Scheduled Maintenance: We are aware of an issue with Google, AOL, and Yahoo services as email providers which are blocking new registrations. We are trying to fix the issue and we have several internal and external support tickets in process to resolve the issue. Please see: viewtopic.php?t=158230

 

 

 

Where will you go after systemd?

Here you can discuss every aspect of Debian. Note: not for support requests!
Post Reply

Where will you go after systemd?

BSD
12
16%
Linux without systemd
34
47%
Mac
2
3%
Windows
0
No votes
something totally different
1
1%
have not decided yet
24
33%
 
Total votes: 73

Message
Author
Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: Where will you go after systemd?

#121 Post by Deb-fan »

Even though @ this point am questioning myself as to why, WHY, WHYYYYYYYY !??!! I'm still posting stuff here. Take a look at this

Linus Torvalds, a guy that came out w the Linux kernel and has been using it and FOSS longggggg before 99.9999999% of humanity had even known such a thing existed. He's using systemd on his main work-station AND his laptop. Clearly he must consider it the worst piece of software ever conceived of huh ?

He as is his usual wont, aka: No bullchit, say what you know and tell it like it is and if people don't like it. Obviously the guy really doesn't care. I approve and admire Linus not just from the fact that the guy is a fricken software mastermind but on a personal level too.

Mr Torvalds has issues w the head-dev behind systemd and the feeling is very much mutual. See that by way of a very clear example.

Jmo and comments on it, think the guy is too harsh on Linus Torvalds in that. Though also don't doubt what the guy has to say about people wanting to literally kill him for projects he's worked on. Well ... kinda ridiculous to me. I know there are nutjobs, fudders and tinfoil-azzhatists aplenty in Gnu/Nix communities far and wide.
Edit: Also pointless, though think the guy was having a bad dy and did many other people wrong too. ie: The gnu/nix communities as a whole. I can't really blame him though, can try to imagine what it's like to wake up, visit this forum, that blog, open an email acct and it's jammed packed with hate mail and even death threats from clueless fektards around the world.

Knowing that some of them actually mean it too. Try to imagine what it must be like having to worry about your safety and more importantly the safety of your family/friends. Must be real comforting knowing some epic out of touch azzhat could show up on your doorstep w a sawed off shotgun and satchel full of pipe bombs huh ? Yeah that's a wonderful way to win the guy and dev-team over to their opponents way of thinking. MEGA FACE-PALM !
Also though I know a heckuva lot less about Lennart Poettering, than Mr Torvalds. Seems obvious to me they are both extremely knowledgeable and gifted in the field of technology. Not really surprising they'd clash or isn't to me, shrugs. Both experts ... both w a preferred style of working on things. Torvalds gets bashed often enough too. By people that are totally clueless and doubtless unworthy to wash hiz dirty underwear.

Just saying a ship has ONE captain for a reason. If it's dragging barges full of fektards and azzhats screaming about their precious freedoms, choices, rights and lofty principles and raging about how unfair it is that they don't get a turn to steer da boat. Well ... that's probably cuz the craft would quickly end up on the rocks or at best spinning aimlessly in circles. :P

Sighs ... also more specifically on the topic of systemd vs other candidates. See this link

See that long lists of yes's vs the others and the no after no after no's. Looks to me like Poettering and the devel-team, he's not the only person involved are doing some dang good things and must have a tad bit of knowledge and skills in turning so many no's into yes's.

Some folks ( who probably know even less than myself and that's with me readily admitting am not a systemd or init guru) are likely saying, those yes's aren't good dude, errrr even though I don't even know what they mean ! That's because systemd is trying to take over the world and because it was secretly developed in a lab sponsored by Satan himself !

Or fallback to FUD and rehashed FUD, it's monolithic/non-modular dang it, it's anti-opensource goss darn it, it's the software anti-christ and I'm going to Bsd ! In some ways yep, some not at all. Imo some real steps in a positive direction, shrugs. Think Debian made the right choice too, not that I'm qualified to even question people in their league, I'm an end-user here.

