Contributing guidelines

[odyslam]

Hey @jacekkolasa,

I want to align this repo with the others, in the work that I am doing to remove friction in the contribution process.

Can you please walk me through the contribution process for the dashboard?

• What are the parts that have to do with Netdata Cloud (thus the contributor does not care)
• How should a contributor develop? (simply fork and edit files? Any hints on the structure and/or workflow?
• From the readme I understand that I don't need to do anything on the agent installation, it talks directly with the API so I just edit my fork, it is served by some static server in the npm scripts and it connects to the agent that is running on my computer, right?

[jacekkolasa]

Hey @odyslam,

We need to ask also @novykh and @netdata/cloud-fe , because I'm not sure I can answer your questions (I am not the main maintainer of this repo anymore). I've heart we'll have another refactor soon, so we need to be careful here, to avoid situation when a contributor works on a piece of code that will be soon dead anyway.
We'll be changing the code style and used libraries heavily, so IMO we'll need to provide some instructions - which tools to use, which patterns, etc. That is because users can get wrong impression by guessing the pattern based on current code (for example it's written mostly in Typescript, but after refactor probably it will be removed).

I would say that after the refactor we should create a short list of "Help Wanted" issues.

What are the parts that have to do with Netdata Cloud (thus the contributor does not care)

Most of the code is shared between Agent Dashboard and Netdata Cloud. Generally, Netdata Cloud uses everything that is exported by https://github.com/netdata/dashboard/blob/master/src/index-npm.ts

How should a contributor develop? (simply fork and edit files? Any hints on the structure and/or workflow?

Yes, just fork and they can start editing. I've put the instructions on how to start working on dashboard, it's basically just single line https://github.com/netdata/dashboard/#npm-run-startdashboard if you already have nodejs and Netdata installed.

From the readme I understand that I don't need to do anything on the agent installation, it talks directly with the API so I just edit my fork, it is served by some static server in the npm scripts and it connects to the agent that is running on my computer, right?

Yes, exactly. You can also use external agent, we can add that info to readme if needed. You just need to modify https://github.com/netdata/dashboard/blob/master/src/utils/server-detection.ts#L39 line for example to return "https://frankfurt.my-netdata.io/"

[odyslam]

1. Given the devenv, where everything runs inside a container, using npm run start:dashboard and a local installation of the agent is just fine, I think. Adding an option to use Frankfurt it would make things even better
2. devenv has already nodejs and the entire toolchain, so we are good to the front
3. I think the biggest issue is the one which you start with, namely the structure of the code and the refactor that you will do, the tools, the patterns, etc.

I don't expect huge contributions in the dashboard, but given that we are an OSS-first project, I feel that we at least should document it. I think this is what users would expect and perhaps it will nurture contributions in this front.