VSIDO systemd and upstart discussion

[ozitraveller]

I prefer this analysis of why systemd is the better choice. And I hope commonsense prevails.

We seem to be at the point of the process where at least those of us who
did early investigation are stating conclusions.  I think I have enough
information to state mine, so will attempt to do so here.

This is probably going to be rather long, as there were quite a few
factors that concerned me and that I wanted to investigate.

The brief summary is that I believe Debian should adopt systemd as its
default init system on Linux.  There are two separate conceptual areas in
which I think systemd offers substantial advantages over upstart, each of
which I would consider sufficient to choose systemd on its own.  Together,
they make a compelling case for systemd.  This position would have
substantial implications for upgrade paths and for non-Linux ports; I'll
discuss a bit of that below, but most of it in the separate branch of this
bug report that Ian opened on that topic.

Below, I first discuss the other choices before us besides systemd and
upstart.  Then I look at a straight technical comparison between the two
init systems, and finally look at issues of maintenance, community,
ecosystem, and portability.  The three main criteria on which I was
evaluating both systems were technical capabilities, surrounding
ecosystem, and portability.  The latter two turned out to be deeply
entangled, so I discuss them together.

Read more:
https://lists.debian.org/debian-ctte/2013/12/msg00234.html

[VastOne]

^ As do I, the problem is that the deck is stacked on that TC and the consensus is that it will be a 5-3 vote for upstart based on what has been documented

And I have thrice read all of this... my eyes are bleeding   

[ozitraveller]

Thrice!

upstart can't replace all of systemd, so why just replace a component of systemd with upstart.

It should go to a community vote!

If TC go for upstart all TC ctte positions should be declared vacant and elections held!

Where to next FreeBSD, Haiku, .....


I'm totally dismayed by their stupidity....

[statmonkey]

I always hope that logic and rationality prevail ozitraveller but the problem with sense is that it is anything but common - if you want proof read that whole Ian Jackson post.  He comes off to me like a spurned female rather than an adult making a professional decision.

In the post you linked to I was struck by these paragraphs:

One of the points that I think may have been obscured in the discussion,
but which is important to highlight, is that basically all parties have
agreed that Debian will adopt large portions of systemd.  systemd is an
umbrella project that includes multiple components, some more significant
than others.  Most of those components are clearly superior to anything we
have available now on Linux platforms and will be used in the distribution
going forward.

In other words, this debate is not actually about systemd vs. upstart in
the most obvious sense.  Rather, the question, assuming one has narrowed
the choices to those two contenders, is between adopting all the major
components of systemd including the init system, or adopting most of the
major components of systemd but replacing the init system with upstart.
Either way, we'll be running udev, logind, some systemd D-Bus services,
and most likely timedated and possibly hostnamed for desktop environments.

I think this changes the nature of the discussion in some key ways.  We're
not really talking about choosing between two competing ecosystems.
Rather, we're talking about whether or not to swap out a core component of
an existing integrated ecosystem with a component that we like better.

I think I had read this before but not sure I "got" it.  So conceptually I am imagining one of my team members coming to me and saying:

1. We have a rival solution that has some nice features but it also has some clear cut disadvantages to our in house solution. But it's no problem because even though they have been working on it for 3 years and haven't addressed those issues they are trivial to address and they will after we buy it.

2. Oh by the way the solution that we have in house is better and adds a lot more and also is an integrated solution.  In fact, even if we select the flawed rival we will keep all the integrated stuff from the in house solution and the flawed rival will rework their stuff to work with the innovative stuff from our in house team. But the in house solution isn't portable to other 'nix's and even though they say that it can be addressed at some point I think it yet they won't. The rival soltuion won't dove tail as well as a fully integrated solution would but it will be trivial to fix and they will do so after we buy it.

3. One more thing, I know all current systems use parallel daemons and IPv6 has been around for years (as have sockets) but they didn't write those into their rival solution.  But it's trivial and they will do so after we buy it.

4. Oh and one more little thing. In order to accept this rival solution they want us to sign an agreement that all our contributors license their code through them and of course in case of disputes over ownership I would guess they would win since they hold the licensing and I know that is a violation of our founding principles but it is trivial and they say that maybe they would look at reviewing it after we buy it.

5. BTW the in house solution was worked out by a large portion of our community and has been accepted by some of our most loyal members of the community and the rival is a small group that operates behind a corporate logo and we really don't have any influence or input directly into what they do and they haven't even named a project lead but that is trivial and I am sure they will fix it after we buy it.

6. So in conclusion I have decided that we will swap out that portion with the rival teams because the guys who do the in house team didn't accept my suggestions and I don't like them they are hard to work with and they picked on me and took my lunch money.

After I told the guy to clean out his desk I would probably start over.  I mean really this whole thing is just surreal to me.  Why is there any discussion?  What strikes me in reading these is that the guys who support Upstart either have no technical support for their positions (it's always about response times and niceties or in the case of Steve its some obscure point that no one cares about).  The systemd supporters give facts, details about where systemd is better etc.  and actual processes they went through in detail to examine the differences and how they came to the conclusion.  It's a little Wizard of Ozish "pay no attention to the man behind the curtain".

I really have nothing to add and guess at this point I am just venting my frustration.  In a way it's like watching a friend die on the operating table for no reason.  You can clearly see the doctor making mistakes and he is just cavalierly going along.  I hope I am wrong and there is an upset here but I am preparing for the grieving process just the same.  My apologies if this is the same stuff over again.  I did at least try to make some new points.  Not that it matters.

[jedi]

@statmonkey, wow just wow!  Would you mind if I posted that on my blog?  Giving you full credit of course!  It's been a stampede there the last couple of days because I copied a couple of my posts here to there. (jedsdesk.com)  Let me know...

