[linux-elitists] How about a fork of Jessie without systemd?

Post by Bob Bernstein
To anyone's knowledge are there projects afoot to create a new linux
distro representing, basically, Debian Jessie but without all the
systemd and Network Manager um stuff. Or, have I thus basically
defined a fork of Wheezy? Perhaps one of those is underway?

Devuan is a thing. It even furnishes vdev (not tried) as a way to give
udev the heave-ho. It's probably pretty alpha, still.

But is that the right problem to solve?

I've found it trivial to migrate a Debian Jessie system sideways to
OpenRC, which I've found a pleasantly well engineered and somewhat
BSDish relief from AT&T-derived SysV cruft. Works fine - ergo, no
compelling present need for a distro fork even for dedicated systemd
haters, IMVAO. (Views Differ[tm].)

LWN.net's article at the time of the OpenRC package being new to Debian
(but already solid) from three years ago might still merit reading:
http://lwn.net/Articles/512719/

At least some in the Debian community are particularly annoyed by the
systemd team's unwillingness to take patches for portability to kernels
beyond Linux. That led Adam Borowski to jokingly dismiss OpenRC
because it lacks "a hostile upstream". More seriously, Leigh pointed
out that OpenRC uses some of the same features as systemd, but does so
with portability in mind: [...]

--
Cheers, "The trouble ain't that there is too many fools,
Rick Moen but that the lightning ain't distributed right."
***@linuxmafia.com -- Mark Twain
McQ! (4x80)

Rick Moen

2015-08-01 04:42:17 UTC

--
Cheers, "The trouble ain't that there is too many fools,
Rick Moen but that the lightning ain't distributed right."
***@linuxmafia.com -- Mark Twain
McQ! (4x80)

Rick Moen

2016-05-14 20:45:36 UTC

Post by Rick Moen
I've found it trivial to migrate a Debian Jessie system sideways to
OpenRC, which I've found a pleasantly well engineered and somewhat
BSDish relief from AT&T-derived SysV cruft. Works fine - ergo, no
compelling present need for a distro fork even for dedicated systemd
haters, IMVAO. (Views Differ[tm].)
LWN.net's article at the time of the OpenRC package being new to Debian
http://lwn.net/Articles/512719/
At least some in the Debian community are particularly annoyed by the
systemd team's unwillingness to take patches for portability to kernels
beyond Linux. That led Adam Borowski to jokingly dismiss OpenRC
because it lacks "a hostile upstream". More seriously, Leigh pointed
out that OpenRC uses some of the same features as systemd, but does so
with portability in mind: [...]

Completing that task:

I've now completely documented how to run Debian 8 'Jessie' with OpenRC
and _keep_ it on OpenRC without allowing any dependency hairballs to
drag the system onto systemd.
http://linuxmafia.com/kb/Debian/openrc-conversion.html (I have also
added said new page to
http://without-systemd.org/wiki/index.php/Main_Page#Other_useful_sites .)

I paid careful attention to Marc Merlin's point about Debian package
problems because of complex dependency chains, using a little
scripting to create what I believe is a complete list of where problems
will lie -- and some suggestions for how to fix those.

--
Cheers, "My life has a superb cast,
Rick Moen but I cannot figure out the plot."
***@linuxmafia.com -- Ashleigh Brilliant
McQ! (4x80)

Ed Carp

2016-05-14 23:38:49 UTC

I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were - systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

I've now completely documented how to run Debian 8 'Jessie' with OpenRC
and _keep_ it on OpenRC without allowing any dependency hairballs to
drag the system onto systemd.
http://linuxmafia.com/kb/Debian/openrc-conversion.html (I have also
added said new page to
http://without-systemd.org/wiki/index.php/Main_Page#Other_useful_sites .)
I paid careful attention to Marc Merlin's point about Debian package
problems because of complex dependency chains, using a little
scripting to create what I believe is a complete list of where problems
will lie -- and some suggestions for how to fix those.
--
Cheers, "My life has a superb cast,
Rick Moen but I cannot figure out the plot."
McQ! (4x80)
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set
for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2016-05-15 00:14:22 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

You are kidding, right? Have you ever worked at that level? Tried to
manage processes and services in a sane manner? I've been dealing with
that problem since my first paid Linux job in 2000, and it didn't get
finally resolved until systemd. There were lots of things wrong with
how things "were" before, perhaps you never worked down there?

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

And as far as ego goes, do you not take pride in the work you do?
systemd was written by people working for _all_ distros, it was not the
work of just one person at all, to act otherwise is a huge disservice to
those other developers.

Just one word of advice, almost _everyone_ who has worked on a distro at
that level of the "plumbing" has helped out with systemd as it is
somehow obvious to them that this is a much better solution than what
they previously had. I guess you don't trust those developers, which is
fine, but you should seriously consider just why you feel that way.

Post by Ed Carp
No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

Makes sense, why would you want to try to keep a fork from your "parent"
distro? Companies that actually funded their developers are even
smarter than that.

Remember back when Linux users were the ones pushing the boundries of
things, solving real problems and being happy to handle major changes in
the quest to making something better? I sure do. The real question is
why have the others on this list, who have been complaining about a
system library being replaced that they didn't have anything to do with,
have forgotten about why they started using Linux in the first place.

I mean, if you like openrc so much, why not just move back to solaris,
there is an "open" version floating around somewhere that desperately
needs users and developers...

Or just use Gentoo or slackware if you don't like change and wish to
remain in the 1990's.

</get off my lawn>

greg k-h

Ed Carp

2016-05-15 07:35:54 UTC

I'm certainly not going to dignify this personal attack by responding,
except to say that (1) I manage about 1200 RHEL servers, and (2)
systemd certainly has *not* made my life easier, not by a long shot.
My team uses Ansible to deploy and manage these systems, I will say
that Ansible hides a lot of the systemd messiness in the service
module - but it's a completely different way of doing things, and IMO
not for the better.

And if you get off attacking people for expressing their opinion, I
guess you *do* have a problem.

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?
And as far as ego goes, do you not take pride in the work you do?
systemd was written by people working for _all_ distros, it was not the
work of just one person at all, to act otherwise is a huge disservice to
those other developers.
Just one word of advice, almost _everyone_ who has worked on a distro at
that level of the "plumbing" has helped out with systemd as it is
somehow obvious to them that this is a much better solution than what
they previously had. I guess you don't trust those developers, which is
fine, but you should seriously consider just why you feel that way.

Post by Ed Carp
No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

Makes sense, why would you want to try to keep a fork from your "parent"
distro? Companies that actually funded their developers are even
smarter than that.
Remember back when Linux users were the ones pushing the boundries of
things, solving real problems and being happy to handle major changes in
the quest to making something better? I sure do. The real question is
why have the others on this list, who have been complaining about a
system library being replaced that they didn't have anything to do with,
have forgotten about why they started using Linux in the first place.
I mean, if you like openrc so much, why not just move back to solaris,
there is an "open" version floating around somewhere that desperately
needs users and developers...
Or just use Gentoo or slackware if you don't like change and wish to
remain in the 1990's.
</get off my lawn>
greg k-h
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set
for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2016-05-15 18:29:29 UTC

A: No.
Q: Should I include quotations after my reply?

http://daringfireball.net/2007/07/on_top

Post by Ed Carp
I'm certainly not going to dignify this personal attack by responding,

Yet you did :)

And it wasn't "personal" at all, please go re-read it again.

Post by Ed Carp
except to say that (1) I manage about 1200 RHEL servers, and (2)
systemd certainly has *not* made my life easier, not by a long shot.
My team uses Ansible to deploy and manage these systems, I will say
that Ansible hides a lot of the systemd messiness in the service
module - but it's a completely different way of doing things, and IMO
not for the better.

So perhaps your issue is with Ansible and how it handles systemd? How
is that systemd's fault? Have you brought it up with the Ansible
developers? Have they worked to help you resolve your objections?

Post by Ed Carp
And if you get off attacking people for expressing their opinion, I
guess you *do* have a problem.

You said a number of things that aren't true (unix way and all that) and
I protested that statement, not anything specifically about you at all.
You were objecting to the work that I, and people I personally trust,
have done and have given to you and the world for free. You feel that
somehow we are doing it just because we wanted to make _your_ life more
miserable. Did we force you to use systemd? No, the vendor _you_ trust
made that decision, perhaps you shouldn't trust that vendor and move to
a different one. You have choices, no one is _forcing_ you to use
Linux, least of all the developers who write it.

