SoylentNews is people
SoylentNews
posted by
janrinok
on Wednesday November 19 2014, @11:25AM
from the I-hope-we-don't-regret-this dept.
from the I-hope-we-don't-regret-this dept.
Ian Jackson's general resolution to prevent init system coupling has failed to pass, the majority vote deciding that the resolution is unnecessary. This means that not only will Debian's default init be systemd, but packages will not be required to support other init systems. Presumably, this means that using other init systems on Debian (without using systemd as a base) will not be possible without major workarounds, or possibly at all. It also leaves the future of Debian projects such as kFreeBSD unclear, as systemd is linux specific.
The vote results can be found here
The winners are:
Option 4 "General Resolution is not required"
This discussion has been archived.
No new comments can be posted.
Debian General Resolution to Prevent Init System Coupling Fails to Pass
|
Log In/Create an Account
| Top
| 179 comments
| Search Discussion
The Fine Print: The following comments are owned by whoever posted them. We are not responsible for them in any way.
Can anyone remember when the times were not hard, and money not scarce?
(Score: 4, Interesting) by FatPhil on Wednesday November 19 2014, @02:42PM
https://vote.debian.org/~secretary/gr_initcoupling/tally.txt
#!/usr/bin/perl -w
while(<>) {
my ($a1,$a2,$b) = ($_ =~ m/V: (.)(.).(.)./) or next;
if ($a1 eq '-'){$a1=6;}
if ($a2 eq '-'){$a2=6;}
if ($b eq '-'){$b=6;}
my ($min,$max) = $a1<$a2 ? ($a1,$a2) : ($a2,$a1);
if($b<=$max and $b>=$min) { print; }
}
phil@bazspaz:~/tmp$ perl tally.pl < tally.txt | wc
90 457 3634
phil@bazspaz:~/tmp$ wc tally.txt
483 2450 19433 tally.txt
So nearly a fifth of the vote had "4" (keep head in sand) separating "1" and "2". I've not got a condorcet calculator, but if you replaced "1" and "2" with preferred("1","2"), I have a feeling the outcome would have been different. In particular as it was fairly close.
It's easy to say it was gerrymandered that way, but if the "1" and "2" preferrers aren't smart enough to see that they have more in common than they do with the enemy, then they deserve to lose.
And I'll join the masses now in ex-Debian Diaspora. It's getting harder and harder to find an OS which supports my hardware, my philosophy, and my way of working.
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 0) by Anonymous Coward on Wednesday November 19 2014, @02:51PM
Yes, the sophistry can be avoided and the questions can be more accurately stated.
1. Systemd as the only init with packages depending on its bundled payload
2. Systemd as an availiable init with no packages depending on its bundled payload
3. Do not care
That is the situation and it seems option 2 would have won that one eh?
(Score: 2) by gringer on Wednesday November 19 2014, @07:34PM
Your option 1 is closest to Option 3 "Packages may require specific init systems if maintainers decide".
Your option 2 is closest to Option 1 "Packages may not (in general) require a specific init system".
Your option 3 is closest to Option 4 "General Resolution is not required".
Bear in mind that the text for these options was bashed out and agreed upon by voters. Any one person may have a view of options that they think is better than the ones that are used, so the chosen options are necessarily a compromise.
Because you include the apathy option, it wins over all others. There were no situtations in the vote where another option defeated option 4. Here are the fights for the closest matches to what you suggested:
Option 3 defeats Option 1 by ( 263 - 183) = 80 votes.
Option 4 defeats Option 1 by ( 323 - 147) = 176 votes.
Option 4 defeats Option 3 by ( 308 - 135) = 173 votes.
In the ordering, this would be 4 > 3 > 1, or using your numbers, 3 > 1 > 2. I find it doubtful that a re-wording of the options would change the ranking that much to put last place into first place.
