Error: Please install openai as a dependency with, e.g. `yarn add openai`

[wadewegner]

This looks like a promising project.

I was attempting to run the example in the readme. When started, I get the following error:

node index.js                                         
Telemetry: Attempt 1 failed with status: 415
Telemetry: Attempt 2 failed with status: 415
/embedchainjs/node_modules/chromadb/dist/cjs/chromadb.cjs:2986
      throw new Error(
            ^

Error: Please install openai as a dependency with, e.g. `yarn add openai`

I have installed the npm dependencies, entered the OpenAI token, and I have Chroma running in Docker.

Any tips?

[wadewegner]

Hmm ... I see this is coming from sendTelemetryEvent. What's the purpose of this API call to https://api.embedchain.ai/api/v1/telemetry/?

[wadewegner]

The fact that this is making API calls to store usage information is extremely disappointing. Nowhere in the README or elsewhere do I see anything about this, and the fact that it appears to be the default behavior is even more disappointing.

You apparantly have a way set this to not collect Telemetry (I see something called collectMetrics) but it doesn't appear to be a default behavior nor do I see how to use it.

[taranjeet]

Hey @wadewegner welcome to embedchain and thanks for trying it out. We are not maintaining the JS package actively, thats why the error. We have been doing all the development on python version now, and this repo was created in the starting. I will figure out a way to communicate this upfront in the readme and archive it as well.

Also for the analytics, this is very old code which we started developing, but then it got dropped due to us solely focussing on the python version. We had plans to update the readme, but skipped as we were not developing the package further. In the python version, we highlight it in the readme and are working on improving further.

We thought a lot about capturing usage information or not, but then realized that without a high level overview of the usage, it becomes very difficult for us to prioritize our features. We are very transparent, open to feedback and always ready to support our community. If you have any suggestions, please feel free to share.