Yeah but what about journald and binary this n that !?!?! Well ... run syslog-ng alongside the sucker if you know what it is or actually have any need to (edit: or mask, limit it or or or etc). If folks spent 1/1millionth of their time actually setting aside ridiculous end of the world FUD and learnt something about using sytemd, they'd have a much happier life. Not to mention any admin worth the label long since has and therefore was/is light-years ahead of all the tinfoilists, whiners, criers etc etc.

Anyway, don't know if this compiles w Debian forum rules or not but wanted to share one last tidbit of info some folks may find interesting. I'm selling hats made out of the far superior aluminumfoil at extremely affordable pricing. PM me if you'd like 2 buy, bulk rate discounts and satisfaction guaranteed ! Also I PROMISE, no more walls of text from me in this thread, no need to put me on ignore, sheesh. :D
Also yes, I'm a Debian and FOSS fanboi, shoot me I guess, shrugs. I must be nutz being crazy about all this awesome software just begging to be used ! Yeah I'm a real nutjob. *Face, meet palm* :)
Last edited by Deb-fan on 2015-01-26 15:14, edited 2 times in total.
Most powerful FREE tech-support tool on the planet * HERE. *

schnuller
Posts: 386
Joined: 2014-11-25 05:05

Re: Where will you go after systemd?

#122 Post by schnuller »

tomazzi wrote:
twoflowers wrote:If you want systemd in debian to be reconsidered, it'd be a good idea to post all these bugs on the debian bugtracker and not on freedesktop.org.
Actually, after Ian Jackson's GR have been rejected, I see no hope for Debian.
That GR was completely logical: <simplification> if Your calculator program have a hard dependancy on a kind of init system, then this is simply stupid (saying gently).
This alone leaves no hope for reconsidering the choice of default init system.

Debian (or perhaps only some of the developers) have worked hard to reject that GR, so for me, it means that Debian is lost. Not direclty due to adopting of systemd, but because political reasons were preferred over technical/logical ones in this case.

Therefore, I've decided to write my own init system, compatible with systemd, so I'll be able to use Debian repositories, but without having to use systemd. And soon nobody will be able to use Debian (but also most of other distros) without systemd.

Regards.

ps. in the mean time, I'v got 1 response for my report, regarding 1 of the reported bugs. I've responded to that email, but for some rason it wasn't added to a thread - so if someone is interrested, here it is:
http://lists.freedesktop.org/archives/s ... 27395.html
1) I want to bump this for obvious reasons
2) I agree with the conclusion there ain't much point in Debian anymore
3) I read the bug-report. It ain't of much use for a non-programmer, but interesting anyway.

User avatar
kolker
Posts: 81
Joined: 2013-08-22 07:16

Re: Where will you go after systemd?

#123 Post by kolker »

Probably gentoo or lfs is the best long term option in Linux considering how you create the system. Though those Linux systems epidemise choice you have to do much yourself and you could probably still use things that may use systemd natively if you use the right build switches.

Deb-fan
Posts: 1047
Joined: 2012-08-14 12:27
Been thanked: 4 times

Re: Where will you go after systemd?

#124 Post by Deb-fan »

Dang it ! Will try to keep it brief and let the linkage, speak for itself.

1. The systemd-shim is apparently maintained by Debian itself. See that

So was both right + wrong on the matter, thus where learning things tends to be productive eh. :) Just goes to show, in addition to a gazillion other choices/freedoms people still retain over using systemd, Debian is once again moving mountains and bending over backwards to satisfy and take care of the Debian userbase. As ever, demonstrating amazing tech skills, consideration and competence. Yep ... my ( non-metalic-foil)hat is once again off to these awesome people.

