Spent two hours or more trying to update a vendor's long running unholy Linux kernel fork before realizing it was quicker and simpler to take mainline and patch in the vendor files from the unholy fork. Why are vendors like this? [sigh]

@cda In general I think hw vendors have trouble attracting top software talent (sw support being viewed as necessarily evil to help sell silicon, lower salaries, fewer roles, etc), so getting a clean patchset that's up to date, much less navigating the painful path to submitting things to mainline is pretty rare. Management lacks the interest in paying for more than the bare minimum, sw staff lacks sufficient power or motivation to argue for doing things right.

@swetland yeah this shows up in firmware libraries / SDKs where it's optimized for convenience of the vendor adding HW variants or eval kits but basically nothing else.

@cda Oh gods yes... I'm dealing with an MCU vendor right now that has an absurdly convoluted framework for building their little samples that involves piles of intermediate code generated from templates and the first week was just "how do I extract the headers and useful re-useable code from this disaster so I can build it from my build system instead of their mess".

If I never see another bespoke Si vendor IDE it'll be too soon. (using native code, Java, Python, *and* Lua!)

Sign in to participate in the conversation

private mastodon instance