Szerző: T.J. Duchene Dátum: Címzett: dng Tárgy: Re: [Dng] Is forking enough?
On Wed, 3 Dec 2014 12:20:24 +0000
Gregory Smith <gregorysmith1974a@???> wrote:
If I may call you "Greg" or "Gregory"? =)
I'm not going to waste personal time responding to your views on women
or other groups, nor am I going to egg you on or provoke you. I'm just
saying this not to antagonize you, but so you understand that my
responses are intended to be as friendly as possible. Even if I feel
your comments are personally insulting, I'm willing to ignore the whole
matter.
If you can offer something less controversial than political
views I would very much appreciate your comments in another area. You
came here to work with Devuan, so please, may we focus on how we can
make Devuan better?
> >>> Linux userland didn't always use crap like DBUS, policykit, etc.
> >>> One could just revive the old branches, one by one. They worked
> >>> fine. Often had better performance. The code still exists in a
> >>> findable form thanks to debian publishing source ISOs every
> >>> release.
This is somewhat true, but most projects maintain concurrent revision
systems, such as git, svn or heaven forfend: cvs. You can use that to
download older versions of their work from their main repositories.
Speaking very generally, of course, projects seldom throw away code.
Most of them archive it in what are known as branches. If you are
interested in packaging older revisions of code for Devuan, you can
download and maintain the branches that you want to.
> >>> Linus loves systemd aswell. All the linux hire-ups are fucks these
> >>> days (maybe they always were). Maybe the kernel needs to be forked
> >>> eventually too (GRSecurity basically does this with their epic
> >>> patch, so it's not impossible).
I realize that this might be a minority opinion, but there is nothing
wrong with systemd that more development time cannot cure. POSIX has no
standard for init. I just do not feel that systemd is ready for use in
mission critical areas, that we have a stake in.
As for maintaining their own version of the kernel, every distribution
does that when they maintain their own patchset to fix bugs.
Eventually, if things go as they should, and if those patches are
acceptable, those patches get merged back upstream. Generally speaking
again, these minor changes never go far, the kernel is a huge
undertaking: millions of lines of code. That's not something that
anyone forks completely without considerable thought, and I do not
think that it would be wise for Devuan to waste limited resources on
such a task at present.