Right now it's very difficult to keep the development of all of the different applications and backends in sync, because there's not an explicit version for the protocol, but previously, there have been things called "v1", "v2", "v3", etc.
To avoid confusion, we should use semver with git tags to version the protocol, with version 0.x.y incrementing x whenever there is a breaking change and y whenever there is a non-breaking change. This version number can be incremented as much as we want, but it gives a reference point to keep things in sync.
The version 0.4.0 is chosen to be bigger than all of the informal version numbers I've seen in discussions, so that there is no possibility of confusion.
Right now it's very difficult to keep the development of all of the different applications and backends in sync, because there's not an explicit version for the protocol, but previously, there have been things called "v1", "v2", "v3", etc.
To avoid confusion, we should use semver with git tags to version the protocol, with version
0.x.y
incrementingx
whenever there is a breaking change andy
whenever there is a non-breaking change. This version number can be incremented as much as we want, but it gives a reference point to keep things in sync.The version 0.4.0 is chosen to be bigger than all of the informal version numbers I've seen in discussions, so that there is no possibility of confusion.