The BSDs are great, they need all the users they can get, and Linux
needs the competition, please move to them if you really like the
old-style init system.

And again people, this is linux-elitists, how "elite" are you if you
complain about having to learn a new tool?

bah humbug,

greg k-h

/dev/rob0

2016-05-16 16:46:27 UTC

Post by Ed Carp
I'm certainly not going to dignify this personal attack by
responding,

Yet you did :)
And it wasn't "personal" at all, please go re-read it again.

+1

Ed, I completely agree with Greg here (about the "attack", not
necessarily about systemd.) Fragile feelings and weak egos on
mailing lists like this make for a toxic environment and lead to
greater difficulties in communication.

Even if it WAS an attack, so what? Be mature and responsible, just
stick to the issues. Whining about personal attacks is every bit as
unbecoming as are the attacks.

Post by Ed Carp
And if you get off attacking people for expressing their opinion,
I guess you *do* have a problem.

You said a number of things that aren't true (unix way and all
that) and I protested that statement, not anything specifically
about you at all.

Indeed, a case of PKB.

Post by Greg KH
You were objecting to the work that I, and people I personally
trust, have done and have given to you and the world for free. You
feel that somehow we are doing it just because we wanted to make
_your_ life more miserable. Did we force you to use systemd? No,
the vendor _you_ trust made that decision, perhaps you shouldn't
trust that vendor and move to a different one.

Nevertheless, systemd is making it difficult for GNU/Linux
distributors to continue without systemd. You [Greg] mentioned
Slackware upthread, and I know that Slackware is feeling the pinch.
I'm sure that's directly related to the unusually long time we're
seeing since the previous Slackware release.

Post by Greg KH
You have choices, no one is _forcing_ you to use Linux, least of
all the developers who write it.
The BSDs are great, they need all the users they can get, and Linux
needs the competition, please move to them if you really like the
old-style init system.

Right, this might be the way forward for us. I like Linux for my
laptops & desktops because of the excellent hardware support, and
mostly ... just because. Inertia. I've been in GNU/Linux a very
long time. I know it well.

Post by Greg KH
And again people, this is linux-elitists, how "elite" are you if
you complain about having to learn a new tool?

Yes, I can learn *BSD if I choose to do it, and I already have some
insight into non-GNU implementations of various tools. I'd rather
stay with Linux, but you're quite right, no one is forcing me to
remain.

That said, I'm not as vehemently opposed to systemd on end-user
machines, so I might stay with Linux laptops and move to a BSD for
other uses.

Also, I can't deny the good points Greg has made, because I'm a mere
admin/enduser. I am fairly familiar with the cruft that comprises
Slackware init scripts, so I am sure that Better Ways exist.

Post by Greg KH
bah humbug,

--
http://rob0.nodns4.us/

Ruben Safir

2016-05-16 18:58:54 UTC

Post by /dev/rob0

Post by Ed Carp
I'm certainly not going to dignify this personal attack by
responding,

Yet you did :)
And it wasn't "personal" at all, please go re-read it again.

Oh BULLSHIT. This is the Linux elitist list where insults are made
with such sublty and sophistication and not even the abuser knows the
real nature of the insult.

Want a Toxic environment, try the IRT at 3PM in August...

Post by /dev/rob0
Even if it WAS an attack, so what? Be mature and responsible, just
stick to the issues.

Really...

We really need a babysitter to tell everyone to play nice.

This generation has spent far to much time sucking on "My Little Pony's"
and watching that purple dinosaur and not enough time playing unsupervised
sports like stickball.

Take two Qualudes and call me in the morning...

Don Marti

2016-05-16 20:26:34 UTC

Post by /dev/rob0
Nevertheless, systemd is making it difficult for GNU/Linux
distributors to continue without systemd. You [Greg] mentioned
Slackware upthread, and I know that Slackware is feeling the pinch.
I'm sure that's directly related to the unusually long time we're
seeing since the previous Slackware release.

It's not systemd making it difficult for distributions
to continue without systemd. It's upstream software
(for a good example, see the "hplip" package in
Rick's article.)

* Package has some random init-related problem.

* Someone writes or contributes a systemd-related
solution.

* Nobody has any real issues with the solution.

* The package comes to depend on systemd.

Hard to argue with, since it means that the package
maintainers are now maintaining one line in a systemd
unit file instead of some shell script, so they can
spend more time on work that solves problems for
their users.

Post by /dev/rob0

Post by Greg KH
And again people, this is linux-elitists, how "elite" are you if
you complain about having to learn a new tool?

So far I haven't seen a good reason to learn 2
different init systems -- systemd for laptops
and basically-stock VMs, and something else for
custom servers. I figure I can work better with 1.1x
systemd skills than with 0.95x skills in 2 different
init systems. (Basically the same reason I stick
with a couple of basic Linux distributions.)

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Ruben Safir

2016-05-16 21:55:59 UTC

It's not systemd making it difficult for distributions
to continue without systemd. It's upstream software
(for a good example, see the "hplip" package in
Rick's article.)
* Package has some random init-related problem.
* Someone writes or contributes a systemd-related
solution.
* Nobody has any real issues with the solution.
* The package comes to depend on systemd.
Hard to argue with, since it means that the package
maintainers are now maintaining one line in a systemd
unit file instead of some shell script, so they can
spend more time on work that solves problems for
their users.

Post by /dev/rob0

Post by Greg KH
And again people, this is linux-elitists, how "elite" are you if
you complain about having to learn a new tool?

Really, that was what SuSE said. They said they hated systemd but the amount of
time they needed to bring new shell scripts up to date with every release
convinced them to change.

and you know what, with all that time they are now producing the worst and
buggiest OS's they ever created.... tah dah.

Everything systemd does is bad and there is NO WAY that systemd can do what
my custom shell scripts do....exect that it can call my shell scripts.

The stated purpose of systemd is make all Linux distros look the same...
that is straight out of Potterings mouth.

I don't support that effort and neither should you,

Post by Don Marti
--
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Greg KH

2016-05-16 23:18:13 UTC

Post by Ruben Safir
The stated purpose of systemd is make all Linux distros look the same...
that is straight out of Potterings mouth.

Not just his, many others, including myself.

And that's a very good thing, if you don't understand the horrid mess
that has lived at that low layer for decades, and is totally different
in different major distros and how it is really hard for developers to
target all of the different ones for their software (think apache
startup scripts, how do you write a single one for all distros?) then
you completely miss the point of why systemd was created.

Post by Ruben Safir
I don't support that effort and neither should you,

I totally support that effort, just look at how much easier it has made
all of the application developer's lives.

And it's freed up your distro developer's time to focus on other more
important things than mucking around in the low levels of things, trying
to package up daemons in a way that works properly for their specific
init.rc layout and logic and helper scripts. They can fix bugs, work on
upstream making things better, and lots of other wonderful things that
they now are doing.

Yes, it's a learning curve when your favorite distro moves over to it,
but be honest, the majority of the layout, structure, and logic, came
from Debian. So much so that in a way, the Debian structure "won" and
is now in all distros. Why wouldn't you want that?

If you want to learn yet-another-way to enable networking on your
distro, great, have fun without systemd, all of the rest of us will just
learn it once, and have that skill transfer to all others automatically.

Have fun with openrc, I've used it for many years, and haven't missed it
at all.

greg k-h

Ruben Safir

2016-05-17 02:32:36 UTC

Post by Ruben Safir
The stated purpose of systemd is make all Linux distros look the same...
that is straight out of Potterings mouth.

Not just his, many others, including myself.

Just his. And what he has created is a total clusterfuck.

Post by Greg KH
And that's a very good thing, if you don't understand the horrid mess
that has lived at that low layer for decades,

It is great and should never be messed with. It didn't need to be made
uniformed, especially is such an imcompantent fashion. there wwas nothing
wrong with it for decades, and it could and will continue for decades more
after systemd is dead and burried.

Post by Greg KH
and is totally different
in different major distros and how it is really hard for developers to
target all of the different ones for their software (think apache
startup scripts,

There is NOTHING wrong with my apache scripts and they have worked for
decades.

Post by Greg KH
how do you write a single one for all distros?)

Seems they've done pretty well with /usr/local/apache/bin/apachectl start.

Or you can start it by hand.

Or you can write your own.

There was no problem here until systemd got involved and made it impossible to
stop apache.

Post by Greg KH
then
you completely miss the point of why systemd was created.

