Py-FSRS has a concept of "learning_steps", but why doesn't the library fsrs-rs-dart have it?

[temeipu-offical]

Let me outline my requirements. I am looking to develop an application that supports client-side card generation, calculates the next review time, and performs offline optimization while also supporting server optimization. It is required that the client and server optimizer codes be consistent. The fsr-rs-dart package is only suitable for client-side operation, but the server intends to use the Py-FSRS package. However, it has been discovered that the versions and usages of the FSRS in these two packages seem to be inconsistent. Could you please suggest a good solution?

[L-M-Sherlock]

There are two different series of fsrs lib: fsrs-rs-[lang] and [lang]-fsrs. The former is a series of binding of fsrs-rs, which is used by Anki. Anki implements its own scheduling logic including "learning_steps", so fsrs-rs doesn't have a concept of "learning_steps".

[temeipu-offical]

Which version of FSRS is used by fsrs-rs-[lang] and [lang]-fsrs respectively?

[L-M-Sherlock]

You need to check the specific repo for the version. Some implementations have been updated to FSRS-6, but some haven't.