Re: There is no choice
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA512
On 09/23/2014 at 10:19 AM, Lisi Reisz wrote:
> On Tuesday 23 September 2014 14:53:15 Joe wrote:
>
>> As Lisi and others have pointed out, the design of systemd and
>> the decision of Debian to make systemd the 'default', or in
>> practice the only init system, are set in stone. There's nothing
>> to discuss.
>
> I have never said anything of the kind. I have said only that it
> is up to the developers to decide what they do. They are
> volunteers.
>
> If someone wants to maintain or develop something else, then
> obviously they can. There is no stone about it. But DO something
> constructive. Maintain sysvinit. Write patches for the
> applications that depend on systend so that they do not depend on
> systemd.
As I have pointed out repeatedly, the problem is not that the
applications depend on systemd; in fact, often they don't.
They depend on external functionality, which happens to be implemented
in systemd, and (at least initially) only in systemd.
There are two possible solutions to this: change systemd so that that
functionality is implemented in a way which does not require systemd to
be PID 1 (and preferably so that the functionality is entirely external
to, and merely depended on by, systemd), or implement the functionality
outside of systemd in such a way that the applications can depend on
your implementation instead.
That latter approach is what the systemd-shim project has been trying to
do; certainly, helping to do so would be a constructive thing to do. But
that project has bugs, and will perpetually be playing catch-up to
systemd, as long as the systemd project does not embrace (which is
distinct from "not reject") the idea of implementing this functionality
outside of systemd itself - and that would be the former approach.
systemd-shim, and the projects it draws on such as cgmanager, are a
workaround for the dependency problems caused by flaws in the design of
systemd. A true fix would have to happen in systemd itself, by changing
that design. However, since that design appears to be actively
intentional, and since changing it in a way which would address those
flaws would be a major project and might involve dropping some features,
it seems highly unlikely that the systemd developers would agree to
making such a change.
> Etc. But stop winging. Stop taking over this list with endless
> useless complaints.
I agree that pure complaints do only so much good, and that that good is
- - at least in practice - fairly limited.
But complaints are not all that has been happening here, and while the
complaints have not been shut down by the posts objecting to the
systemd-discussion traffic, it looks to me as if much of the discussion
which is not purely complaints and might be more potentially useful
*has* been cut off in response to those objections.
If you're addressing only the few people who are doing nothing but
complaining, then okay, fine (although I'm not 100% positive there is
anyone still posting complaints who isn't also, at least occasionally,
saying something worth noting). But the comments I see from you read to
me as if they are directed at everyone who's taking part in the
systemd-related discussions here (aside possibly from people with
specific "I saw this problem, how do I fix it?" questions), which is
overly broad-brush IMO.
> And listen to what anyone else is saying. You want to be heard.
> Hear others.
This I do agree with.
- --
The Wanderer
The reasonable man adapts himself to the world; the unreasonable one
persists in trying to adapt the world to himself. Therefore all
progress depends on the unreasonable man. -- George Bernard Shaw
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1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=/rt3
-----END PGP SIGNATURE-----
Reply to: