Re: systemd .service file conversion
On Fri, May 31, 2013 at 06:12:38PM +0200, Ondřej Surý wrote:
> On 31. 5. 2013, at 15:53, Jeff Epler <jepler@unpythonic.net> wrote:
> > The idea that somehow users of non-linux kernels don't matter or don't
> > even exist as debian users is one of the most frustrating bits of this
> > whole thread.
>
> I would happily support any non-linux kernel arch in form of integrating patches, but the reality is that Hurd and FreeBSD kernels are just toys.
>
> That doesn't mean the toys are not important (...all work and no play...), they are, but they must not stop the inovation.
[...]
Please don't talk about 'innovation' in Linux. There really isn't
anything particularly new going on here, just incremental development
of useful features. There is also plenty of work being done to add
new features to the FreeBSD kernel (not all of which gets merged back
into it) and apparently a fair amount on Hurd. Their major weakness
at the moment is in hardware support (this is an extremely weak point
for Hurd), and that can mostly be dealt with by making them work
efficiently in VMs.
The real problem we have is that these three kernels are not matching
each other's features, and if we want to run mostly the same userland
on all of them then it can only rely on the common subset of features.
That potentially leaves Debian trailing behind Linux-only
distributions that aren't limited in this way.
I think it makes more sense for all the non-Linux kernels to be
supported through debian-ports, allowing porters to apply kernel-
specific workaround patches, drop unsupportable packages, and release
on their own schedule. There is a risk, of course, that this results
in continuing divergence if patches don't get fed 'upstream' to the
main Debian archive.
Ben.
--
Ben Hutchings
We get into the habit of living before acquiring the habit of thinking.
- Albert Camus
Reply to: