Dear OpenAI & Community,
I’ve been working with the ChatGPT Plugins extensively and have come to appreciate their functionality and versatility. There’s one area of the implementation that I believe could benefit from a bit of improvement - the use of Conversation IDs.
As you may know, a Conversation ID is provided when making API requests from ChatGPT to plugins. This ID is unique and identifies the current conversation context, which is crucial for maintaining coherent and context-aware interactions. However, this Conversation ID is distinct from the ID used for the chat history. Having two separate identifiers can create confusion and necessitate additional mapping work when tracking the flow of conversation or debugging issues.
Therefore, I propose unifying these identifiers - specifically, making the Conversation ID provided in the API requests from ChatGPT to the plugin the same ID used for the chat history.
Doing so would provide the following benefits:
- Simplicity: It would streamline the process of tracking conversations, as we would only have to keep track of a single identifier.
- Debugging: It would simplify troubleshooting, making it easier to link interactions with their corresponding history.
- Consistency: It would create a more consistent experience, which is always a plus in developer experience.
Of course, I’m aware that there might be complexities or implications that I’m not seeing. I’m eager to hear your thoughts and discuss potential obstacles or alternatives to this approach.
Thank you for taking the time to consider this suggestion.
Brian 