchain, with redundant copies stored on clients and clients using some consensus protocol to agree on what the real history is.
I see this as a big issue for content going forward. I get that reddit is one single point of failure but I wonder how you can keep knowledge in a static state when most instances lose money and it’s a hobby. What happens when the owner grows tired.
Could also be an issue from a data privacy perspective. If I want to delete comments I left, but the instance I made them on has shut down how can I delete them?
How often do you revisit forum posts you made in 2007? Some content doesn’t need to be preserved
It’s not about visiting a forum post I made in 2007. It’s about visiting a forum post about an obscure issue that someone solved in 2007.
relevant XKCD
This is so stupidly relevant to strange motherboard problems. I hope that lemmy is able to be scraped by web crawlers to help assist with these sorts of problems in the future.
That part won’t change. If your current instance was federated with an obscure community/instance before that community/instance disappeared, then you will still have the content from back then and will be able to find the discussion and solution
Even if my instance never interacted with that instance?
Well no, but then you weren’t going to find it anyway, even if the other instance was still around
I would argue the instance still gets indexed by Google and if it was around I could Google search for the result.
Ok, but in that instance, unless that niche instance that disappeared never federated with anyone, then their content is going to be available on the instances they did federate with before they went away, and those will continue to show in google
It would certainly be nice to replace stack overflow with a good Lemmy instance, but have the data guaranteed to remain around.
That still works with replication. What won’t work is having any discussion going forward be replicated over all replications. But the replication works fine for archival purpouses.
Perhaps each client should keep data. Something like a blockchain, with redundant copies stored on clients and clients using some consensus protocol to agree on what the real history is.
Blockchain is the worst possible solution for this. There are much better versions of this.
Also, no need for clients to keep gigabytes or even terrabytes of data. That’s what you have instances for.