estimate_token_count units bug

[Zaczero]

What version of Codex is running?

0.0.0

What subscription do you have?

N/A

Which model were you using?

No response

What platform is your computer?

No response

What terminal emulator and version are you using (if applicable)?

No response

What issue are you seeing?

estimate_token_count uses estimate_reasoning_length without converting it to token count via approx_tokens_from_byte_count, causing ~4x overcount.

codex/codex-rs/core/src/context_manager/history.rs

Line 100 in faeb08c

} => estimate_reasoning_length(content.len()) as i64,

What steps can reproduce the bug?

when getting back in history, at 25% context usage, the compaction starts as if 100% context was used

What is the expected behavior?

No response

Additional information

i think your style of naming is error prone and could be better

or you could make use of strong aliasing

https://doc.rust-lang.org/reference/items/type-aliases.html

[etraut-openai]

Thanks for the bug report and analysis. This will be addressed in the next release.