| 1 |
On Sun, 16 Feb 2014 15:16:36 -0600 Canek Peláez Valdés wrote: |
| 2 |
> On Sun, Feb 16, 2014 at 2:58 PM, Volker Armin Hemmann |
| 3 |
> <volkerarmin@××××××××××.com> wrote: |
| 4 |
> > Am 16.02.2014 21:08, schrieb Canek Peláez Valdés: |
| 5 |
> >> On Sun, Feb 16, 2014 at 12:59 PM, Volker Armin Hemmann |
| 6 |
> >> <volkerarmin@××××××××××.com> wrote: |
| 7 |
> >> [ snip ] |
| 8 |
> >>> or it is an idiotic decision. Because features means complexity. |
| 9 |
> >> Yeah, like the kernel. |
| 10 |
> >> |
| 11 |
> >>> Complexity means bugs. |
| 12 |
> >> Bugs get reported, bugs get fixes. Life goes on. |
| 13 |
> |
| 14 |
> You didn't answered this, did you? |
| 15 |
|
| 16 |
Bugs are different. Bugs in the critical system components are |
| 17 |
critical to the whole system. If Libreoffice or browser |
| 18 |
segfaults, some data may be lost and inconvenience created, but the |
| 19 |
system will continue to run. If PID 1 segfaults — everything is |
| 20 |
lost, you have a kernel panic. That's why critical components should |
| 21 |
be as simple and clean as possible. |
| 22 |
|
| 23 |
SysVinit code size is about 10 000 lines of code, OpenRC contains |
| 24 |
about 13 000 lines, systemd — about 200 000 lines. Even assuming |
| 25 |
systemd code is as mature as sysvinit or openrc (though I doubt this) |
| 26 |
you can calculate probabilities of segfaults yourself easily. |
| 27 |
|
| 28 |
> >> All of them are different tools providing one capability to systemd as |
| 29 |
> >> a whole. So systemd is a collection of tools, where each one does one |
| 30 |
> >> thing, and it does it well. |
| 31 |
> >> |
| 32 |
> >> By your definition, systemd perfectly follows "the unix way". |
| 33 |
> >> |
| 34 |
> > |
| 35 |
> > no, it isn't. |
| 36 |
> > |
| 37 |
> > How are those binaries talk to each other? |
| 38 |
> |
| 39 |
> dbus, which is about to be integrated into the kernel with kdbus. |
| 40 |
|
| 41 |
And this is a very, very bad idea. Looks like you don't know matter at |
| 42 |
all: to begin with kdbus protocol is NOT compatible dbus and special |
| 43 |
converter daemon will be needed to enable dbus to talk to kdbus. The |
| 44 |
whole kdbus technology is very questionable itself (and was |
| 45 |
forcefully pushed by RH devs), anyway it is possible to disable this |
| 46 |
stuff in kernel and guess what will be done on my systems. |
| 47 |
|
| 48 |
> > Looks broken. Broken by design. The worst form of broken. |
| 49 |
> |
| 50 |
> By your opinion, not others. |
| 51 |
|
| 52 |
That is not just an opinion. There is a science and experience behind |
| 53 |
system's design. And all that science was ignored during systemd |
| 54 |
architecture process if there was any at all. |
| 55 |
|
| 56 |
Best regards, |
| 57 |
Andrew Savchenko |
Archives