1 |
On Tue, Nov 11, 2014 at 07:18:30AM -0500, Rich Freeman wrote |
2 |
> On Tue, Nov 11, 2014 at 12:20 AM, Walter Dnes <waltdnes@××××××××.org> wrote: |
3 |
> > |
4 |
> > What worries me is that Lennart has been able to get modifications |
5 |
> > done to the kernel, e.g. kdbus. I know this'll sound paranoid, but how |
6 |
> > long before he pushes a patch that requires systemd to run the linux |
7 |
> > kernel? |
8 |
> |
9 |
> Apologies if this comes across a bit agro, but: |
10 |
> |
11 |
> 1. kdbus isn't in the kernel (though that seems likely to happen at some point) |
12 |
> 2. it does sound paranoid |
13 |
> 3. if he pushes a patch that requires systemd I'd be shocked if Linus merged it |
14 |
> |
15 |
> Can you find one example of any situation where the linux kernel has |
16 |
> ever required any specific implementation of anything in userspace as |
17 |
> a matter of policy in its 23 year history? I'm sure you could find |
18 |
> some examples of cases where there just happened to be one de-facto |
19 |
> implementation of something, but even that might be tough with all the |
20 |
> diversity in the linux world. |
21 |
|
22 |
It might not be an official official requirement, but if the upstream |
23 |
gets rolled into systemd, then we depend on the "goodwill" of systemd |
24 |
devs not to go and break anybody else's userspace implementation. |
25 |
Lennart and "goodwill" do not belong in the same sentence. How's |
26 |
systemd-shim working out for Debian??? I'm old enough to remember the |
27 |
OS/2-versus-Windows wars. At one point, IBM had Windows 3.1 running |
28 |
inside of OS/2. Then Microsoft issued "a minor update" (Windows 3.11) |
29 |
and it no longer ran inside OS/2. It took a while for IBM to get |
30 |
Windows 3.11 running inside OS/2. That's the kind of hostility that |
31 |
non-Lennart userspace software faces. |
32 |
|
33 |
> Linus himself has articulated some of the reasons why kdbus is likely |
34 |
> to get merged. It fills in a gap in Linux as compared to many |
35 |
> competing operating systems, and it is logical to implement at the |
36 |
> kernel level. That is generally the criteria for getting stuff into |
37 |
> the kernel, and is basically good software design. The linux kernel |
38 |
> is all about stable userspace ABIs - if there is only one |
39 |
> implementation of something it is probably because nobody was bothered |
40 |
> enough to write another. |
41 |
|
42 |
We Gentoo folks got our wakeup call ( literally |
43 |
http://lists.freedesktop.org/archives/systemd-devel/2014-May/019657.html ) |
44 |
from the guy who signs himself as "Lennart Poettering, Red Hat" |
45 |
(nuff said). |
46 |
|
47 |
I don't know how long udev will run standalone without systemd. My |
48 |
desktop PC is running mdev. See https://wiki.gentoo.org/wiki/Mdev and |
49 |
https://wiki.gentoo.org/wiki/Mdev/Automount_USB |
50 |
|
51 |
eudev is also an option. Hopefully, device management doesn't get |
52 |
forced into systemd. |
53 |
|
54 |
-- |
55 |
Walter Dnes <waltdnes@××××××××.org> |
56 |
I don't run "desktop environments"; I run useful applications |