[VastOne]

@statmonkey, epic post man... did you write that on the fly or have to prepare it beforehand? 

Very succinct!   

[statmonkey]

@statmonkey, wow just wow!  Would you mind if I posted that on my blog?  Giving you full credit of course!  It's been a stampede there the last couple of days because I copied a couple of my posts here to there. (jedsdesk.com)  Let me know...

You don't need to remind me by the way, I visit your site already. 

Sure Jed anything in this thread for that matter.  What's mine is yours. Hmm, there's an idea. Would you like my ex-wife's phone number by the way? 

v-ger.  I am a fly guy you should know that by now    (checking my fly now in fact)

[jedi]

@statmonkey, thanks for that, and as to the phone number, nahhh.  Happily married 20 years now.  If this one doesn't work out I think I'll call it quits on marriage!

[ozitraveller]

@statmonkey, awesome post and I can sense your frustration running right through that explantion/comment. I have been following along with the discussion and cannot for the life of me understand how upstart could possibly be considered. I'm honestly dumbfounded! Hopefully work will not stop on systemd and the decision, if upstart is chosen, will be reversed. I feel sorry for the devs having their excellent work dismissed for "crap".

[statmonkey]

I prefer this analysis of why systemd is the better choice. And I hope commonsense prevails.

Read more:
https://lists.debian.org/debian-ctte/2013/12/msg00234.html

I have just reread Russ Alberry's position paper.  I think that is at least THHHRRRIIICCCCEE!!! The v-ger magic number and I really think it should be on the required reading list.  Not only is it the anti-thesis of the other posts in that thread it is exactly what it should be.  Well thought out and reasoned.  Supported by facts, logic and rationale.  Among the highlights not yet posted here:

Debian has
more than enough hard integration problems to solve without creating new
ones for ourselves unnecessarily.  But that's the key word: unnecessarily.
If we do have a reason for doing this, we should seriously consider it.

Therefore, I believe the burden of proof is on upstart to show that it is
a clearly superior init system along some axis, whether that be
functionality or portability or flexibility or maintainability, to warrant
going to the effort of disassembling a part of the systemd ecosystem and
swapping in our own component.

I should state up front that, in making these sorts of decisions around
free software projects, I have a relatively high future discount rate.  In
other words, I give substantially less credit to something that does not
exist now but could exist in the future.

Right now, neither systemd nor upstart work on non-Linux platforms.
Therefore, right now, adopting either of them means that we either
jettison our non-Linux ports or adopt a transition plan that retains
support for sysvinit scripts.

Note: they are hammering systemd for not being portable and yet ... yep it's true they are not portable and have presented no plan for getting there.

So, in short, I consider portability to be a possible benefit of upstart,
but I'm inclined to discount that advantage for several reasons.  One,
it's not yet actually materialized and still may not, and two, systemd
porting looks like it's going to be on the table regardless.

One of the reasons why I'm leery of the future portability argument for
upstart, and one of the reasons why I'm leery of upstart in general, is
that I'm quite worried upstart will prove to be a blind alley.

Red Hat adopted upstart but never did a wholescale conversion, and then
  abandoned upstart in favor of systemd.  ... But I
  think it's still worth noting that the only non-Ubuntu major adopter of
  upstart backed away from it.

I'm concerned that, if we adopt upstart, in two or three years we'll end
up wanting to do the same thing that Red Hat did, back out, and switch to
systemd.  That would be a huge amount of wasted effort.

This is a conclusion that I totally agree with.  What is not being said is the whole why would this happen.


WHAT IF - the why is ...  That Canonical could give a flying diminutive rodents lower locopodus whether upstart works for Debian or not.  Similar to what might have happened with Gnome and Unity, they worked them like circus dogs to get their stuff into and tested with Gnome, took patches from Gnome devs under the CLA, patched Gnome code without GPL'ing those patches and passing them along and then abandoned the Gnome devs.  Wrapping it all up with a little bow by stealing a significant base of users and pointing fingers at Gnome and saying look those guys never followed through. Plus parading Unity as if they made it out of whole cloth.  Of course the Gnome guys could not do anything but look like fools because Canonical didn't share their code and they thought they were in a relationship. They could not do anything because of the CLA and were just stuck looking like chumps.

In this theoretical case it's to get access to the Debian install base, integrating into systemd to the level of the parts being inseparable, getting help from the Debian dev community all under CLA, then not passing back that code and then dumping Debian, taking the now fully integrated and inseparable Upstart/udev/logind/etc. spinning it as their own (lets see last time they called it unity, this time maybe they call it ReachAround) and acting like they are the wronged party.  Those who fail to learn from history are doomed to repeat it.  Think Lucy and the football for you old guys like me.

[ozitraveller]

If I planted/buried upstart in my backyard what would it grow into? A stunted mutant maybe.....

Now about the ex-wife's phone number??

[statmonkey]

If I planted/buried upstart in my backyard what would it grow into? A stunted mutant maybe.....

Now about the ex-wife's phone number??

How about me burying my ex in your backyard and I give you upstarts phone number? 

[VastOne]

Oh my... the dogs have been let loose... 

[statmonkey]

Oh my... the dogs have been let loose... 

Hmm on second glance those final two paragraphs do seem a little harsh 

There now you won't have to worry about libel quite so much.

[PackRat]

Where to next FreeBSD, Haiku, .....

Where to next? I'll stick my neck out here - so bookmark this post  - if upstart is adopted, the purists devs will leave and fork (probably not the correct word here) debian into a new distro with systemd. More power to them if they do.