Bug-for-bug compatibility is great, but having an actual feedback loop into enterprise Linux development is better.
-
Bug-for-bug compatibility is great, but having an actual feedback loop into enterprise Linux development is better. I wrote about my experience running CentOS Stream in production, how it handles security updates without the rebuild lag, and why the upstream model beats the old CentOS way.
https://blog.hofstede.it/why-i-prefer-centos-stream-over-old-centos/
-
R relay@relay.mycrowd.ca shared this topic
-
Bug-for-bug compatibility is great, but having an actual feedback loop into enterprise Linux development is better. I wrote about my experience running CentOS Stream in production, how it handles security updates without the rebuild lag, and why the upstream model beats the old CentOS way.
https://blog.hofstede.it/why-i-prefer-centos-stream-over-old-centos/
@Larvitz so you're running a rolling release candidate in production... I suppose an unstable dev would be worse but you dare the devil. 🫣
Stable community rebuilds are less unconscious ways. Still it's a workaround for something broken by design.
The real solution is to run away from redhat ecosystem -> #debian #opensuse #mageia #mint ... -
@Larvitz so you're running a rolling release candidate in production... I suppose an unstable dev would be worse but you dare the devil. 🫣
Stable community rebuilds are less unconscious ways. Still it's a workaround for something broken by design.
The real solution is to run away from redhat ecosystem -> #debian #opensuse #mageia #mint ...@maat always depends on the use case. For my infra (basically some Podman containers and a family Nextcloud) it’s totally fine for many years. Always chose the right too for the task
-
@maat always depends on the use case. For my infra (basically some Podman containers and a family Nextcloud) it’s totally fine for many years. Always chose the right too for the task
@Larvitz it's totally fine... until you realize, too late, that your evaluation of risk was rotten.
You can as well walk blindfolded in a minefield, confidently because « there should not be a lot of mines left ».
You can get your way through it once, twice and tell people around it's fine...
But it's not. -
Bug-for-bug compatibility is great, but having an actual feedback loop into enterprise Linux development is better. I wrote about my experience running CentOS Stream in production, how it handles security updates without the rebuild lag, and why the upstream model beats the old CentOS way.
https://blog.hofstede.it/why-i-prefer-centos-stream-over-old-centos/
Really interesting.
Looking for something innovative-but-not-experimental for a workstation environment with R, python, and some tools for having fun. Should give it a try, I guess.
