karthink

@mocknen Thank you for the reminder, I missed this when I added o1-preview support.. It required more changes in gptel than provided by this PR, but it should all be...

See some prior discussion about this in #315. Could you suggest 1. When this token counting should happen, keeping in mind that gptel can be used from anywhere? 2. How...

> Perhaps if the provider (for example anthropic) has a token counting end point, `gptel-inspect-query` could have an option to `gptel-query-cost` that queries the endpoint and just displays the result...

@axelknock thanks for digging through the documentation! I've made a note. > If the variation is not that high, for a v0.1 implementation we could simply use the Anthropic count_tokens...

Well there's no easy heuristic way to solve this. I can provide an option where you can populate major-mode -> mode string associations (e.g. `((enh-ruby-mode . "ruby"))`) for use by...

N.B.: I just added `gptel-rewrite-directives-hook`, to which you can add functions to get specialized refactor instructions based on context. I'm still debating adding a `gptel-mode-replacement-alist` with some common transformations like...

> I think even a quick and dirty hotfix that fixes it for 70-80% users is better > than the one where an average gptel user would need to figure...

For this topic, yes, I plan to add a `gptel-mode-name-remap-alist` populated with some common replacements so this should be fixed soon.

Yes, see my [first comment](https://github.com/karthink/gptel/issues/346#issuecomment-2259075893) above. These variables (like `major-mode-remap-alist`) are very bare, it doesn't help.

Thank you for recording these here. I'll add a replacement-alist soon. > Another potential replacement > https://github.com/brotzeit/rustic popular major-mode for Rust development, `rustic-mode` is probably better to be replaced with...