Re: [PATCH v2 00/53] MIPS: SGI-IP27: add support for the IP35 (Bedrock hub) machines

[Jakub Kicinski]

On Thu, 22 Oct 2026 20:28:25 +0100 (BST) Maciej W. Rozycki wrote:
> On Wed, 21 Oct 2026, Jakub Kicinski wrote:
> > > The users are the people running these machines, the same population
> > > as for IP27 and IP30, and the ioc3-eth driver in net/ exists for them
> > > today. The series was developed and tested on a Fuel and a Tezro in
> > > regular use, so "nobody is using it with Linux" is true only in the
> > > sense that the support isn't merged yet; that was equally true of IP30
> > > when it went in for 5.5, and of every port before it.  
> > 
> > Not sure if you followed the discussions at the recent maintainer
> > summit but we are now allowed to be more opinionated about code
> > we merge. Merging code just to test that code seems like entertainment
> > for the authors, not useful development.  
> 
>  I thought the event was by invitation only.

LWN posts a great summary.

>  Anyway, this seems to me like discouraging enthusiasts who do their work 
> outside their day job from making contributions.  Their natural target are 
> commercially worthless systems, since working for free on commercially 
> valuable systems would be taking away someone's bread and I think grossly 
> undeserved support for corporations.  Taking this stance effectively would 
> be cutting Linux off from its origins and making it a commercial project.

Or. Perhaps. By pushing this useless retro-hobbist hogwash you're
contributing to the burn out of genuine maintainers, pushing the
project into the hands of people who do have anterior motives.
Who is to say..

>  Please let's not go down this route and only base the acceptance, or not, 
> of submissions on quality merits.  Many of us took this route to Linux, by 
> showing competence in public with work done for free, and some of us still 
> enjoy such side project even if we do have a day job now that involves 
> Linux development.  Let's not close this option to people.

Who cares about quality of code nobody will run?