| 1 |
On Sun, Feb 16, 2014 at 2:58 PM, Volker Armin Hemmann |
| 2 |
<volkerarmin@××××××××××.com> wrote: |
| 3 |
> Am 16.02.2014 21:08, schrieb Canek Peláez Valdés: |
| 4 |
>> On Sun, Feb 16, 2014 at 12:59 PM, Volker Armin Hemmann |
| 5 |
>> <volkerarmin@××××××××××.com> wrote: |
| 6 |
>> [ snip ] |
| 7 |
>>> or it is an idiotic decision. Because features means complexity. |
| 8 |
>> Yeah, like the kernel. |
| 9 |
>> |
| 10 |
>>> Complexity means bugs. |
| 11 |
>> Bugs get reported, bugs get fixes. Life goes on. |
| 12 |
|
| 13 |
You didn't answered this, did you? |
| 14 |
|
| 15 |
>>> And you don't want complexity in PID1 or init. Let those 'features' be |
| 16 |
>>> handled by their own specialists. |
| 17 |
>> Almost all the features of systemd live outside of PID 1. |
| 18 |
>> |
| 19 |
>>> You know, the unix way. Do one thing, do it well. |
| 20 |
>> This is from my desktop machine: |
| 21 |
>> |
| 22 |
>> /usr/lib/systemd/systemd-reply-password |
| 23 |
>> /usr/lib/systemd/ntp-units.d |
| 24 |
>> /usr/lib/systemd/systemd-coredump |
| 25 |
>> /usr/lib/systemd/systemd-hostnamed |
| 26 |
>> /usr/lib/systemd/systemd-binfmt |
| 27 |
>> /usr/lib/systemd/systemd-localed |
| 28 |
>> /usr/lib/systemd/systemd-machined |
| 29 |
>> /usr/lib/systemd/systemd-sleep |
| 30 |
>> /usr/lib/systemd/system-generators |
| 31 |
>> /usr/lib/systemd/system-generators/systemd-system-update-generator |
| 32 |
>> /usr/lib/systemd/system-generators/systemd-gpt-auto-generator |
| 33 |
>> /usr/lib/systemd/system-generators/systemd-efi-boot-generator |
| 34 |
>> /usr/lib/systemd/system-generators/systemd-fstab-generator |
| 35 |
>> /usr/lib/systemd/system-generators/systemd-getty-generator |
| 36 |
>> /usr/lib/systemd/system-generators/gentoo-local-generator |
| 37 |
>> /usr/lib/systemd/systemd-fsck |
| 38 |
>> /usr/lib/systemd/systemd-bootchart |
| 39 |
>> /usr/lib/systemd/systemd-shutdown |
| 40 |
>> /usr/lib/systemd/systemd-random-seed |
| 41 |
>> /usr/lib/systemd/system-sleep |
| 42 |
>> /usr/lib/systemd/systemd-remount-fs |
| 43 |
>> /usr/lib/systemd/user-generators |
| 44 |
>> /usr/lib/systemd/systemd-sysctl |
| 45 |
>> /usr/lib/systemd/systemd-timedated |
| 46 |
>> /usr/lib/systemd/catalog |
| 47 |
>> /usr/lib/systemd/system-shutdown |
| 48 |
>> /usr/lib/systemd/systemd-udevd |
| 49 |
>> /usr/lib/systemd/systemd-multi-seat-x |
| 50 |
>> /usr/lib/systemd/systemd-cgroups-agent |
| 51 |
>> /usr/lib/systemd/systemd-user-sessions |
| 52 |
>> /usr/lib/systemd/systemd-journal-gatewayd |
| 53 |
>> /usr/lib/systemd/systemd-quotacheck |
| 54 |
>> /usr/lib/systemd/systemd-shutdownd |
| 55 |
>> /usr/lib/systemd/systemd-modules-load |
| 56 |
>> /usr/lib/systemd/systemd-backlight |
| 57 |
>> /usr/lib/systemd/systemd-ac-power |
| 58 |
>> /usr/lib/systemd/systemd-initctl |
| 59 |
>> /usr/lib/systemd/systemd-readahead |
| 60 |
>> /usr/lib/systemd/systemd-journald |
| 61 |
>> /usr/lib/systemd/systemd-activate |
