1 |
On 22.02.2014 21:21, Stroller wrote: |
2 |
> |
3 |
> On Sat, 22 February 2014, at 10:38 am, Yuri K. Shatroff |
4 |
> <yks-uno@××××××.ru> wrote: |
5 |
> |
6 |
>> On 22.02.2014 11:40, Mark David Dumlao wrote: |
7 |
>>> [ ... ] Even as the complex beast it has become systemd is still |
8 |
>>> simpler than the alternative of having abominations of unreliable |
9 |
>>> shell scripts checking to see which version of grep and sed is |
10 |
>>> used to split the command line, or whether the system uses |
11 |
>>> tempfile or mktemp, or depending on perl. |
12 |
>> |
13 |
>> Well, simpler yeah, supporting only one kernel of specific versions |
14 |
>> is always simpler than trying to support everything from SunOS to |
15 |
>> NetBSD. This way, if the kernel supported only e.g. Intel |
16 |
>> IvyBridge+ with one chipset family, one graphics (VESA) and so on, |
17 |
>> it would have been incredibly simpler. |
18 |
> |
19 |
> I’m doing a (free) operating system (just a hobby, won’t be big and |
20 |
> professional like gnu) for 386(486) AT clones. … PS. Yes – it’s free |
21 |
> of any minix code, and it has a multi-threaded fs. It is NOT |
22 |
> protable (uses 386 task switching etc), and it probably never will |
23 |
> support anything other than AT-harddisks, as that’s all I have :-(. |
24 |
|
25 |
Good luck! |
26 |
|
27 |
> Linux did indeed once support only one CPU family and one or two |
28 |
> hard-drives, and the reason that it now supports more is that people |
29 |
> dug into the code, submitted patches and got it fixed. |
30 |
> |
31 |
> Had all the original Linux developers spent their time on the |
32 |
> comp.os.minix list, complaining "oh, those splitters, they're trying |
33 |
> to fragment the Minix community" and "this Linus guy should be |
34 |
> putting his effort into improving the Minix kernel", where would we |
35 |
> be today? |
36 |
|
37 |
Actually I don't get what you are arguing [against]. |
38 |
|
39 |
> It's almost hilarious the volume of traffic expended here on this |
40 |
> subject, especially that by the naysayers. When I first learned of |
41 |
> systemd I did not feel favourably towards it, but those ranting |
42 |
> against it have only given Canek a platform to convince me. |
43 |
|
44 |
I partially agree. In an emotional discussion the most probable winner |
45 |
(as seen from outside) is the calm one. |
46 |
But being calm doesn't refute all technical and `political` stuff. |
47 |
I personally was going to try systemd about a week ago when the |
48 |
discussion started. Now I'm quite convinced not to do this in the near |
49 |
future. |
50 |
No calm arguments of systemd's supporters, such as the complexity of |
51 |
shell scripts, the simplicity of systemd compared to the Kernel, the |
52 |
ease of use of journald tools, the shitload of troubles of configuring |
53 |
syslog, the replacement for all network setup tools, the good intents of |
54 |
Red Hat, etc etc, didn't convince me. Emotions pass, results remain. |
55 |
|
56 |
> And whilst I'm still of two minds on which init system I'd ideally |
57 |
> prefer, I am not under any delusions that I can influence the |
58 |
> developers of the Gentoo distro or those of the Linux kernel (who |
59 |
> AIUI are adding kdbus to support systemd), either by ranting about it |
60 |
> here or otherwise. |
61 |
|
62 |
No delusions, there will always be an alternative. |
63 |
Nobody actually has disagreed yet with my words that in a couple of |
64 |
years systemd is going to dominate "90%" (meaning the majority of) linux |
65 |
distros. But "10%" hopefully will remain without it. Anyway since |
66 |
systemd is not intending to support any other kernels, we'll probably |
67 |
see other OS or stuff like Debian/kFreeBSD develop more intensively. |
68 |
Yet, of course, these alternatives will necessarily be poorer supported |
69 |
and one will have to take effort to migrate - to either the distro he |
70 |
used, but the version with systemd, or a different distro/OS. |
71 |
|
72 |
> The amount of energy spent on this, you could have established a fork |
73 |
> and written code by now - if y'all really want to prove your point, |
74 |
> that's the way to do it. |
75 |
|
76 |
What point? |
77 |
I personally am terribly satisfied with the SysV init and shell scripts |
78 |
so what am I to fork and write? |
79 |
What a fork to establish? A fork of debian, to maintain it w/o systemd? |
80 |
Let that be done by debianners maybe, if they so desire. |
81 |
|
82 |
As for `ranting`, I do see a point in such talks (until these get |
83 |
personal), as I learn many new things (both from the posts and while |
84 |
trying to prove/refute the points) and I always try to ask a concrete |
85 |
question and answer a concrete question. Note: I do repeat *I* here |
86 |
because you answered *my* post. |
87 |
|
88 |
In any case, no offense, your reply is a rant, too. |
89 |
|
90 |
-- |
91 |
Best wishes, |
92 |
Yuri K. Shatroff |