Updates on GNU/Hurd progress

19 points by FedericoSchonborn


nemin

I really don't know what to think of Hurd.

On one hand, it's evident that its developers are putting in the time and the effort. It's not Linux-distro n+1 (since it's not even Linux), nor is it a toy, nor is it something so single-purpose that it'd only ever be useful to a very specific group.

The fact that so much can be built on it is also really impressive. I was most surprised that they got Rust to work, feels kind of poetic to see something so ancient and modern touch bases.

On the other hand, I wonder if Hurd will ever see much traction... It took Linux decades to be "respected" by big players (I'm mainly thinking of desktop-related stuff, drivers, support, etc.) and---despite being very well-usable for years---there's still people jeering on it about an always just-out-of-reach "Year of the Linux Desktop".

With all its cool resilience and microkernel design, I'm not sure Hurd has something that'd differentiate it for the average user, who might not really care about the kernel itself, just what their system can run and what the ecosystem is like.

I guess I could see usage in places where resilience is a must and systems must be able to recover from near-catastrophic failure states, but to me that feels like a niche that's both small and might already be filled by a hypervisor. And even then, it'd take some miraculous disruption to make big companies move their entrenched systems from one architecture to another.

I really don't want to sound dismissive of the amount of time and effort and ingenuity spent on Hurd, but I wonder if they lost the race before they even really started when Torvalds dropped his kernel and beat them to a punch (followed by Hurd's long dormancy). I don't think Linux will be our "forever system", surely eventually there will be something better or Linux will evolve to such a point it's completely different from what we know today, but I'm not convinced Hurd is the one to take its throne.

ethoh

Still hasn't managed to move beyond MACH.

Lots of sideways progress, but no forward advancement.

mperham

Is this anything but an academic exercise? Is there any demand for it?