No I didn't Iwas created so that all linux systems, regardless of distro
could be controlled by a central authority, preferably one at redhat.

Post by Ruben Safir
I don't support that effort and neither should you,

I totally support that effort,

Well stop because your killing Linux.

Post by Greg KH
And it's freed up your distro developer's time to focus on other more
important things than mucking around in the low levels of things, trying
to package up daemons in a way that works properly for their specific
init.rc layout and logic and helper scripts. They can fix bugs, work on
upstream making things better, and lots of other wonderful things that
they now are doing.
Yes, it's a learning curve when your favorite distro moves over to it,
but be honest, the majority of the layout, structure, and logic, came
from Debian.

Ummm no. But aside from that, Debian is not a brillantly built distro either.

Post by Greg KH
So much so that in a way, the Debian structure "won" and
is now in all distros. Why wouldn't you want that?
If you want to learn yet-another-way to enable networking on your
distro, great, have fun without systemd, all of the rest of us will just
learn it once, and have that skill transfer to all others automatically.

There is no "all the rest of us" there is just people like you and the
rest of US.

Post by Greg KH
Have fun with openrc, I've used it for many years, and haven't missed it
at all.

That is just a pitty you have abandoned that.

Seriously, you need to dump systemd and us a FREE os.

Post by Greg KH
greg k-h
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2016-05-17 03:26:13 UTC

Post by Greg KH
then
you completely miss the point of why systemd was created.

No I didn't Iwas created so that all linux systems, regardless of distro
could be controlled by a central authority, preferably one at redhat.

There is no "control", and least of all, by Red Hat.

Post by Ruben Safir
I don't support that effort and neither should you,

I totally support that effort,

Well stop because your killing Linux.

_I'm_ killing Linux?

Hah, that's the best thing I've heard all week. Thanks for the laugh,

greg k-h

Teh Entar-Nick

2016-05-17 19:50:04 UTC

Post by Greg KH
in different major distros and how it is really hard for developers
to target all of the different ones for their software (think apache
startup scripts,

There is NOTHING wrong with my apache scripts and they have worked for
decades.

Hold up.

The apache2 init scripts were one of the most embarrassing warts in
Debian for the past decade at least. Unmaintained init script code that
couldn't handle graceful control of apache on a loaded system was part
of why many folks considered this the last straw and jumped ship to a
newer httpd.

I'm sure they worked great for you on your personal server, but for
production workloads they couldn't handle a server that was swamped with
requests. They couldn't reliably *shut down* apache, and thus the
standard `/etc/init.d/<gubble> restart` DWIM invocation stopped becoming
reliable.

I haven't got a horse in the init systems race, but the apache init
scripts themselves were shameful and neglected.

Ruben Safir

2016-05-17 08:17:56 UTC

Post by Greg KH
Not just his, many others, including myself.
And that's a very good thing

And that is the summation of the reason why you are having such a
violent reaction against it from the community.

enough .. nothing can be left to be said.

I will ask why the Kernel Programming Challenge is not
working, or responding, if you have a contact there.
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Greg KH

2016-05-17 14:37:49 UTC

Post by Greg KH
Not just his, many others, including myself.
And that's a very good thing

And that is the summation of the reason why you are having such a
violent reaction against it from the community.

Nope, no reaction anymore, only from the crazies like yourself, everyone
else has realized there is no problem here :)

You have yet to give any _technical_ reason why systemd is bad, ranting
about control and conspiracy things is fun and all, but really, not
something you should actually take seriously. So while this has been a
fun thread to see you go crazy on, I think I'm going to bow out for now
until you post anything based on facts.

Have fun!

greg k-h

Ruben Safir

2016-05-19 14:23:53 UTC

Post by Greg KH
Not just his, many others, including myself.
And that's a very good thing

And that is the summation of the reason why you are having such a
violent reaction against it from the community.

Nope, no reaction anymore, only from the crazies like yourself, everyone
else has realized there is no problem here :)
You have yet to give any _technical_ reason why systemd is bad, ranting
about control and conspiracy things is fun and all, but really, not
something you should actually take seriously. So while this has been a
fun thread to see you go crazy on, I think I'm going to bow out for now
until you post anything based on facts.

A few closing remarks on this. I have given numerous very specfic
technical problems with this, but as usual with systemd proponents, your
ability to discern what you don't want to hear is non-existant.

What you want to here is that there is a bug with this or that, I'm telling
you I don't like the design. That is a technical answer. I'm specifically
telling you what in the design I don't like, which includes its top down control
of the system, its binary journals, its absorption of services which are not
part of the traditional init system, its derooting of X11, its monitoring
of sockets, its nondependency on the shell to make things work,
its interference with gtty processes and the login function, and its absorption and adoption of dbus.

Which part of this technical conversation is not being followed now, and I
can try to answer in more detail, althogh your not listening so I doubt it is
worth it.

Now, addressing the issue of conspriacy theory, I'm frankly disapoint in you,
talk to older and more experienced individuals like this. You owe me an
apology. As someone who has listened to countless conspiracy thearies
about Jews and Banking and Finance and Control, of even control of Linux,
after all GNU is a Jewish organization, and how Israel was a conspirer in
9-11, which I lost many firends in, I am deeply and personally insulted by
this off the wall derogatory comment by you.

If we can not poperly note that Systemd is designed to make all the
linux distros look and act the same, without stating this is a form of
top down control, withouth you snarely about conspiracy theories, then I
under estimated your moral compass. You have none then. It is not I, young
man, who is hopping around on bean bags and oversized rubber balls, dressed
like King Friday from Mr Rogers Neighborhood, promoting the invincibility
of systemd as the universal solution for all things Linux.

~~that was you.

Now to address another area, the need for you to control 5000 computers at the
same time means nothing to me. Applications for systemd for cloud computing
is another thing I couldn't care less about. So stop referencing this as if
it is some benifit, because its not. It mioght be a benifit of a very small
segment of professional sys admins, but that vast rest of us couldn't care less.
That being said, we controlled the entire stock market with the korn shell and
BSD style init, so it defies me what problem you have. On the other hand
I don't care. I do care that a whole pool of system development and control
tools have been removed from my hands with system.

I also mention as a techincal fact that counselkit is SLOW to respond and is
messing up terminal access, forget display settings for X11 etc.

I also mentioned that Systemd make for one large monolithic hacking target
and creates a great security risk. I'm sure Leornard Potterings coding
can never be exploited, because we all see what a great coder he is.

Now one for Nick. I don't use debian. I don't know what they do sometimes.
I see a complity in there start up scripts with is crazy. I suggest you try

/usr/local/apache/bin/apachectl restart

When init scripts get to complex, I tend to just rip them out and rewrite them
to my specific needs.

Ruben

Post by Greg KH
Have fun!
greg k-h
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Don Marti

2016-05-17 04:30:23 UTC

Post by Ruben Safir
Everything systemd does is bad and there is NO WAY that systemd can do what
my custom shell scripts do....exect that it can call my shell scripts.

And call them from a flexible set of places. How would
you hook up a shell script to PowerTOP without it?

http://blog.christophersmart.com/2016/05/08/automatic-power-saving-on-a-linux-laptop-with-powertop-and-systemd/

How would you write a shell script to wake up the
system before running?

https://joeyh.name/blog/entry/a_programmable_alarm_clock_using_systemd/

Post by Ruben Safir
The stated purpose of systemd is make all Linux distros look the same...
that is straight out of Potterings mouth.
I don't support that effort and neither should you,

Why not? Why would I want to reduce the chances that
I can help someone with basic administration tasks on
a random Linux system? If somebody posts about not
being able to ssh in, now I can ask for the output of
"systemctl status sshd" even if they're running some
distribution I don't know (and I only really know
Fedora, Debian, and enough CentOS to install EPEL and
pretend it's Fedora.)

(Are there fewer and fewer active Linux distributions
these days? Active as in they do security updates for
base packages in a less-than-scary amount of time?)

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Ruben Safir

2016-05-17 08:35:22 UTC

Post by Don Marti
Why not? Why would I want to reduce the chances that
I can help someone with basic administration tasks on
a random Linux system?

I do that everyday and doing by hand is educational and useful.

Aside from that, systemd does the oposite. It make it impossible
to alter it.

Post by Don Marti
If somebody posts about not
being able to ssh in, now I can ask for the output of
"systemctl status sshd"

Yeah that was one of the first things that stopped working with systemd

... my ssh tunnel to the panix proxy and to all my kids remote machines
around the country which nicely worked until systemd came to SuSE.

