Misrepresentation of RateLimitError as 'Unknown OpenAI Error'

[h-b]

I'm currently using Logseq 0.9.9 under Fedora and the GPT-3 OpenAI plugin v1.13.1. Even after generating a new API key, I'm still getting the 'Unknown OpenAI Error', similar as described in an older issue #55 (comment) - in my case, even /dalle causes the same error and there are three different error messages, depending on what specific function of the plugin I use.

I've tried to follow the debug instructions on https://github.com/briansunter/logseq-plugin-gpt3-openai#debugging, but I'm having trouble finding the 'View -> Toggle Developer Tools' menu, even though I've enabled developer mode in Logseq's advanced settings.

I've been encountering this issue ever since I first installed the plugin, rendering it unusable for me. Despite several updates to both Logseq and the plugin itself since the initial installation, the problem persists.

I use the default settings:

[h-b]

I've managed to open the developer tools. It turns out, to open the view shown at https://github.com/briansunter/logseq-plugin-gpt3-openai#debugging, one needs to use the key combination Ctrl+Shift+I, as Logseq is an Electron app. Here are the error messages that appear in the Console when I use this plugin:

[h-b]

I've managed to diagnose the problem by running a quick test using the OpenAI quickstart Python repository (https://github.com/openai/openai-quickstart-python.git). The issue appears to be an openai.error.RateLimitError. It would have been helpful if the plugin's error message had indicated this directly. To reflect these findings, I've updated the title of this issue accordingly.