Blind vs. mainstream distros

[Linux for blind general discussion]

Le 23/04/2017 à 23:18, Linux for blind general discussion a écrit :
> Debian Testing isn't exactly a rolling release in the same way as Arch.
> Debian Unstable probably comes closer, but even that's not the same. The
> problem is that Arch for example takes all component packages from the
> same desktop version, ie. GNOME, whereas Debian even unstable takes
> parts of 2 or 3 different versions and tries to fit them together.

That's right. I'm investigating about it as I'd like to understand this
asynchronous.

> Worse, back when I ran Debian Unstable on a desktop, I had lots of
> apt-get breakage where packages were broken due to dependent packages
> that had either fallen behind the broken package or new dependencies
> that had been added to the broken package, but were not yet in the
> repository. These things don't necessarily fix themselves in Testing
> unfortunately, as the only requirement for a package to enter Testing
> from Unstable is a 1-week waiting period, which will naturally be just
> as many days behind for the dependency as it was when the packages
> entered Unstable. There is a good reason why Debian's newer repositories
> are called Testing" and "Unstable." Arch at least tries, and for the
> most part does, keep all component packages for the major desktops at
> the same version. The only exceptions I've seen are where some packages
> don't get a point release, so you may see for example MATE 1.18.1 and
> 1.18.2 packages, but you will never see MATE 1.16 and 1.18 packages, Nor
> do you often see packages that break due to missing or old dependencies,
> as those are supposed to be filtered out in Arch testing and earlier. I
> did see a single exception to that where a Python package broke due to
> an outdated dependency, but this is quite rare in Arch, as I have only
> seen that once or twice in more than 6 years.

It's indeed an imrrurcodnt need I think. At least, testing shou;d just
have relevant packages without breakage. Unstable is a really workspace,
so why not. But testing should be clean to be tested. But it's something
which needs to be discussed, defended, by users, by people. So in front
of such problems, I prefer fixing them inside than leaving it due to
them. To make the global project improve. But right, the current
situation is not acceptable.

> ~Kyle
> 
> _______________________________________________
> Blinux-list mailing list
> Blinux-list at redhat.com
> https://www.redhat.com/mailman/listinfo/blinux-list
> 

-- 
Logo Hypra 	
Photo Jean-Philippe MENGUAL *JEAN-PHILIPPE MENGUAL**
DIRECTEUR TECHNIQUE ET QUALITÉ*
adresse84, quai de Jemappes, 75010, Paris
téléphone+331 84 71 06 61 <tel:+33%201%2000%2000%2000%2000>
courieljpmengual at hypra.fr
site webwww.hypra.fr
Facebook Hypra <https://www.facebook.com/hyprasoftware/>Twitter Hypra
<https://twitter.com/Hypra_>Linkedin
<https://fr.linkedin.com/in/jean-philippe-mengual-800133135>