And I started using sshd by compiling it on Slackware ....

Please Don, don't waste my time on this. Its is destroying linux based
systems by making them all conform. Its like waking up one morning to
find all your children have been removed by 6 children who all are
indentical, and none of them any fun.

Systemd is an enslaving system. It is just like windows right down to the
regit system

Post by Don Marti
even if they're running some
distribution

Try Manjaro with openrc.... it works.

Ruben Safir

2016-05-17 08:50:57 UTC

Post by Don Marti
https://joeyh.name/blog/entry/a_programmable_alarm_clock_using_systemd/

cron doesn't work?

Post by Don Marti
Why not? Why would I want to reduce the chances that
I can help someone with basic administration tasks on
a random Linux system?

This is also an example of one of the reasons why systemd has to go.
It is making peoples brains dull to the point where they all sound like
they come from Apple headquarters.

So what are you telling me Don, that if I set up 100 desktop systems
in front of you that you couldn't turn on sshd on them ALL without
breaking a sweat?

Why would you say this?

Don Marti

2016-05-17 14:19:27 UTC

Post by Don Marti
https://joeyh.name/blog/entry/a_programmable_alarm_clock_using_systemd/

cron doesn't work?

On a sleeping Linux laptop? No. If you want to
do it without systemd, you need a separate tool.
Here you go:

http://linux.die.net/man/8/rtcwake

And here's the systemd way to do it:

WakeSystem=true

What's the non-systemd way?

Post by Don Marti
Why not? Why would I want to reduce the chances that
I can help someone with basic administration tasks on
a random Linux system?

This is also an example of one of the reasons why systemd has to go.
It is making peoples brains dull to the point where they all sound like
they come from Apple headquarters.

The company that makes the computers that most of
the creative people I talk to are using, by default?

In the 1990s, when a lot of us learned Linux, the
average home or office computer was pretty crappy.
I got Linux, and I was all like, "Hey, I can multitask
and not have my computer crash! And users of other
OSs can't! Awesome!"

Now what's the point?

Don't all speak up at once -- there is still a lot
of great projects that are possible if you have a
a cross-platform kernel with drivers for everything
from bargain bin devices up to Enterprise Solutions.

But I'm not going to learn a vintage init system
because it enables me to do the stuff I did in
the 1990s.

I'm not dogmatically pro-systemd, and if there's
something useful that you can do with an alternate
init system, I might learn it. Ideas?

Post by Ruben Safir
So what are you telling me Don, that if I set up 100 desktop systems
in front of you that you couldn't turn on sshd on them ALL without
breaking a sweat?
Why would you say this?

I can find and check the configuration file, and
probably Google/cut/paste the right command to start
sshd, but I probably wouldn't trust myself to do all
the tweaky stuff right (is it set to come back up on
reboot, is the right port open, do I have the latest
security update) anywhere but Debian or Fedora.

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

/dev/rob0

2016-05-16 22:08:58 UTC

Post by /dev/rob0
Nevertheless, systemd is making it difficult for GNU/Linux
distributors to continue without systemd. You [Greg] mentioned
Slackware upthread, and I know that Slackware is feeling the
pinch. I'm sure that's directly related to the unusually long
time we're seeing since the previous Slackware release.

It's not systemd making it difficult for distributions
to continue without systemd. It's upstream software
(for a good example, see the "hplip" package in
Rick's article.)

Yes, you are right, and in fact I started thinking about that word
choice right after sending. Thanks for the clarification.

Post by Don Marti
* Package has some random init-related problem.
* Someone writes or contributes a systemd-related
solution.
* Nobody has any real issues with the solution.
* The package comes to depend on systemd.
Hard to argue with, since it means that the package
maintainers are now maintaining one line in a systemd
unit file instead of some shell script, so they can
spend more time on work that solves problems for
their users.

--
http://rob0.nodns4.us/

Ruben Safir

2016-05-15 09:22:24 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

You are kidding, right? Have you ever worked at that level? Tried to
manage processes and services in a sane manner?

Yes and systemd sucks

Post by Greg KH
I've been dealing with
that problem since my first paid Linux job in 2000, and it didn't get
finally resolved until systemd. There were lots of things wrong with
how things "were" before, perhaps you never worked down there?

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?
And as far as ego goes, do you not take pride in the work you do?
systemd was written by people working for _all_ distros, it was not the
work of just one person at all, to act otherwise is a huge disservice to
those other developers.
Just one word of advice, almost _everyone_ who has worked on a distro at
that level of the "plumbing" has helped out with systemd as it is
somehow obvious to them that this is a much better solution than what
they previously had. I guess you don't trust those developers, which is
fine, but you should seriously consider just why you feel that way.

Post by Ed Carp
No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

Makes sense, why would you want to try to keep a fork from your "parent"
distro? Companies that actually funded their developers are even
smarter than that.
Remember back when Linux users were the ones pushing the boundries of
things, solving real problems and being happy to handle major changes in
the quest to making something better? I sure do. The real question is
why have the others on this list, who have been complaining about a
system library being replaced that they didn't have anything to do with,
have forgotten about why they started using Linux in the first place.
I mean, if you like openrc so much, why not just move back to solaris,
there is an "open" version floating around somewhere that desperately
needs users and developers...
Or just use Gentoo or slackware if you don't like change and wish to
remain in the 1990's.
</get off my lawn>
greg k-h
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Ruben Safir

2016-05-15 09:28:35 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

2000 hmmm

You would have loved HPUnix or AIX at SIAC

So it took you that much time to make Linux look and act like Windows. Actually
it is not fair to blame just you for this.

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

Or we can just get rid of systemd

Post by Greg KH
And as far as ego goes, do you not take pride in the work you do?
systemd was written by people working for _all_ distros, it was not the
work of just one person at all, to act otherwise is a huge disservice to
those other developers.

Actually, NOT. Unless Pottering works for all distros now.

Post by Greg KH
Just one word of advice, almost _everyone_ who has worked on a distro at
that level of the "plumbing" has helped out with systemd as it is
somehow obvious to them that this is a much better solution than what
they previously had. I guess you don't trust those developers, which is
fine, but you should seriously consider just why you feel that way.

Post by Ed Carp
No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

Ruben Safir

2016-05-15 09:57:35 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

2000 hmmm
You would have loved HPUnix or AIX at SIAC
So it took you that much time to make Linux look and act like Windows. Actually
it is not fair to blame just you for this.

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

Or we can just get rid of systemd

Actually, NOT. Unless Pottering works for all distros now.

I know why, because systemd is obtuse, turned my system upside down and obtuse.

The very idea that this has been brought from the Gods on high and therefor
needs to be loved is LAUGHABLE. Free Software is not a corporation and YES
I hate the New Coke.

Maybe it is time for ***YOU*** to ask why this keeps coming up again and again.

The answer is because it is so alien to anything done in Free Software
previously, it is so stifling, and so centralized, that is it HATED and
being VIOLENTLY rejected...

Anyway, Manjaro with Openrc, is the best of the bunch as I can tell.
they guy maintaining it seems to have a good handle on things.

https://sourceforge.net/projects/manjaro-openrc/

Post by Ed Carp
No matter - hooray for Rick! Thank you! I hope I don't have to worry
about this when Linux Mint goes systemd in the next release
(reportedly)...

Try Manjaro with openrc
I'm just gonna just bypass the rest of this ranting
It is always good to have a single vector to control and entire system
What will happen when Watson takes over and exploits systemd and takes over
all those android phones....using big data from the clouds...
Now I'm going back to listening to the Velvet Unground...
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2016-05-15 18:32:15 UTC

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

Or we can just get rid of systemd

What specifically are you not liking about it?

Actually, NOT. Unless Pottering works for all distros now.

Do you think that somehow only 1 person did all of this work? Come on
now, that's looney. Look at the author list of systemd, it included
developers from _all_ distros.

Post by Ruben Safir
What will happen when Watson takes over and exploits systemd and takes over
all those android phones....using big data from the clouds...

systemd is not on your android phone, yet, give me time :)

Ruben Safir

2016-05-16 19:03:41 UTC

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

Or we can just get rid of systemd

What specifically are you not liking about it?

Actually, NOT. Unless Pottering works for all distros now.

Do you think that somehow only 1 person did all of this work? Come on
now, that's looney. Look at the author list of systemd, it included
developers from _all_ distros.

No I don't he did like what? 80% of it?

