This is a more general and structured plan of what features a server has to have, as simple explained as possible.
Do you agree? @upm-developers/owners
This already contains the ideas of @NoahPeeters , just to catch it in a nutshell, that we have a reference.
However:
There is no ID for a server, because the server knows where it gets its messages from and thus it knows if it can trust them (i.e. whether it came from a server he syncs with).
The server does not return an url to download the package, it justs sends the package itself. This is easier. The packages should be on the server, not somewhere else.
I dont call the passwords for packages IDs, because ID means identification, which is not the purpose of the password of a package.
I did not go too deep in the implementation because thats something we discuss when the general idea is stable!
Also added new idea: the maintainer can check packages, too
This is a more general and structured plan of what features a server has to have, as simple explained as possible. Do you agree? @upm-developers/owners This already contains the ideas of @NoahPeeters , just to catch it in a nutshell, that we have a reference. However:
There is no ID for a server, because the server knows where it gets its messages from and thus it knows if it can trust them (i.e. whether it came from a server he syncs with).
The server does not return an url to download the package, it justs sends the package itself. This is easier. The packages should be on the server, not somewhere else.
I dont call the passwords for packages IDs, because ID means identification, which is not the purpose of the password of a package.
I did not go too deep in the implementation because thats something we discuss when the general idea is stable!
Also added new idea: the maintainer can check packages, too
Happy discussing.