We basically always goes through all microservices when generating client code, and if one of those services isn't built or failed in a previous build, then this is going to crash.
We should:
Give a better error information about what is going on (maybe say go there and build your services again because something is up with them)
~Have a separate command for building all microservices and/or a flag for the generate-client to do a clean build before generating all clients code (would be nice to test timing consumption on this)~ (this already exists as beam project build)
We should attempt to build all services, and if one fails in the middle, aggregate its error into a final exception at the end of all service builds.
Discussion in slack about this: https://disruptorbeam.slack.com/archives/C03P1J5NR97/p1717595501347599
We basically always goes through all microservices when generating client code, and if one of those services isn't built or failed in a previous build, then this is going to crash.
We should:
generate-client
to do a clean build before generating all clients code (would be nice to test timing consumption on this)~ (this already exists asbeam project build
)We should attempt to build all services, and if one fails in the middle, aggregate its error into a final exception at the end of all service builds.