Post by Ruben Safir
What will happen when Watson takes over and exploits systemd and takes over
all those android phones....using big data from the clouds...

systemd is not on your android phone, yet, give me time :)

THAT I have no doubt, and THAT IS EXACTLY the attitude problem towards this
product that I'm worried about.

I love you Greg but you can shove systemd where the sun don't shine.

Love Truly

Ruben

Post by Greg KH
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2016-05-16 19:39:58 UTC

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

Or we can just get rid of systemd

What specifically are you not liking about it?

Actually, NOT. Unless Pottering works for all distros now.

Do you think that somehow only 1 person did all of this work? Come on
now, that's looney. Look at the author list of systemd, it included
developers from _all_ distros.

No I don't he did like what? 80% of it?

Nope.

How many different people do you think contributed to systemd, without
going and looking at the source repo? 10? 20? 30? I'll let you guess
to see how close you are...

Post by Ruben Safir
What will happen when Watson takes over and exploits systemd and takes over
all those android phones....using big data from the clouds...

systemd is not on your android phone, yet, give me time :)

THAT I have no doubt, and THAT IS EXACTLY the attitude problem towards this
product that I'm worried about.
I love you Greg but you can shove systemd where the sun don't shine.

Great, the sun doesn't shine inside my phone's case, which is where this
is going to go :)

And really, why would you care? When was the last time you had to
handle process management on your phone? When was the last time you
touched the init system of your phone? And if you did, what was the
reaction you had when you did so (hint, if you have touched that code,
and you like it, well, then you really want something older than what
the BSDs even provide...)

I assume you also hate the init system in an Apple phone for the same
reason you hate systemd? I guess we all need our windmills to tilt
against, have fun!

greg k-h

Ruben Safir

2016-05-16 22:02:15 UTC

Post by Ruben Safir
THAT I have no doubt, and THAT IS EXACTLY the attitude problem towards this
product that I'm worried about.
I love you Greg but you can shove systemd where the sun don't shine.

Great, the sun doesn't shine inside my phone's case, which is where this
is going to go :)
And really, why would you care?

The fact that you ask that question means either your trolling me
or the most ignorant programming genius to walk the earth.

Post by Greg KH
When was the last time you had to
handle process management on your phone?

My phone is a land line, FWIW.

Post by Greg KH
When was the last time you
touched the init system of your phone?

Can you get to it? I doubt it without hacking through it.

Post by Greg KH
And if you did, what was the
reaction you had when you did so (hint, if you have touched that code,
and you like it, well, then you really want something older than what
the BSDs even provide...)
I assume you also hate the init system in an Apple phone for the same
reason you hate systemd?

I don't touch anything Apple for the resons I don't touch systemd.

Post by Greg KH
I guess we all need our windmills to tilt
against, have fun!

Manjaro and openrc.

Less attitute.... more and freer software ...faster and more secure computers.

Ruben
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Ruben Safir

2016-05-16 22:05:19 UTC

Post by Greg KH
Nope.
How many different people do you think contributed to systemd, without
going and looking at the source repo? 10? 20? 30? I'll let you guess
to see how close you are...

No matter how you bend it, systemd is controlled lock stock and barrel
by Pottering and his employer.
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Don Marti

2016-05-15 17:18:13 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

I have written both init scripts and systemd unit
files, for server-side packages of comparable
complexity. From the point of view of the package
maintainer, systemd does have an edge. You're writing
fewer total lines and telling a well-documented
program what to do, not invoking shell functions that
can run arbitrary stuff.

I also run a "mainstream" desktop distribution (with
systemd) and the thing starts up faster than my Linux
laptops used to. Hard to break out how much of this
is systemd socket activation, though. (this is where
someone links to the obvious LWN article I should
have cited.)

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

I don't get the "Unix Philosophy" thing.

Posts about "this functionality is broken on systemd
but not OpenRC" are great, but "Unix Philosophy"?
What did that get us? A bunch of failing vendors in
the 1990s, and the inevitability of Windows NT.

http://gabriellacoleman.org/blog/?p=1729

The "philosophers" of Unix let themselves be rounded
up and made irrelevant.

If you want to build a better init system, build a
better init system. But philosophy-based OS advocacy
is a failure. Designing working software based on
philsophy is like writing real network software based
on the OSI 7-layer burrito model.

(Nothing against Rick's article.
http://linuxmafia.com/kb/Debian/openrc-conversion.html
I enjoyed reading it. No philosophy, which I
appreciate.)

Post by Greg KH
Remember back when Linux users were the ones pushing the boundries of
things, solving real problems and being happy to handle major changes in
the quest to making something better? I sure do. The real question is
why have the others on this list, who have been complaining about a
system library being replaced that they didn't have anything to do with,
have forgotten about why they started using Linux in the first place.

One cool systemd feature that I want to try is
multiseat:

https://bluehatrecord.wordpress.com/2015/01/15/multiseat-configuration-in-fedora-21/

Debian is one of the distributions I regularly
use, so I wouldn't mind setting up a system with
an alternate init system if there is something I
want to do that works with OpenRC but not systemd.
Right now I don't see a compelling reason not to use
the "out of the box" init system.

Post by Greg KH
I mean, if you like openrc so much, why not just move back to solaris,
there is an "open" version floating around somewhere that desperately
needs users and developers...
Or just use Gentoo or slackware if you don't like change and wish to
remain in the 1990's.

Or if you want to go the other direction, Plan 9 or GNU Hurd.

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Ruben Safir

2016-05-15 23:42:00 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

That is just NOT true. The shell does what it is told...nothing arbitrary.
It is flexible and it DOES exactly what I want without having to use
systemd.

Post by Don Marti
I also run a "mainstream" desktop distribution (with
systemd) and the thing starts up faster than my Linux
laptops used to. Hard to break out how much of this
is systemd socket activation, though. (this is where
someone links to the obvious LWN article I should
have cited.)

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

I don't get the "Unix Philosophy" thing.

yes you do...

Post by Don Marti
Posts about "this functionality is broken on systemd
but not OpenRC" are great, but "Unix Philosophy"?
What did that get us? A bunch of failing vendors in
the 1990s, and the inevitability of Windows NT.
http://gabriellacoleman.org/blog/?p=1729
The "philosophers" of Unix let themselves be rounded
up and made irrelevant.
If you want to build a better init system, build a
better init system.

systemd is __NOT__ an init system
so lets stop pretending that we are comparing apples here when
we are comparing apples and windows...

<snip>

Umm, no you don't Linux users where forging social change

There has never been a technological alteration of the OS as signifant as the installation
of systemd as the mommy application between me and my system. It's placement into the
middle of the OS allow makes it a complete non-starter.

Post by Don Marti
The real question is

Post by Greg KH
why have the others on this list, who have been complaining about a
system library being replaced that they didn't have anything to do with,
have forgotten about why they started using Linux in the first place.

Actually, there were compalints about REAL innovations not moving fast enough and
the implementation of threads and forks being merged. And there were intense conversations
about Apaches adoption of threads and Java being foisted on the application layers and so on.

This is not even a real innovation. In fact it will kill innovation as it absorbs
one independent project after the next, with no end in sight.

The reason why systemd is so HATED is because it breaks OS's paridign by putting a single
application incharge of nearly the entire OS, which is why I call it a Kernel Wrapper..
because it is.

It has eliminated getty processes and emulates them, very slowly I might add, for sentimental
reasons, it monitors processes and refuses killall -9 signals, it restructured init traditional
systems and threw a bone to shell script writers, it absorbed login, X11, pts, removed plain english logs
with an easily corruptable single binary, it has decreased overall performance it a crawl, etc etc etc.

And it is gonna be attached... just watch. It is gonna be one big fat ugly hackers target and
there will be nobody to patch it.

It has turned Linux into a windows clone and I have NO INTEREST.

This is not a theorectical conversaiton. I've been dealing with the horrors of ALL of systemd
for 3 years, and in about 14 days - it is finally being dispacts, happely for Manjoro of openrc.

I'm not even interested in debating this any longer. People advocating for systemd are NUTS.
There is no point in arguing with them and I'm not interested, any longer, in anyones independent
accesement on its "advantages". Its advantages are disadvantaged, big bother top down designs.
This is a huge step backwards. The revolution of Linux that you are talking about was decenitralization,
collaberation, indpendence, community, and cooperation. These are not in the vocabulary of systemd's
design or its creators ambitions.

