Backport PR #1534 on branch 0.10.x (Prepare v0.10.8) (#1535)

Co-authored-by: Philipp A <flying-sheep@web.de>

docs/release-notes/0.10.8.md

@@ -1,15 +1,13 @@
-### 0.10.8 {small}`the future`
+### 0.10.8 {small}`2024-06-20`
 
 ```{rubric} Bugfix
 ```
 
 * Write out `64bit` indptr when appropriate for {func}`~anndata.experimental.concat_on_disk` {pr}`1493` {user}`ilan-gold`
 * Support for Numpy 2 {pr}`1499` {user}`flying-sheep`
 * Fix {func}`~anndata.experimental.sparse_dataset` docstring test on account of new {mod}`scipy` version {pr}`1514` {user}`ilan-gold`
+
 ```{rubric} Documentation
 ```
 
 * Improved example for {func}`~anndata.experimental.sparse_dataset` {pr}`1468` {user}`ivirshup`
-
-```{rubric} Performance
-```

docs/release-notes/0.10.9.md

@@ -0,0 +1,10 @@
+### 0.10.9 {small}`the future`
+
+```{rubric} Bugfix
+```
+
+```{rubric} Documentation
+```
+
+```{rubric} Performance
+```

docs/release-notes/release-latest.md

@@ -1,5 +1,8 @@
 ## Version 0.10
 
+```{include} /release-notes/0.10.9.md
+```
+
 ```{include} /release-notes/0.10.8.md
 ```
 

src/anndata/_io/zarr.py

@@ -106,8 +106,6 @@ def read_dataset(dataset: zarr.Array):
         return value
     elif isinstance(value.dtype, str):
         pass
-    elif issubclass(value.dtype.type, np.str_):
-        value = value.astype(object)
     elif issubclass(value.dtype.type, np.bytes_):
         value = value.astype(str).astype(object)  # bytestring -> unicode -> str
     elif len(value.dtype.descr) > 1:  # Compound dtype