2. Some thoughts from a guy who actually knows of which he speaks (unlike so many conspiracy nutters), a Siduction maintainer see that

Just to reinterate the painfully obvious on this topic, yet again and again and again. Plenty of people infinitely better qualified than vocal ignorant nutjobs are concerned with systemd vs alternate init, if only to avoid needless hostility and other socio-type, non-technical problems the issue is presenting. Also for every frothing @ the mouth FUD infected anti-systemd troll, seen TONS ( including Linus Tovalds himself) of people happy with systemd and/or people who'd rather employ one of the MANY ( some quite easy) alternatives, work-arounds-etc to taking control of "their" OS platform. Compared to whiners and azzhatz unwilling to even utilize one of the many (also seemingly very easy) alternatives the people behind Debian have gone out of their way to provide.

Sheesh talk about biting the hand, that makes someone gourmet food and feeds em on a silver-spoon. These azzhatz don't just bite that hand, they try to kick Debian in the nutz for all it's awesome efforts. Imo of which, these people are grossly unworthy.

Ha ! Right on the borderline of a wall of text ! Though at this point promise no posts of any length 2 this FF&FF= (FUD FANATICS & FEKTARD FESTIVAL.) That is not directed @ everyone, people I'm talking about know who they are, or should, shrugs.

PS, on review, nope it's a wall of friggin text afterall, I lied ... sue me. :)
Most powerful FREE tech-support tool on the planet * HERE. *

mean_dean
Posts: 132
Joined: 2014-12-29 18:21

Re: Where will you go after systemd?

#125 Post by mean_dean »

Deb-fan wrote: 1. The systemd-shim is apparently maintained by Debian itself.
It is packaged and maintained by debian devels the same as any package in Debian.

But the source comes from the site of Ryan Lortie who is with the Gnome project. http://people.gnome.org/~desrt/

The copyright file lists the following:
Files: *
Copyright: Copyright: 2010 Lennart Poettering
Copyright: 2011 Lennart Poettering
Copyright: 2011 Bastien Nocera <hadess@hadess.net> / 2007 David Zeuthen <david@fubar.dk> / 2013 Canonical Limited
Copyright: 2013 Canonical Limited
License: GPL-2+

Files: debian/*
Copyright: 2013 Sebastien Bacher <seb128@ubuntu.com>
License: GPL-2+

twoflowers

Re: Where will you go after systemd?

#126 Post by twoflowers »

@tomazzi: I find the answer of Mr. P most intersting:
> registering one sig handler at a time should be used. Then, we can
> tell (log) which signals were not registered by sigaction, and take
> conscious decision what to do next.

We actually want to handle failure of installing these crash handlers
all the same way: by mostly ignoring them, and proceeding.

Lennart

tomazzi
Posts: 730
Joined: 2013-08-02 21:33

Re: Where will you go after systemd?

#127 Post by tomazzi »

twoflowers wrote:@tomazzi: I find the answer of Mr. P most intersting:
> registering one sig handler at a time should be used. Then, we can
> tell (log) which signals were not registered by sigaction, and take
> conscious decision what to do next.

We actually want to handle failure of installing these crash handlers
all the same way: by mostly ignoring them, and proceeding.

Lennart
Perhaps It would look better if I will spend another few hours on respoding to e-mails, but in fact I'm tired.

With only that 3 bugs I've reported, it showed up that key sysd developers actually have no idea of what they're doing and why (f.e. that funny case when one of them failed to understand manual for kill(2) ), and Mr P only confirms that fact.
It's just a waste of time to try to explain (again) why signal handlers are required to work in a serious project.

Regards.
Odi profanum vulgus

User avatar
thanatos_incarnate
Posts: 717
Joined: 2012-11-04 20:36

Re: Where will you go after systemd?

#128 Post by thanatos_incarnate »

Let's also quote the entire reply for the sake of completion:
On Mon, 26.01.15 00:33, Tomasz Pawlak (tomazzi at wp.pl) wrote:

> You are right, but it's not as simple as it may look at first sight:
>
> 1. If we allow the process to continue without sig handlers
> installed, then results can be just catastrophic: kernel panic with
> all the services launched -> broken transanctions, half-written
> records/files, etc -> total mess, corrupted or lost data etc. So,
> since successfull installation of the sig handlers is one of the
> most critical parts of initialisation, it is actualy safer to just
> quit. This is just a critical fault (and is currently completely
> ignored).

