Wouldn’t an archiving service be improved by utilizing a federated, decentralized model? A large decentralized network of archiving servers that can not only A) act as redundant backup for archived sites but also B) act as different points of entry for the archiving process.
The whole reason you’re attempting to archive things like this is for both freedom of access and also preservation of information, right? Having that kind of service in a centralized form seems like a huge problem.
Not many people can mirror what is probably a couple of petabytes.
And ensuring preservation and availability across unreliable nodes is pretty difficult, requires considerable redundancy. How many people are willing to do that especially when it’s legally difficult?
edit: and how many people would keep their mirrors up when FBI/Europol start cracking down on people hosting mirrors.
edit: There seems to have been an effort to decentralize internetarchive on dweb. But I’m not sure how serious dweb is, it seems to have been a vehicle to promote crypto (the blockchain kind) at least in part.
I also found out that KDE (the linux desktop) is funded by big meat (Tönnies), through Blue Systems. They sponsor dweb too. Just random aside.
Essentially the same built-in problems that crypto has. The model works for blogging, like twitter and reddit clones. It’s just not the right set up for frequent requests or a huge amount of storage.
That being said, I hope a few of the mirrors are exploiting the current geo-political climate. Canada may eventually fold, but a server in Russia or China ought to be well enough removed.
The issue with that is you need volunteers with the infrastructure to run the nodes, and each one would be subject to as much legal liability and the person running a centralised service
Wouldn’t an archiving service be improved by utilizing a federated, decentralized model? A large decentralized network of archiving servers that can not only A) act as redundant backup for archived sites but also B) act as different points of entry for the archiving process.
The whole reason you’re attempting to archive things like this is for both freedom of access and also preservation of information, right? Having that kind of service in a centralized form seems like a huge problem.
Not many people can mirror what is probably a couple of petabytes.
And ensuring preservation and availability across unreliable nodes is pretty difficult, requires considerable redundancy. How many people are willing to do that especially when it’s legally difficult?
edit: and how many people would keep their mirrors up when FBI/Europol start cracking down on people hosting mirrors.
edit: There seems to have been an effort to decentralize internetarchive on dweb. But I’m not sure how serious dweb is, it seems to have been a vehicle to promote crypto (the blockchain kind) at least in part.
I also found out that KDE (the linux desktop) is funded by big meat (Tönnies), through Blue Systems. They sponsor dweb too. Just random aside.
Essentially the same built-in problems that crypto has. The model works for blogging, like twitter and reddit clones. It’s just not the right set up for frequent requests or a huge amount of storage.
That being said, I hope a few of the mirrors are exploiting the current geo-political climate. Canada may eventually fold, but a server in Russia or China ought to be well enough removed.
The issue with that is you need volunteers with the infrastructure to run the nodes, and each one would be subject to as much legal liability and the person running a centralised service