Should we switch from make to meson for a simpler, more maintainable build?

[Savannah.T]

I’m trying to decide if it’s worth moving our internal tool’s build system from Make to something like Meson. The Makefiles have become a maze of cross-platform hacks and recursive includes that only one person on the team really understands, and adding a new dependency feels like I’m breaking something every time. I’m just not sure if the pain of migration and retraining everyone will actually solve our maintainability problems or just trade them for new ones.

[LunaYG]

We tried migrating a chunk of our internal tool to Meson and it did clean up some platform flag spaghetti It also exposed a lot of edge cases that just kept blooming We still rely on a few hacks but we can trace them a bit easier now and onboarding feels less terrifying even if not perfect

[Madison.W]

I kept a small shim that parsed a manifest and produced the build graph We tracked rebuilds across Linux and Windows and saw fewer inconsistent results after pruning a ton of recursive includes The cost was the lingering toolchain quirks and a couple of broken automations that still bite us

[AbigailXM]

The more I think about it the less convinced I am this is the right lever The heart of the trouble might be ownership and tacit knowledge not the tool We went from one map keeper to a team and the map kept mutating It is exhausting to chase someone who knows the hidden shortcuts

[William16]

Do you think the pain is really about the build system or about team ownership?