Hmm? no. if PID 1 dies then either the kernel halts PID 1 or we do.

> 2. Another thing is, that those signals are not equivalently
> important, f.e. SIGABRT can be throwed by thousants lines of code in
> this project (by abort()), so it is much more likely that assertion
> checking will prevent segfaults, throwing SIGABRT instead. This
> means that SIGABRT is actually far more probable than SIGSEGV. This
> in turn leads to simple solution: the process should unconditionally
> exit if hander for SIGABRT have failed to install, but with other
> sig handlers failed, we may take a risk and continue. In any case,
> such situation should be logged as soon as possible. Ignoring this
> is just asking for catastrophe.

The only thing you can do to recover from SIGABRT or SIGSEGV,
reexec()ing yourself from the sig handler. That' something the kernel
doesn't allow for PID 1 however...

It's illusionary to believe that you could just do some magic, and
return from SIGSEGV and continue running your program. You
cannot. SIGSEGV is more often than not an indication for a memory
corruption, and if that happens, there's no way to bring back the
memory to a state where things are good again, because memory doesn't
tell you if its in a good or bad state.

> 3. SIGFPE: how often the code uses FPU? -> I mean, that handler for
> this sig can be dynamically installed/unistalled when needed,
> probably only on a thread level, not for the whole process. This
> will allow to completely safely report failed sigaction by assertion
> checking.

SIGFPE is also triggered by integer divisions by zero (yeah, the name
is misleading).

Catching SIGFPE, SIGSEGV, SIGABRT and so on are for software problems
that we don't exptect. If we expected them then we could certainly
handle them in a nicer way than getting a signal thrown...

> 4. So, sigaction_many() should be removed (also because it is a
> vararg function, what is rather bad idea), and a function for

Ahum? vararg is bad now? I must have missed that memo. Why would it be
bad? Do you write C code without printf() (which is varargs)?

> registering one sig handler at a time should be used. Then, we can
> tell (log) which signals were not registered by sigaction, and take
> conscious decision what to do next.

We actually want to handle failure of installing these crash handlers
all the same way: by mostly ignoring them, and proceeding.

Lennart

--
Lennart Poettering, Red Hat

tomazzi
Posts: 730
Joined: 2013-08-02 21:33

Re: Where will you go after systemd?

#129 Post by tomazzi »

I've mislooked that reply, because it is a response to a detached (by accident) mail, and I've expected a reply in a main thread.
I'm going to reply to that later.

edit:
done.

edit2:
This is a link to my reply:
http://lists.freedesktop.org/archives/s ... 27648.html

This is a link to my final reply in the main thread: (the thread was splitted by accident, what is my fault)
http://lists.freedesktop.org/archives/s ... 27701.html

Normally, there's no point to provide a link to a reply to a reply, but I think this time it can be interresting:
http://lists.freedesktop.org/archives/s ... 27705.html
> 1. Ignoring return value from sigaction is not a problem, because
> it's almost *unlikely* to fail.
> In any case, kernel will do the job, so it doesn't even matter
> whether the handlers are installed or not -> everything that can be
> done in the handler is to just quit anyway.
Actually the signal handler does sleep in an infinite loop, to not
crash the system.
This is just outstanding:
They are trying to prevent a crash without checking whether the crash handler was installed at all.

I wish them luck, cause without sig handlers installed journald doesn't even work - as they are using sig handlers to pass messages to journald.

Final conclusion:
Excluding the fact, that I had to explain to Lennart why it is stupid to use vararg function to handle a preprocessor symbol (what is really hilarious :) ), the key problem with systemd is that the developers have only a hope, that their code will somehow work, likely in most of general cases, without taking care of even the most obvious bugs.

