[arch-general] Lennart Poettering on udev-systemd

[Brandon Watkins]

On Tue, Aug 14, 2012 at 10:55 AM, Baho Utot <baho-utot at columbus.rr.com>wrote:

> On 08/14/2012 10:32 AM, Brandon Watkins wrote:
>
>> On Tue, Aug 14, 2012 at 9:28 AM, Jelle van der Waa <jelle at vdwaa.nl>
>> wrote:
>>
>>  On 08/09/12 22:00, Anthony ''Ishpeck'' Tedjamulia wrote:
>>>
>>>> I think what he was saying wasn't that systemd is hard but switching is
>>>>
>>> hard irrespectively of what you're switching to.
>>> Because the devs made systemd being able to use rc.conf?
>>>
>>> It takes less then a day to use systemd, but I am not forcing you to use
>>> it.
>>>
>>> --
>>> Jelle van der Waa
>>>
>>> Yeah, I found systemd very easy to learn. The wiki page is great, and
>>>
>> after switching to it I prefer it because I just find it a lot easier to
>> deal with than sysvinit IMO. For example I find systemd's .service files
>> so
>> much cleaner and easier to understand than initscripts, they are also
>> portable and can be included in upstream packages.
>>
>> This "Oh my god systemd is hard and I'm being forced to use it!" FUD I
>> keep
>> seeing is getting pretty ridiculous... Even if arch does someday switch to
>> systemd, I'm sure initscripts will be supported for quite some time,
>> giving
>> plenty of time to learn/transition (again really not that hard) in the
>> event that that ever happened.
>>
>> Arch has always been a bleeding edge constantly changing distro, if you
>> want everything to stay the same forever, use debian. No matter what
>> happens with this whole sysvinit vs systemd kerfuffle, you will never be
>> "forced" to use systemd in arch, just like you've never been forced to use
>> sysvinit...
>>
>
> I don't think you fully understand the issue.
>
> If udev was still a "stand alone package" and not part of systemd as it is
> now....
> Then systemd would be an alternative init system and all the other init
> systems would not be impacted and one could use any of the system init
> methods he chooses.  If you would want systemd becames it works for you
> great...knock yourself out...but on the other hand when this thing becomes
> fully matured then systemd will be the only one that works well with udev
> and everyone else be damned.
>
> Lennart Poettering by his own submission stated that he wanted udev as
> part of systemd and that he doesn't care about any other init system that
> would use udev.  As with Lennart it seems as it's my way or the
> highway...which indeed is the problem.
>
>
> Poettering didn't kidnap the udev developers and force them to merge with
systemd. And yes I am aware of his comments regarding udev, I saw a comment
elsewhere that I think explains what he meant pretty well:

"What he's saying is "non-systemd systems are dead in our eyes because no
one is maintaining them; we will maintain udev without systemd as promised,
but don't ask us to spend our time making it pretty; if you want that pay
someone to do that for you".

I don't see what's unclear here."


Lets take a hypothetical situation: If udev someday only works well with
systemd (which is wild speculation...) then if there is enough interest, an
alternative would appear for people who don't use systemd. If there isn't
enough interest in other init systems and an alternative then you could
suck it up and switch.


Also, I will state once again that I think people are
highly exaggerating the "difficulty" of transitioning an arch install to
systemd, its quite simple. If arch were to one day switch to systemd and
not support initscripts, it would not be the end of the world (and again
this is wild speculation/FUD in the first place...)