I am working on a point to point media tool right now and it’s super challenging to get right. There is no server. There are just peers.
Client/server is very easy, because there is a centralized common point of access that manages everything. In a serverless model you have to connect via identity that has no fixed address.
Why is a mixed model not used? Where the server exists and serves both the meta-data and acts as a seeder with support from other users? This seems to be the best of both worlds?
Comments
I am working on a point to point media tool right now and it’s super challenging to get right. There is no server. There are just peers.
Client/server is very easy, because there is a centralized common point of access that manages everything. In a serverless model you have to connect via identity that has no fixed address.
Why is a mixed model not used? Where the server exists and serves both the meta-data and acts as a seeder with support from other users? This seems to be the best of both worlds?
Because that requires creation of new original technology. Most developers struggle to deploy REST using a framework that does all the work for them.