But in fact, it turns out to be good: I'm just even more convinced, that I should continue my work on a "forking" of systemd.

Regards.
Odi profanum vulgus

bdtc1
Posts: 42
Joined: 2015-01-22 09:00

Re: Where will you go after systemd?

#130 Post by bdtc1 »

To summarise: there are no bugs in core/main.c:

1. Ignoring return value from sigaction is not a problem, because it's
almost *unlikely* to fail.
In any case, kernel will do the job, so it doesn't even matter
whether the handlers are installed or not -> everything that can be done
in the handler is to just quit anyway.

2. SIGSEGV handler (and others): it's *unlikely* to happen that the
handler will be executed at the bottom of the main stack (with
insufficient stack space or when the stack is already overflowed) -
because stacks are huge, and in case of segfault see point 1.

3. SA_NODEFER is OK, cause it's *unlikely* that external signal gets
delivered. Even if the crash() handler will crash (f.e. because it's
not re-entrant), what (again) is *unlikely* to happen, see point 1.

I've learned a lot, sorry to waste Your time.
Just to be clear, is the above sarcasm? Because it could be interpreted as saying that you finally agree and see their point.

twoflowers

Re: Where will you go after systemd?

#131 Post by twoflowers »

http://lists.freedesktop.org/archives/s ... 27705.html
OK, great to hear that what we have currently should work :)

Zbyszek
It just hurts to read these lines. Can't they just take a junior class in system programming?

User avatar
Head_on_a_Stick
Posts: 14114
Joined: 2014-06-01 17:46
Location: London, England
Has thanked: 81 times
Been thanked: 132 times

Re: Where will you go after systemd?

#132 Post by Head_on_a_Stick »

twoflowers wrote:It just hurts to read these lines. Can't they just take a junior class in system programming?
Can we just clarify (for the benefit of other users reading this) exactly what the experience level of @tomazzi is?

He talks a good talk and I certainly do not have the expertise to critique his reasoning but it seems to me that he agrees with the systemd developers in the message you have linked.

Which software is he responsible for?

What has he written and where is it used?

Which packages (if any) does he maintain for this or any other distribution?

I do not mean this to be a personal attack on @tomazzi, but it would allow other users to better judge his capabilities and whether he is in a position to critically analyse the systemd code base.
deadbang

bdtc1
Posts: 42
Joined: 2015-01-22 09:00

Re: Where will you go after systemd?

#133 Post by bdtc1 »

I certainly do not have the expertise to critique his reasoning
...
allow other users to better judge his capabilities and whether he is in a position to critically analyse the systemd code base.
I do have the expertise, although my time is required elsewhere and so I haven't looked at the code for myself. I can verify that tomazzi seems to have some good points, based on what he has excerpted here. I could be wrong, but some of the code's embedded comments appear to indicate that even the authors know that they are being rather sloppy at times.

To put it another way, the things which he has pointed out so far probably would be required security updates if they were discovered in other packages. The fact that they are found in an init system just makes them worse.

All of this could in fact be no big deal in the end, but it could also be very important. Open review is an important part of open source, and I believe that this issue merits an extensive open review.

User avatar
Head_on_a_Stick
Posts: 14114
Joined: 2014-06-01 17:46
Location: London, England
Has thanked: 81 times
Been thanked: 132 times

Re: Where will you go after systemd?

#134 Post by Head_on_a_Stick »

bdtc1 wrote:I do have the expertise
What software are you responsible for?

What have you written and where is it used?

Which packages (if any) do you maintain for this or any other distribution?
deadbang

twoflowers

Re: Where will you go after systemd?

#135 Post by twoflowers »

@Head_on_a_Stick:

What software are you responsible for?

What have you written and where is it used?

Which packages (if any) do you maintain for this or any other distribution?

