[cfarm-users] cfarm110 and cfarm111 decommissioning on February 27

[Thorsten Glaser]

On Sun, 15 Feb 2026, Jonathan Wakely via cfarm-users wrote:

>Maybe in an ideal world. But maintainers of free software should not be
>obliged to purchase proprietary software licenses to port their code to
>that platform, or go to extraordinary lengths to do so.

But these who want to, don’t hinder them.

(I don’t have a budget for these purchases, but I do want to port my
code to all those platforms.)

>The only obligation should be to accept patches for supporting such
>systems, and give them fair consideration. That means users/vendors of
>those systems can submit patches to the project, and the maintainers won't

Haha, as if. When mksh was added to Android, Google told me to not
even expect patches from them, that, if I want to see them if/when
they patch that I have to look at their git repo myself.

>Then the market should provide patches.

As a maintainer, perhaps I want to be able to change those patches
to suit my style better or to avoid breaking other systems, and then
I still need to test on those systems. Back-and-forth with those with
access to systems… won’t fly.

Also, if I port things now, retro enthusiasts can still use it in
a decade or more.

>I think you are misunderstanding or misrepresenting what happens. If the
>vendor no longer supports it, the number of users will drop dramatically.

Until it raises again.

>That makes porting to those systems useful for a smaller and smaller group

So what? Even a small amount of users is nice to support.

And then, there *is* the effect of weird compilers finding bugs
that other compilers missed or have as extensions.

(That being said, I just found a miscompile in xlc 16 that xlc 12
does not have. But that just makes me rearrange the code in a way
I had planned already anyway.)

>*Give up on weird ISAs and platforms*

Fuck NO. That’s a monoculturism attitude that leads to worsening
in software quality and a spiral of death.

>Companies should be paying for this directly: if pyca/cryptography actually
>broke on HPPA or IA-64, then HP or Intel or whoever should be forking over

If you continue this thought, we would never have gotten support
for FOSS Unix on i386 because Intel would not have paid for it.

Everything has to start somewhere, and whatever hardware someone
can get their hands on as spare (i.e. not their main working systems)
is a good start.

Additionally, having to throw away hardware because of loss of
support is e-waste and bad for the environment. I already hate
the “smartphone” world for it, let’s avoid doing so for normal
computers. (I vocally criticised various distros’ plans for i386,
armel, etc. for the same reason.) The hardware still works and
does its job decently.


On Sun, 15 Feb 2026, Martin Guy via cfarm-users wrote:

>I don't care about the corporate platform or indeed about the
>corporation itself but I do care about people who, for whatever reason,
>are using them and would be able to provide better solutions for the
>people who depend on their work by "leveraging" (as they say in
>Fordspeak) the immense richness that open source provides.

Yes!

>It is in every hardware and OS company's interests to make it as easy
>as possible for OSdevs to ensure their stuff works on their platform
>for the benefit obtained by everyone from free
>compilers/libraries/support apps and I have contacted "IBM support"
>highlighting this thread and suggesting generosity with licenses.

Thanks.

bye,
//mirabilos
-- 
Gestern Nacht ist mein IRC-Netzwerk explodiert. Ich hatte nicht damit
gerechnet, darum bin ich blutverschmiert… wer konnte ahnen, daß SIE so
reagier’n… gestern Nacht ist mein IRC-Netzwerk explodiert~~~
	(as of 2021-06-15 The MirOS Project temporarily reconvenes on OFTC)