# Testing low-level layer to harden astropy's core stability (GSOC 2026)

**URL:** <https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378>\
**Category:** Astropy\
**Tags:** astropy, question, gsoc\
**Created:** [March 12, 2026, 7:08pm UTC](https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378 "2026-03-12T19:08:23Z")\
**Posts on this page:** 4\
**Page:** 1

<div class="post-metadata">

**Author:** ![Abu-bakkar-siddique](https://dub1.discourse-cdn.com/flex005/user_avatar/community.openastronomy.org/abu-bakkar-siddique/32/744_2.png) [@Abu-bakkar-siddique](https://community.openastronomy.org/u/Abu-bakkar-siddique)\
**Post date:** [March 12, 2026, 7:08pm UTC](https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378/1 "2026-03-12T19:08:23Z")

</div>

Hi all,

I need some clarification on the scope and scale of the proposed project [hardening astropy’s core stability](https://openastronomy.org/gsoc/gsoc2026/#/projects?project=hardening_astropy's_core_stability) about testing low-level layer extensions.

> Does _ **building a more fundamental test suite for this low-level layer** _ mean testing the functions exposed by the compiled `.so` modules, OR does it mean more in-depth testing of internal helper functions in `.c` files as well?

TIA

---

<div class="post-metadata">

**Author:** ![neutrinoceros](https://dub1.discourse-cdn.com/flex005/user_avatar/community.openastronomy.org/neutrinoceros/32/589_2.png) [@neutrinoceros](https://community.openastronomy.org/u/neutrinoceros)\
**Post date:** [March 13, 2026, 5:12pm UTC](https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378/2 "2026-03-13T17:12:39Z")

</div>

Hi,

It could be both. Though I think right now astropy as _zero_ C testing going on and as far as I understand that would require an entire new layer of architecture and these helper functions \*might\* proove too volatile to be worth the effort, so we’d need to first justify that it’s worth doing, and I wouldn’t advise attempting it now.

---

<div class="post-metadata">

**Author:** ![Abu-bakkar-siddique](https://dub1.discourse-cdn.com/flex005/user_avatar/community.openastronomy.org/abu-bakkar-siddique/32/744_2.png) [@Abu-bakkar-siddique](https://community.openastronomy.org/u/Abu-bakkar-siddique)\
**Post date:** [March 13, 2026, 5:44pm UTC](https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378/3 "2026-03-13T17:44:05Z")

</div>

So would it be reasonable to assume (for now) that if python-callable extension functions have sufficient test coverage, then the helper functions are indirectly validated through those tests?

---

<div class="post-metadata">

**Author:** ![neutrinoceros](https://dub1.discourse-cdn.com/flex005/user_avatar/community.openastronomy.org/neutrinoceros/32/589_2.png) [@neutrinoceros](https://community.openastronomy.org/u/neutrinoceros)\
**Post date:** [March 15, 2026, 11:40am UTC](https://community.openastronomy.org/t/testing-low-level-layer-to-harden-astropys-core-stability-gsoc-2026/1378/4 "2026-03-15T11:40:53Z")

</div>

Sorry for the delay in my answer. I’m not checking on this forum daily. The answer is yes