When I see systemd it makes me depressed. I feel like I've now lived to my 50's and I've worked for
30 years trying to empower individuals through technology... but with the dual forks of Android
and systemd, I feel I haven't made even the slightest dent. We are charging into a very bleak and
dark future and there seems to be nothing I can do drive history in a different direction. I
underestimated the danger of young people not having experience of previous generations and spending
their short lifetimesi, instead of learning from their parents, to surrounded completely by a corperate
marketing plan that has the time and patience to span generations.

Your like, "Hey your a slave like that, and you have lost control of your life, " and they are like, i
"Yeah but I am a well fed slave I get free and limitless porn, and an HIV prevent drug, and I got 50%
off this neat Coach bag that I can show everone on my social networking feed that Watson is listening in on".

Post by Don Marti
One cool systemd feature that I want to try is

Yeah well maybe after I can coax it to give me a terminal quickly and run X11 on a second display
and kill apache when I SAY SO. Also its journal grows too quickly slowing everything down, probably because
everything is now dumped into that one binary...

That just kills my apache debugging, and has made me restart my server more times in the last 6 months than
they previous decade....

Multiseat, and all those years the kids were playing crossfire off the main server upstairs...they were all
"single seat"

<snip>
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Don Marti

2016-05-16 00:04:35 UTC

Post by Ruben Safir
This is not a theorectical conversaiton. I've been dealing with the horrors of ALL of systemd
for 3 years, and in about 14 days - it is finally being dispacts, happely for Manjoro of openrc.

Three years of problems that affect the
distribution(s) you use? Can you share some links to
bug reports you have filed, or mailing list questions
you have asked?

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Ruben Safir

2016-05-16 00:51:28 UTC

Three years of problems that affect the
distribution(s) you use? Can you share some links to
bug reports you have filed, or mailing list questions
you have asked?

Nope, not interested. I just want to get off it as fast as i can.
I don't want it fixed, and I want it and the entire pardign it created
gone...

blotted from existence would be best.

Like the Mouse said, you can keep the cheese just let me out of the trap.

Jim Thompson

2016-05-17 04:41:13 UTC

Post by Ruben Safir
The shell does what it is told...nothing arbitrary.

You're delusional.

Ruben Safir

2016-05-17 08:37:29 UTC

Post by Jim Thompson

Post by Ruben Safir
The shell does what it is told...nothing arbitrary.

You're delusional.

No it does something OTHER than what you program it to do because there is a
a secret AI intercepting your shell scripts before they can execute.

But for the rest of us, we have no such gremblins in out computers.
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Ruben Safir

2016-05-17 08:41:11 UTC

Post by Jim Thompson

Post by Ruben Safir
The shell does what it is told...nothing arbitrary.

You're delusional.

BTW - If I recall correctly you presented that video on systemd that
sealed the deal for me to never use it and that it could never be fixed.

You gave the most inacurate presentation I can ever remember
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Tomasz Rola

2016-05-23 21:51:53 UTC

Post by Ed Carp
I'm particularly annoyed by systemd, period. Nothing wrong with the
way things were

Oh. I am not sure if this is good idea to chime in, but this sorry
thread is itching me and I somehow cannot resist. Sorry, everybody.

So you are happy with systemd, fine. And you claim it is
great. Cool. Myself, I could not care less if it is good or bad, given
that I have no technical need it could help me to with. But I have to
care, because judging how things seem to be I expect there will be
growing pressure to lock Linux in. It is possible - for a while - to
have init processes as they used to be, but this is not going to
last. Gentoo and Slackware will have to adopt systemd too, sooner or
later. And everybody who is going to run Linux kernel will have to run
systemd.

At least this is how I see it.

But IMHO it does not make you right, only self rightuous. You are the
one who asked in earlier post what is a non-systemd way of turning
Linux laptop into an alarm clock. And I would say, the non-systemd way
is to ask a question first, like "why on Earth would I want to do such
thing".

Ruben Safir, if you are reading this: it looks to me that you are
making an error too. Linux is already lost. Systemd has been adopted
into it. There were many people who made such decision, they probably
had a good reason to force it their way, majority had no reason to
oppose, and you are wasting your time. No amount of talking will undo
this change, because there were strong players (I do not know names,
does it matter?) who wanted this change to happen and so it
happened. Either accept the facts and love systemd or accept the facts
and move to another platform, where hopefully majority is still
unspoiled. FreeBSD seems to me like a good choice and if not, I can
either improve it a bit or move on, whereas I have not much
interest in improving systemd. There is still some time (medicine
calls it "grace period") before I will have to run systemd with no opt
out. The last two or so system upgrades (Debian) forced me to chase
and kill some unwanted demons (they trashed my logs with garbage,
fucked with some software and refused to die and maybe some more). The
writing was on the wall for years, I should have seen it earlier.

[...]

Post by Ed Carp
- systemd goes completely against the Unix idea of one
small utility doing one thing well, and it just comes across as change
for the sake of change as a monument to someone's ego.

Really, you want to drag this old disputed argument out again? Tell me
how your old init system was a "do one thing well" program?

I don't get the "Unix Philosophy" thing.

I do not care about "philosophy" all that much, but thing either does
a job, or can be improved or it does not do the job *I* want it to do
and should be replaced (and in such case, little does it matter how
many other things it does because I do not need them).

Windows NT, being touted as UNIX killer by marketing reps? And
silently discouraged by sales reps (because it was not doing the
advertised job). The one who required separate cpu for each service
ran (like, you wanted ftp, http and mail - buy 4-cpu machine or spend
rest of your life trying to log in, but even many cpu could not help
much). Windows "Not Today"?

So this is what inevitable things look like.

Post by Don Marti
http://gabriellacoleman.org/blog/?p=1729
The "philosophers" of Unix let themselves be rounded
up and made irrelevant.

There are many kinds of philosophers.

[...]

One cool systemd feature that I want to try is
https://bluehatrecord.wordpress.com/2015/01/15/multiseat-configuration-in-fedora-21/

Yes? If I am right, this was being done with X-server since 1980-ties
and with text terminals since about half a century. Wake me up when
you can do "multiseat" across a continent.
--
Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature. **
** As the answer, master did "rm -rif" on the programmer's home **
** directory. And then the C programmer became enlightened... **
** **
** Tomasz Rola mailto:***@bigfoot.com **

Greg KH

2016-05-23 23:18:17 UTC

Post by Tomasz Rola
So you are happy with systemd, fine. And you claim it is
great. Cool. Myself, I could not care less if it is good or bad, given
that I have no technical need it could help me to with. But I have to
care, because judging how things seem to be I expect there will be
growing pressure to lock Linux in. It is possible - for a while - to
have init processes as they used to be, but this is not going to
last. Gentoo and Slackware will have to adopt systemd too, sooner or
later. And everybody who is going to run Linux kernel will have to run
systemd.
At least this is how I see it.

But what's wrong with that? Before you were all running a fork of init
scripts, all hacked up in odd ways that were not compatible across any
distro. What has changed here?

Would you rather run something as pid-1 that only a few developers have
ever looked at (i.e. your specific fork), or one that hundreds of
developers have looked at and are maintaining?

thanks,

greg k-h

Tomasz Rola

2016-05-24 00:37:49 UTC

But what's wrong with that? Before you were all running a fork of init
scripts, all hacked up in odd ways that were not compatible across any
distro. What has changed here?

This kind of argument has disputable value. I expect there will
"always" be an OS or flavour of some with "fork of init scripts" and
they will choose to not run systemd or they will choose to give it as
an option. And if Apache folks wanted to run theirs on such OS, they
will have to somehow support such scripts anyway. And so on. So I do
not see systemd as anything close to solving a problem. The problem is
there, granted, and it would be nice to have a solution but I would
like to have few options, rather than being told I have to use systemd
in order to use Apache. In such case, I will go with something that
does not depend on systemd, just like I go now with things that do not
depend on Windows. Because once you start depending on Windows, you
are in a prison. Once you start depending on Java, you are in another
prison. And so on. Once you start depending on Linux... you cannot
switch to FreeBSD overnight and this makes life a little bit
complicated and interesting.

Post by Greg KH
Would you rather run something as pid-1 that only a few developers have
ever looked at (i.e. your specific fork), or one that hundreds of
developers have looked at and are maintaining?

I would rather run something that I find acceptable. So far I have
read too much critique of systemd to just install it without
objection and further inspection.

So I would not welcome systemd on my computer, but I am fine with
others happily running it (if this is what they want). But then there
is my suspicion there is going to be no choice. Right now it is
transition time, but like I wrote pressure etc will do the job.

