Re: Q: our priority: our users, or free software?

Hi,

On 3/25/26 5:10 AM, Lucas Nussbaum wrote:

> This is very nice when the interests of our users and of the free
> software community are aligned. But sometimes they aren't. And we
> probably have at least some users who care a lot more about how well
> Debian works for them, than about free software.

I'd argue that long-term, the interests of our users and of the free 
software community are always aligned because only free software can 
ensure that our users' needs will also be met in the future.

The various nVidia graphics cards I have collected over the years have 
slowly fallen out of support by the non-free drivers, but they are still 
perfectly good cards -- so as a user, I rely on the free software 
community after vendor support has run out.

Pretty much any argument that Debian needs to ship the proprietary 
drivers because "users are a priority" has focused only on the short 
term, where we have full vendor support. This works only as long as 
there are some greybearded free software zealots that reject the most 
convenient option, and instead implement something that can be supported 
by the community and is not beholden to any single external entity whose 
goals align with the interests of neither our users nor the free 
software community.

That is another balance we must strike in the project, so this is kind 
of an extension to the question: should Debian invest resources 
(manpower, money, publicity, ...) into things that do not provide a 
short term benefit, but ensure the long term viability of the free 
software ecosystem, or is that outside our wheelhouse?

For example, are we content with the "user-friendly" approach of 
Microsoft signing our secure boot shim, or do we want to join a lobbying 
effort to require user manageable SB keys in ARM64 firmware?

> Sometimes non-free tools, or infrastructure clearly based
> on non-free software, can help us make Debian better for our users.

That is a view that assumes that users are consumers. Some of them are, 
but for the longevity of the free software ecosystem, we need to turn 
some of these users into contributors, and an important aspect of that 
is to give them access to tooling for experimentation.

Pretty much all of the older developers started out by installing a 
compiler and experimenting without a clear direction.

Any non-free tools or infrastructure we have, is however available only 
for the particular purpose of improving something in Debian, and we 
cannot make it available to users, so we really need to be careful to 
not accidentally create an additional barrier to entry here.

If a tool is useful, a free software version should be made. Alleviating 
the immediate need for such a tool makes sure that the free software 
version will be made only if enough greybearded free software zealots 
come together to build it.

It would be nice if we could have a free software community that is not 
entirely reliant on the greybeards. A bit of outreach and diversity 
might help here -- my feeling is that the western hemisphere nerds have 
become a bit too complacent lately in assuming the good will of 
corporations ensuring their continued access to generic computing.

    Simon