From 9961b7be53a3cbf741dd6ffa4e0276bfb59ae686 Mon Sep 17 00:00:00 2001
From: "Lorenzo (Mec-iS)" <tunedconsulting@gmail.com>
Date: Sat, 24 Oct 2026 05:33:50 +0100
Subject: [PATCH] docs: surface the 0.7.0 datasets/std_rand breaking change in
 README, crate docs and AGENTS

---
 AGENTS.md  |  2 +-
 README.md  | 20 ++++++++++++++++++--
 src/lib.rs |  5 +++++
 3 files changed, 24 insertions(+), 3 deletions(-)

diff --git a/AGENTS.md b/AGENTS.md
index d8e4a4d5..59dca180 100644
--- a/AGENTS.md
+++ b/AGENTS.md
@@ -47,7 +47,7 @@ Key features defined in `Cargo.toml`:
 
 - `ndarray-bindings` — optional `ndarray` integration.
 - `serde` — serialization support (also pulls in `typetag`).
-- `datasets` — built-in sample datasets; implies `std_rand` and `serde`.
+- `datasets` — built-in sample datasets; implies `serde` only. Since 0.7.0 it does not imply `std_rand` (breaking change, #416). Enable `std_rand` together with `datasets` to seed the generators from OS entropy.
 - `std_rand` — enables standard RNG facilities in `rand`.
 - `js` — for `wasm32-unknown-unknown` in-browser usage.
 
diff --git a/README.md b/README.md
index b219b2f9..4baad5d1 100644
--- a/README.md
+++ b/README.md
@@ -37,7 +37,7 @@ Add to Cargo.toml:
 
 ```toml
 [dependencies]
-smartcore = "^0.6"
+smartcore = "^0.7"
 ```
 
 For the latest changes from the `main` branch:
@@ -55,6 +55,22 @@ Optional features (examples):
 
 Check Cargo.toml for available features and compatibility notes.
 
+## Breaking changes in 0.7.0
+
+> [!WARNING]
+> The `datasets` feature no longer enables `std_rand` (#416).
+>
+> - The generators `make_blobs`, `make_circles` and `make_moons` no longer take their seed from the OS RNG.
+> - Without `std_rand` they use a fixed seed and return the same output on every call.
+> - With `std_rand` they take their seed from OS entropy, as before.
+>
+> To keep the 0.6 behaviour, enable both features:
+>
+> ```toml
+> [dependencies]
+> smartcore = { version = "^0.7", features = ["datasets", "std_rand"] }
+> ```
+
 ## Quick start
 
 Here is a minimal example fitting a KNN classifier from native Rust vectors using DenseMatrix:
@@ -103,7 +119,7 @@ Recent refactors emphasize reusable components in trees/forests and expanded mul
 ## Data access and readers
 
 - CSV readers: Read matrices from CSV with configurable delimiter and header rows, with helpful error messages and testing utilities (including non-IO reader abstractions).
-- Dataset generators: make_blobs, make_circles, make_moons for quick experiments.
+- Dataset generators: make_blobs, make_circles, make_moons for quick experiments. Since 0.7.0 they use a fixed seed unless you also enable `std_rand` (see the breaking changes note above).
 - Built-in datasets (feature-gated): digits, diabetes, breast cancer, boston, with serialization utilities to persist or refresh .xy bundles.
 
 
diff --git a/src/lib.rs b/src/lib.rs
index 99af7502..b1608bdd 100644
--- a/src/lib.rs
+++ b/src/lib.rs
@@ -39,6 +39,11 @@
 //! ```
 //! Check `smartcore`'s `Cargo.toml` for available features.
 //!
+//! **Breaking change in 0.7**: the `datasets` feature no longer enables `std_rand`. The
+//! generators `make_blobs`, `make_circles` and `make_moons` then use a fixed seed and
+//! return the same output on each call. Enable `std_rand` together with `datasets` to
+//! seed the generators from OS entropy, as in 0.6.
+//!
 //! ## Using Jupyter
 //! For quick introduction, Jupyter Notebooks are available [here](https://github.com/smartcorelib/smartcore-jupyter/tree/main/notebooks).
 //! You can set up a local environment to run Rust notebooks using [EVCXR](https://github.com/google/evcxr)