And it would be wrong if things went that way. See, I appreciate that
people can use software they choose. You can edit your code with
whatever editor you like and I am cool about it as long as I can edit
mine with Emacs. If there was Linuxwide adoption of some, say,
Sublimetext and in such way that I could expect more and more
obstacles with compiling & installing my copy of Emacs, this would not
make me happy. And it does not matter if Sublimetext is better
editor. In this fictional case, Emacs is a bit more than just editor
and Sublimetext is not going to replace all of it.

Likewise, I am surprised that people are using Unity (tried it, not
liked it) but again, I see nothing wrong about it, but myself, I run
fvwm. I was using kde and old gnome for few years, but once they
started to misbehave, I went back to fvwm. Old configs almost worked
after 15 years, I just had to do few edits (the years do not add up
because I was also long time afterstep user, but they screwed it and I
never looked back).

It seems to be so much different with systemd. There is a number of
people who dislike it for various reasons, and I am willing to pay
attention. Yet one distro after another, systemd is being chosen as
the right thing to do. Honestly, if this was the case with Emacs, if
you were told this is going to be *the editor* for every Linux from
now on, would you feel it as something good? Even I would feel uneasy
about it, because I have enough reason to understand that Emacs is not
for everybody, and even if it was, everybody should be given a choice
if and when to start using it. And I still keep using Vim quite often,
even learn doing new things with it, because Emacs is not for every
use case either.
--
Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature. **
** As the answer, master did "rm -rif" on the programmer's home **
** directory. And then the C programmer became enlightened... **
** **
** Tomasz Rola mailto:***@bigfoot.com **

Rik van Riel

2016-05-24 01:19:00 UTC

Â
But what's wrong with that?Â Â Before you were all running a fork of
init
scripts, all hacked up in odd ways that were not compatible across any
distro.Â Â What has changed here?

The "init scripts are different on different systems"
problem may not be solved by systemd (though they will
look the same on more systems), but systemd does solve
a number of actual problems.

For one, it has a reliable way of detecting whether
daemons are still running, more so than the old style
shell scripts ever could. Â It relies on cgroups to
do this.

Secondly, it can use the same cgroups code to set up
containers, and make sure the software inside the
containers is running. The old shell scripts did not
have a way to start things in containers at all.

--
All Rights Reversed.

Tomasz Rola

2016-05-24 13:24:45 UTC

Post by Rik van Riel

Post by Greg KH
But what's wrong with that? Before you were all running a fork of init
scripts, all hacked up in odd ways that were not compatible across any
distro. What has changed here?

The "init scripts are different on different systems"
problem may not be solved by systemd (though they will
look the same on more systems), but systemd does solve
a number of actual problems.

I think there is intetesting pattern among systemd supporters:

1. strong dislike of scripting, in particular shell scripting, as it seems

2. fascination with short boot time ("in 12 seconds" - nice, mine
boots in 120 seconds, yet I only do it once a day and speeding up may
look attractive, but a cost is prohibitive, so for a while I will keep
doing it my way - if it ever becomes a problem, I will just switch to
booting once a month, thus I will have average daily boot time below
4s)

Post by Rik van Riel
For one, it has a reliable way of detecting whether
daemons are still running, more so than the old style
shell scripts ever could. It relies on cgroups to
do this.

Is it so? Better than this:

=> (627 2): cp ~/bin/ice.sh ~/bin/demonumb.sh

=> (627 3): gvim ~/bin/demonumb.sh

=> (627 11): cat ~/bin/demonumb.sh
#!/bin/sh

ps ax | grep -iE "$1" | grep -ivE 'grep|demonumb' | wc -l

Post by Rik van Riel
(627 8): demonumb.sh boa

1

=> (627 9): demonumb.sh 'bo*'
114

=> (627 10): demonumb.sh 'bo.*'
3

=> (627 12): demonumb.sh bind
1

=> (627 13): demonumb.sh lpd
0

My "script" can see some improvement in a future. Right now it is just
a oneliner in a file, maybe 2 minutes of work, maybe 5 (gradually
getting better). Out of curiosity, in what way systemd improves upon
my script? I am not going to use systemd any time soon, so if it does
it better and I could emulate it in a script without too much work,
why not.

OTOH, if I can get 80% of functionality with one liner, next 10% with
20 lines and the remaining 10% would require that I rebuild the whole
way I use my system, then this last 10% is probably not worth the
effort.

Post by Rik van Riel
Secondly, it can use the same cgroups code to set up
containers, and make sure the software inside the
containers is running. The old shell scripts did not
have a way to start things in containers at all.

After reading a bit:

http://serverfault.com/questions/560206/how-to-find-out-cgroup-of-a-particular-process

http://hydra.geht.net/tino/english/faq/debian/squeeze/cgroups/

I believe I can do a lot with cgroups using just scripting and
properly abusing /proc interface. Whether I can do the same or more
then with systemd is not so important as long as I can do whatever is
needed.

Mind you, scripting is not just /bin/sh or /bin/bash, it could as well
be /usr/bin/wish or /usr/bin/awk. But if you insisted on using proper
shell for it, than /usr/bin/zsh comes to help with tons of builtins
and almost 1.5 megabyte of man page:

=> (895 5): man zshall | wc
26408 181799 1394826

Still, cannot see how systemd would make *my* life so much better that
I would want to abandon whatever I am using right now.

As of containers, seems like docker works on FreeBSD and Windows, so
perhaps I could go without needing systemd, again. They may also have
something else for FreeBSD, which I am yet to learn about.
--
Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature. **
** As the answer, master did "rm -rif" on the programmer's home **
** directory. And then the C programmer became enlightened... **
** **
** Tomasz Rola mailto:***@bigfoot.com **

Rik van Riel

2016-05-24 14:56:04 UTC

Post by Rik van Riel
Â

1. strong dislike of scripting, in particular shell scripting, as it seems
2. fascination with short boot time

I did not mention either of those, nor do I really
care about them.

Post by Rik van Riel
For one, it has a reliable way of detecting whether
daemons are still running, more so than the old style
shell scripts ever could.Â Â It relies on cgroups to
do this.

Out of curiosity, in what way systemd improves upon
my script?

Systemd can manage multiple instances of the same
daemon, which is quite common especially when used
in conjunction with containers.

Post by Tomasz Rola
OTOH, if I can get 80% of functionality with one liner, next 10% with
20 lines and the remaining 10% would require that I rebuild the whole
way I use my system, then this last 10% is probably not worth the
effort.

Not worth it for you, probably.

However, when a distro has hundreds of thousands,
or even millions of users, there will be significant
number of users who do rely on the last 20% of
functionality.

Their needs are no less legitimate than yours.

Post by Tomasz Rola
Still, cannot see how systemd would make *my* life so much better that
I would want to abandon whatever I am using right now.

Nobody seems to be telling you to stop abandoning
your scripts.

--
All rights reversed

Tomasz Rola

2016-05-24 23:00:54 UTC

[...]

Post by Rik van Riel

Not worth it for you, probably.
However, when a distro has hundreds of thousands,
or even millions of users, there will be significant
number of users who do rely on the last 20% of
functionality.
Their needs are no less legitimate than yours.

Certainly.

Somehow this thread is sliding into minutiae. Perhaps my
fault. Anyway, my original attempt was to hint that percentage of
users who are going to enjoy systemd's specific functionality is
rather small and with such assumption (which could be wrong) the
universal adoption of systemd looks strange to me.

Or rather: certain group would welcome systemd and some other group
would reject it. Both groups seem to be (small, really small)
minorities to me. The majority was probably not interested or maybe
even did not notice the change. I belong neither to this majority nor
to the welcoming crowd, but instead of rejecting I prefer to
investigate.

As I have written already in one previous mail, any discussion of the
subject is not going to revert systemd's adoption. It may look a bit
like Linux has been hijacked, but since so many decision makers went
for it, I think they should have it. So far no argument convinced me
enough to join the welcoming crowd, but investigation has been fun (it
usually is).

