Open juliangruber opened 11 years ago
I think replicating files (maybe not 'huge', but small and medium definitely) into a browser is a viable use case
On Tue, Sep 17, 2013 at 5:27 PM, Julian Gruber notifications@github.comwrote:
Does it make sense to make backer environment agnostic, so it can work in mostly any environment javascript works in? Initially this would just be a node.js thing.
So, would you want to replicate a browser filesystem to your pc using backer? Or is it fair to say that you wouldn't replicate huge data in that setup anyways, so a much simpler solution can be used?
— Reply to this email directly or view it on GitHubhttps://github.com/juliangruber/backer/issues/4 .
oh, yeah, then it would just work with any webapp that primarily uses one of the browser dbs!
This would limit us to use technologies like WebRTC. What about building some kind of bridge?
I thought about relays: peers that can't decrypt the data, but cache and forward it. Those relays could offer a WebRTC/WS bridge. Web services could offer their own relays and you could add them to your trusted ones and they will be integrated in your private network. Those relays can also be super useful for speeding up file transfers
Since it's just a streaming interface, it can use WebRTC, but doesn't have to. When you're running it on your computer and have a web app that uses it, it can for example pipe its stream over websockets to the node server you're running which then pipes to the backer instance running on your server - or any other computer.
re relays, see https://github.com/juliangruber/backer/issues/6
Does it make sense to make backer environment agnostic, so it can work in mostly any environment javascript works in? Initially this would just be a node.js thing.
So, would you want to replicate a browser filesystem to your pc using backer? Or is it fair to say that you wouldn't replicate huge data in that setup anyways, so a much simpler solution can be used?