Re: [PATCH net v2 10/15] drivers: net: cirrus: mac89x0: Remove this driver
Finn Thain <[email protected]>
| Newsgroups | org.kernel.vger.linux-m68k,org.kernel.vger.linux-doc,org.kernel.vger.linux-kernel,org.kernel.vger.netdev |
|---|---|
| Message-ID | <[email protected]> |
On Thu, 23 Apr 2026, Jakub Kicinski wrote: > We want to be welcoming in the community. But being a part of the > community means participating in _work_. > The work of testing drivers requires that maintainers have ready access to actual hardware. Behavioural simulation like QEMU does not cut it. Therefore, it sounds like you're going to need someone with actual clout to mandate that drivers must have maintainers and those maintainers must have access to working hardware, or those drivers get placed on a removal schedule along with _every_ other orphaned module across the entire tree. That way, you get to chase _all_ freeloaders regardless of whether they are unpaid enthusiasts or foundation sponsors. Surely you'd find out what "welcoming" really means -- but you would get to remove a _lot_ of code. Far more likely, you'd find out that the typical Linux user is actually a retro computing enthusiast because he's using some device that's over 5 years old, and the employer of the maintainer of the relevant driver got assigned to something more profitable years ago and all their old hardware got tossed out in the last office move. (What was that saying about those who live in silica houses?) Anyway, I think there is a better way. Let's score every module based on cost/benefit to the "community". We can use retail sales volume as a proxy for the former quantity (in unadjusted dollars) and use module size (in lines of code) as proxy for the latter. If nothing else, that would show us which modules are more or less valuable to the community, relative to the average module.