Yeah, I concur with the idea. I think that's also the set up for other templates as well. As for me, I went with YAML since I think there should be minimal markup to the data since it's a CV and nothing fancy like a webpage. I don't see myself using images or other custom styles/functions. Even if I did, I think the `eval()` would be enough for minimally styled text and/or the occasional italics or bold. I'm also looking into piping the data in YAML to LLMs to help suggest better phrasing of descriptions of various items and take advantage of better context and structure (function calling?).
Edit: Also, there was the benefit of JSON schema for validation, I guess.
Comments
Yeah, I concur with the idea. I think that's also the set up for other templates as well. As for me, I went with YAML since I think there should be minimal markup to the data since it's a CV and nothing fancy like a webpage. I don't see myself using images or other custom styles/functions. Even if I did, I think the `eval()` would be enough for minimally styled text and/or the occasional italics or bold. I'm also looking into piping the data in YAML to LLMs to help suggest better phrasing of descriptions of various items and take advantage of better context and structure (function calling?).
Edit: Also, there was the benefit of JSON schema for validation, I guess.