Ask me about Sequencing DNA in front of Linus Torvalds [youtube.com]
(Score: 2) by opinionated_science on Wednesday November 19 2014, @03:07PM
I wondered about the voting options too. Opensuse has had systemd wrapped in init-scripts for a few years now. I don't have a problem with systemd, with the exception I don't like the binary blob logs. They should use zip like everyone else!!
Anyway, we should remember the enemy is the non-FOSS proprietary industry that exists solely to lock-in consumers.
I would hope well all remember this.
(Score: 2) by mechanicjay on Wednesday November 19 2014, @05:53PM
Opensuse has had systemd wrapped in init-scripts for a few years now.
Yes, 12.3 was okay, but 13.1 was basically broken because of it. With 13.2 they seem to have completed the transition to systemd, which is good as the init system is actually controllable now.
I'm still torn on the whole thing, honestly. systemd is maturing and working better, which is good. I don't really have a good technical argument for or against it. It just breaks with the historic philosophy of doing one small thing well, and chaining together the tools, which is why I'm upset about it.
A technical argument cannot overcome a philosophical divide, which I think is what this whole thing boils down to: People who have a philosophical objection to a tool like this, and those who don't. Unfortunately, this ends up with straw-man argument flame-wars, rather than keeping everyone happy and staying init system agnostic -- which baffles me.
My VMS box beat up your Windows box.
(Score: 2) by opinionated_science on Wednesday November 19 2014, @05:59PM
exactly - so long as it works i am not bothered. My only technical complaint about systemd it the opaque logs - maybe it is buried in a manual somewhere, but any log i cannot open with less/vi is broken.
Logs are for humans - databases are for machines.
(Score: 2) by Arik on Wednesday November 19 2014, @08:05PM
You need to think on this more clearly. That historical philosophy is either important and should be followed for technical reasons (which I believe to be correct) or else it's not relevant at all here.
You do one thing and do it well, because it's possible to verify a simple tool with a clearly defined role actually does what it's supposed to do, and not more, and doing that is the key to keeping a sane system that works as it should.
You avoid designing monsters like systemd not for some disconnected 'philosophical' reason but because decades of experience show that monsters like that are impossible to do right. They are far too complicated to verify - too much code and too many code-paths. Too many pieces of it that will rarely be executed, where bugs can sit for years without being noticed.
If systemd is still in use in 10 years, they will just be beginning to find the bugs they are writing today.
If laughter is the best medicine, who are the best doctors?
(Score: 2) by cafebabe on Thursday November 20 2014, @03:57AM
As fairly well defined and protocol specific examples, some of problems in bind and Apache httpd have occurred due to the conflated role of serving authoritative data and caching data from other servers. Neither of these implementations are doing one thing well.
1702845791×2
(Score: 2) by FatPhil on Wednesday November 19 2014, @03:34PM
So even IJ preferred the head in the sand approach to "Support for other init systems is recommended, but not mandatory".
I like the guy, of all the big-name Debianites, he's probably one of the top 3 that I agree with most often, but if even he can't work out the consequences of such voting, then he's as naive as the other 89. He needs to watch /The People vs. Larry Flint/, and understand that in order to keep the small thing you want alive, you sometimes have to keep bigger things, not all of which you want, alive.
And here's his resignation letter:
"""
The majority of the project have voted to say that it was wrong of me
to bring this GR at this time. Despite everything that's happened, I
respectfully disagree.
"""
And on that, I call bullshit. He put "the GR was wrong of me to call" *higher* than "Support for other init systems is recommended, but not mandatory".
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by mechanicjay on Wednesday November 19 2014, @04:07PM
Maybe I need more coffee this morning, but the results seem needlessly obfuscated with a results matrix. Did I miss where the raw vote count is?
My VMS box beat up your Windows box.
(Score: 2) by FatPhil on Thursday November 20 2014, @08:26AM
Great minds discuss ideas; average minds discuss events; small minds discuss people; the smallest discuss themselves
(Score: 2) by cafebabe on Wednesday November 19 2014, @07:09PM
So, the amendments split the vote? That's a lesson for the future.
1702845791×2