Mention golang-go in the README.md

[donquixote]

When running the ./scripts/dev as in the "Development setup" in the README.md, I get this:

 => ERROR [bbb-webrtc-recorder builder 4/7] RUN go mod tidy

> [bbb-webrtc-recorder builder 4/7] RUN go mod tidy:
2.452 go: go.mod file not found in current directory or any parent directory; see 'go help modules'

It seems we need to install go / golang-go, and possibly some additional stuff, to make this script work. This should be mentioned in the requirements.

[donquixote]

Also the ./scripts/dev script could verify requirements before it runs.

And I noticed that after a failed run it is better to remove everything and start over. Perhaps it could verify that there are no leftovers from a previous attempt (by looking for a specific file).

[alangecker]

It seems we need to install go

the command go mod tidy is ran inside the docker container. go is not required as an dependency on your system, just inside the container and it is present there and generating the error message, what's missing is the go.mod file (see go.mod file not found in current directory).

For the development setup you need all submodules checked out including bbb-webrtc-recorder. is it possible that you did the git clone without the --recurse-submodules as proposed in the README and the repos/bbb-webrtc-recorder is therefore empty? 🙂

if so following might help:

git submodule init  
git submodule update

[donquixote]

the command go mod tidy is ran inside the docker container. go is not required as an dependency on your system, just inside the container and it is present there and generating the error message, what's missing is the go.mod file (see go.mod file not found in current directory).

Ok this makes sense, my bad.

[donquixote]

is it possible that you did the git clone without the --recurse-submodules as proposed in the README and the repos/bbb-webrtc-recorder is therefore empty?

I did use --recurse-submodules but I think something went wrong in that first run.

I did work (sort of) after installing go and doing a fresh clone. So probably that one step (install go) was unnecessary.

No I am facing the next obstacles:

• #354 The script takes a long time to build everything, and then starts the containers. The only way I found to stop was ctrl+c. After that it seems I need to start all over again to restart. But I might be wrong.
• #351 It is not even clear to me that a successful run of the dev script would bring me any closer to having a docker (compose) setup that could be distributed with another product. I guess that's what that other issue is for.