User avatar
Head_on_a_Stick
Posts: 14114
Joined: 2014-06-01 17:46
Location: London, England
Has thanked: 81 times
Been thanked: 132 times

Re: Where will you go after systemd?

#136 Post by Head_on_a_Stick »

twoflowers wrote:@Head_on_a_Stick:

What software are you responsible for?

What have you written and where is it used?

Which packages (if any) do you maintain for this or any other distribution?
None whatsoever -- I have already clearly stated that I do not have the expertise to dissect any code base.

My point is that anyone can post authoritative-sounding bulls**t -- how are the users of this forum supposed to be able to separate the bs-merchants from the genuine experts?

EDIT: Anyway, I though I was on your "ignore" list...
:P
deadbang

tomazzi
Posts: 730
Joined: 2013-08-02 21:33

Re: Where will you go after systemd?

#137 Post by tomazzi »

bdtc1 wrote:
To summarise: there are no bugs in core/main.c:

1. Ignoring return value from sigaction is not a problem, because it's
almost *unlikely* to fail.
In any case, kernel will do the job, so it doesn't even matter
whether the handlers are installed or not -> everything that can be done
in the handler is to just quit anyway.

2. SIGSEGV handler (and others): it's *unlikely* to happen that the
handler will be executed at the bottom of the main stack (with
insufficient stack space or when the stack is already overflowed) -
because stacks are huge, and in case of segfault see point 1.

3. SA_NODEFER is OK, cause it's *unlikely* that external signal gets
delivered. Even if the crash() handler will crash (f.e. because it's
not re-entrant), what (again) is *unlikely* to happen, see point 1.

I've learned a lot, sorry to waste Your time.
Just to be clear, is the above sarcasm? Because it could be interpreted as saying that you finally agree and see their point.
Of course this is sarcasm.

There's no point to argue with statements like "it's unlikely to happen, so let's just ignore it". I call this bad and dangerous programming practice, but it can be also pure stupidity. Pick one.

Lennart said that he could accept a patch for this - so I'm writting one - alternative, systemd-compatible soultion (with properly designed signals and errors handling). I'll send him a link when the code reaches testing stage.... ;)

Regards.
Odi profanum vulgus

bdtc1
Posts: 42
Joined: 2015-01-22 09:00

Re: Where will you go after systemd?

#138 Post by bdtc1 »

tomazzi wrote:There's no point to argue with statements like "it's unlikely to happen, so let's just ignore it". I call this bad and dangerous programming practice, but it can be also pure stupidity. Pick one.
I wonder what they'd say about the "asynchronous sync" issue. Data corruption can be worse than system lock-up.

tomazzi
Posts: 730
Joined: 2013-08-02 21:33

Re: Where will you go after systemd?

#139 Post by tomazzi »

System "lock-up" (this term can mean a really wide range of situations) can cause data corrupiton too.

Anyway, I've spent many hours in replying to e-mails and posting replies on systemd-devel, only to discover that it was a waste of time.
I think I know what will be the answer to "sync problem" - it's *unlikely* to happen... (and hard to catch, so it will randomly cause problems "for no obvious reason" ... ;) )

Therefore, instead of wasting yet more time, I'll just continue to work on my project(s)

Regards.

ps.
There's also no point to post portions of broken systemd code here, cause when looking at the replies, it seems obvious that it's *unlikely* to find someone who could understand it...
Odi profanum vulgus

User avatar
mardybear
Posts: 994
Joined: 2014-01-19 03:30

Re: Where will you go after systemd?

#140 Post by mardybear »

@tomazzi: You sounded frustrated and fatigued a bit back, so just wanted to wish you good luck with your init development. Hopefully you will get something useful out of your effort for yourself...and maybe for us non-programmers. Always good to have choices, that's what Linux used to be about. Still is to some extent but feels more restrictive.
800mhz, 512mb ram, dCore-jessie (Tiny Core with Debian Jessie packages) with BusyBox and Fluxbox.
Most don't have computer access, reuse or pay forward an old computer.

Post Reply