| 62 |
>> /usr/lib/systemd/systemd |
| 63 |
>> /usr/lib/systemd/systemd-update-utmp |
| 64 |
>> /usr/lib/systemd/systemd-vconsole-setup |
| 65 |
>> /usr/lib/systemd/systemd-logind |
| 66 |
>> |
| 67 |
>> All of them are different tools providing one capability to systemd as |
| 68 |
>> a whole. So systemd is a collection of tools, where each one does one |
| 69 |
>> thing, and it does it well. |
| 70 |
>> |
| 71 |
>> By your definition, systemd perfectly follows "the unix way". |
| 72 |
>> |
| 73 |
> |
| 74 |
> no, it isn't. |
| 75 |
> |
| 76 |
> How are those binaries talk to each other? |
| 77 |
|
| 78 |
dbus, which is about to be integrated into the kernel with kdbus. |
| 79 |
|
| 80 |
> Besides - why is garbage essential for booting in /usr? |
| 81 |
|
| 82 |
Is not. Most of it is optional, in a server I have there are much less binaries. |
| 83 |
|
| 84 |
> Looks broken. Broken by design. The worst form of broken. |
| 85 |
|
| 86 |
By your opinion, not others. |
| 87 |
|
| 88 |
>>> Use text to communicate. |
| 89 |
>> systemd can comunicate basically everything via text: |
| 90 |
>> |
| 91 |
>> centurion ~ # systemctl show sshd.service | head |
| 92 |
>> Id=sshd.service |
| 93 |
>> Names=sshd.service |
| 94 |
>> Requires=basic.target |
| 95 |
>> Wants=system.slice |
| 96 |
>> WantedBy=multi-user.target |
| 97 |
>> Conflicts=shutdown.target |
| 98 |
>> Before=shutdown.target multi-user.target |
| 99 |
>> After=syslog.target network.target auditd.service |
| 100 |
>> systemd-journald.socket basic.target system.slice |
| 101 |
>> Description=OpenSSH server daemon |
| 102 |
>> LoadState=loaded |
| 103 |
>> |
| 104 |
>> For performance reasons, some things are passed or stored as data. Bu |
| 105 |
>> everything works with text also. So, again, it passes your definition. |
| 106 |
>> |
| 107 |
> |
| 108 |
> oh? I can pipe that output into cat or any any daemon I like? Doesn't |
| 109 |
> look like so. |
| 110 |
|
| 111 |
But it does, you can "cat" with journalctl; it's one of its output options: |
| 112 |
|
| 113 |
-o, --output= |
| 114 |
cat |
| 115 |
generates a very terse output only showing the actual |
| 116 |
message of each journal entry with no meta data, not even a timestamp. |
| 117 |
|
| 118 |
>>> That stuff. That makes things easy. And flexible. And replaceable. |
| 119 |
>> Easy to whom? And systemd is more flexible that a lot of init systems, |
| 120 |
>> in my opinion including OpenRC. |
| 121 |
> |
| 122 |
> oh really? because everything is done by the magical Pöttering? |
| 123 |
|
| 124 |
OK, sorry, I thought you wanted to have a civil, serious, technical |
| 125 |
conversation. |
| 126 |
|
| 127 |
I'm done with you in this thread. |
| 128 |
|
| 129 |
Regards. |
| 130 |
-- |
| 131 |
Canek Peláez Valdés |
| 132 |
Posgrado en Ciencia e Ingeniería de la Computación |
| 133 |
Universidad Nacional Autónoma de México |
Archives