Describe the issue:

I was trying to "patch" a structed array (contiguous) with a structed patch array (non contiguous). It turns out that if I use a.view(b.dtype)[0]=b, the result is more or less random, while if I use a.view(b.dtype)[:1]=b, then things works

Reproduce the code example:

import numpy as np
a = np.array([False,False],'?,?,?,?')
b = np.array(True,{'names':['f1'],'formats':['?'],'offsets':[1],'itemsize':4})
a.view(b.dtype)[0] = b
print(a)

Error message:

When I try multiple times in REPL, it's cycling between

[( True,  True, False, False) (False, False, False, False)] # Wrong

and

[(False,  True, False, False) (False, False, False, False)] # Correct

Python and NumPy Versions:

2.3.2
3.12.11 (main, Aug 8 2025, 17:06:48) [Clang 20.1.4 ]

Runtime Environment:

[{'numpy_version': '2.3.2',
'python': '3.12.11 (main, Aug 8 2025, 17:06:48) [Clang 20.1.4 ]',
'uname': uname_result(system='Linux', node='z', release='6.17.0-3-generic', version='#3-Ubuntu SMP PREEMPT_DYNAMIC Thu Aug 28 10:21:42 UTC 2025', machine='x86_64')},
{'simd_extensions': {'baseline': ['SSE', 'SSE2', 'SSE3'],
'found': ['SSSE3',
'SSE41',
'POPCNT',
'SSE42',
'AVX',
'F16C',
'FMA3',
'AVX2'],
'not_found': ['AVX512F',
'AVX512CD',
'AVX512_KNL',
'AVX512_KNM',
'AVX512_SKX',
'AVX512_CLX',
'AVX512_CNL',
'AVX512_ICL',
'AVX512_SPR']}},
{'architecture': 'Haswell',
'filepath': '/home/z/.venv/lib/python3.12/site-packages/numpy.libs/libscipy_openblas64_-8fb3d286.so',
'internal_api': 'openblas',
'num_threads': 12,
'prefix': 'libscipy_openblas',
'threading_layer': 'pthreads',
'user_api': 'blas',
'version': '0.3.30'}]

Context for the issue:

Because it's expected not to affect no-element position for sparse strutured array