BTW, someone tried to show me he sent request in my name to
unsubscribe me (knowing it was not going to succeed but I would
receive notice from mail server). The usual ip from France involved,
92.103.69.51. Assuming same person in all cases, he not only targets
supporters of systemd but also investigators :-). I am thrilled.
--
Regards,
Tomasz Rola

--
** A C programmer asked whether computer had Buddha's nature. **
** As the answer, master did "rm -rif" on the programmer's home **
** directory. And then the C programmer became enlightened... **
** **
** Tomasz Rola mailto:***@bigfoot.com **

Don Marti

2016-05-26 04:52:00 UTC

Post by Tomasz Rola
Somehow this thread is sliding into minutiae. Perhaps my
fault. Anyway, my original attempt was to hint that percentage of
users who are going to enjoy systemd's specific functionality is
rather small and with such assumption (which could be wrong) the
universal adoption of systemd looks strange to me.

Think about it not from the point of users, but from
the point of view of developers of server software.
Does systemd mean that they can spend more time
on application-specific functionality and less on
boilerplate service management code? You're not just
saving lines of shell in an init script.

http://0pointer.de./public/systemd-man/daemon.html#New-Style%20Daemons

Post by Tomasz Rola
As I have written already in one previous mail, any discussion of the
subject is not going to revert systemd's adoption. It may look a bit
like Linux has been hijacked, but since so many decision makers went
for it, I think they should have it. So far no argument convinced me
enough to join the welcoming crowd, but investigation has been fun (it
usually is).

It's more of a do-ocracy here, with the decisions
being made by each individual package that either goes
systemd-dependent or maintains support for running
under other init systems.

The case for supporting non-systemd init systems
is going to have to be made one package at a time.
(I don't see a reason to do this, because systemd
is under LGPL, which is a perfectly fine license.
It's not like making Flash-free versions of web sites,
or podcasts in Ogg Vorbis (when does that MP3 patent
really expire anyway?) or all that stuff.)

Post by Tomasz Rola
BTW, someone tried to show me he sent request in my name to
unsubscribe me (knowing it was not going to succeed but I would
receive notice from mail server). The usual ip from France involved,
92.103.69.51. Assuming same person in all cases, he not only targets
supporters of systemd but also investigators :-). I am thrilled.

Probably just going after list members in general.
Anyone else dealt with this problem on a Mailman list?

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Ruben Safir

2016-05-25 03:57:26 UTC

Post by Rik van Riel
Systemd can manage multiple instances of the same
daemon, which is quite common especially when used
in conjunction with containers.

No it is not common, not common at all, aside which, daemons
should be controlled by there parent...

other designs would be broken.

Ruben Safir

2016-05-25 04:14:07 UTC

Post by Rik van Riel
The "init scripts are different on different systems"
problem may not be solved by systemd (though they will
look the same on more systems), but systemd does solve
a number of actual problems.

What you fail to see is that there is no problem that inti
scripts look different or even act differently on different
systems. That is an __Advantage__.

Conformity sucks when it is too rigid.
--
So many immigrant groups have swept through our town
that Brooklyn, like Atlantis, reaches mythological
proportions in the mind of the world - RI Safir 1998
http://www.mrbrklyn.com

DRM is THEFT - We are the STAKEHOLDERS - RI Safir 2002
http://www.nylxs.com - Leadership Development in Free Software
http://www2.mrbrklyn.com/resources - Unpublished Archive
http://www.coinhangout.com - coins!
http://www.brooklyn-living.com

Being so tracked is for FARM ANIMALS and and extermination camps,
but incompatible with living as a free human being. -RI Safir 2013

Ed Carp

2016-05-25 06:10:34 UTC

Disclaimer: I've been writing software for 40 years. I started out
with Linux in late 1991 or early 1992 as one for the first developers
working on Linux. I'm currently working in a large (1200+ servers)
shop running pretty much 100% RHEL.

The problem I have with this entire discussion is that Linux has
always stood for choice. Disregarding GregH's childish rants and
name-calling (which is easy to do), we are swiftly heading into a
world where systemd is forced upon the Linux world, with no other
choice, no other alternative. If I want to stick with init.d, I will
soon have no choice - major players like RedHat have already moved to
systemd in RHEL7, so if I want a job in a RHEL shop, I have to know
systemd, and the rest of the Linux world (or a lot of it) is going
down that path also.

Is systemd better or worse than what has come before? That's
completely beside the point. In a larger sense, it's not even about
systemd. What IS the point is that the freedom of choice is
disappearing, pushed by a bunch of loudmouth control freaks who want
to force the Linux world into doing things their way - and if we don't
agree, trying to browbeat us into submission with insults,
name-calling and clever put-downs. Really? Is that all you have to
offer? In fact, I wouldn't mind seeing systemd/init.d as an
install-time choice, as long as systemd keeps it's damned code out of
the kernel (where it doesn't belong anyway), but the systemd fanboys
are arrogantly pushing the "my way or the highway" route. For those,
see the link to the image at the end of this email.

So much for freedom of choice.

Loading Image...

Greg KH

2016-05-25 06:37:35 UTC

Post by Ed Carp
The problem I have with this entire discussion is that Linux has
always stood for choice.

Never been true, sorry:
http://www.islinuxaboutchoice.com/

Ed Carp

2016-05-25 07:01:14 UTC

Post by Ed Carp
The problem I have with this entire discussion is that Linux has
always stood for choice.

http://www.islinuxaboutchoice.com/

The problem with this argument is that it's simply not true. For
example, if you don't like GNOME, you have a huge number of choices,
all doing the same thing in different ways, offering different
features. I can switch between window managers easily, and if I use
XFCE, it doesn't mean that everyone else either uses XFCE or is some
sort of Luddite idiot, and it doesn't mean that there are multiple
points of failure, because they're all independent of each other by
design

But systemd doesn't afford one that choice, and that's my point. Even
though systemd was supposedly designed as a "drop-in replacement" for
init.d and the like, can it be pulled out and OpenRC or something else
be dropped in? Can I switch as easily as I can switch window managers?

Greg KH

2016-05-26 04:47:03 UTC

Post by Ed Carp
But systemd doesn't afford one that choice, and that's my point. Even
though systemd was supposedly designed as a "drop-in replacement" for
init.d and the like, can it be pulled out and OpenRC or something else
be dropped in? Can I switch as easily as I can switch window managers?

So you want a distro to provide you the choice of different init
systems, that's different. And a really difficult thing to do, so much
so that almost no one wants to try to maintain such a mess. Gentoo does
this right now, and it's a valiant effort to keep it all working
properly.

But you can't tell someone else what to create, or work on, so if the
distro you know and love decides to change, well, either you need to
accept that, or change distros. That's _your_ choice to make, the
distro has to make their own choice in what they will support, they
can't just offer up all choices, that makes a total mess.

Or, even better, you can make the choice that the Subject: here asks
for, fork the distro and do your own thing. That's why we have so many
distros, because people want to do different things. No one is stopping
anyone from forking Jessie and maintaining the old init systems. The
fact that almost no one is doing that work, might be a sign that maybe
it's not anything that anyone wants to do :)

Have fun!

greg k-h

D. Joe

2016-05-25 12:30:52 UTC

Post by Ed Carp
The problem I have with this entire discussion is that Linux has
always stood for choice.

http://www.islinuxaboutchoice.com/

Quite right. Linux is just a kernel. It isn't about choice. So glad to see
we're keeping it real, here.

*Computing* is fundamentally about choice, though: Read a bit, choose
whether and which bit to write. Repeat...until the entire tapestry of our
increasingly computing-controlled world has been woven.

The extent to which Linux is about computing, now, that's the tricky
question.

Maybe it's about joining the effort to sell a fantasy, in which a service or
appliance does exactly and only what we want, when we want it, how we want
it, pay no attention to the Turing machine behind the curtain. (Who "we"
are also left as an exercise for the reader.)

We must also ask ourselves these fundamental and obviously intrinsically
interlinked questions: "Do you ... hate handicap people? Do you hate
people who do not speak English? Do you only care about your own use-case
for GDM?"

But I see we've already been drawing from that question pool, so, carry on.

--
Joe On ceding power to tech companies: http://xkcd.com/1118/
man screen | grep -A1 weird
A weird imagination is most useful to gain full advantage of
all the features.

Ruben Safir

2016-05-25 21:45:55 UTC

Post by D. Joe

Post by Ed Carp
The problem I have with this entire discussion is that Linux has
always stood for choice.

http://www.islinuxaboutchoice.com/

Quite right. Linux is just a kernel. It isn't about choice.

This gets counted as a bunch of euphorism tossed about by tech people
that don't mean anything when dissect them.

Linux is a kernel and the growth of Linux based operating systems has always
been about choice, and freedom to chose.

You can't just separate the sociological importance of something from its technical
merits and properties. This is not to imply that it has been ONLY about choice, but
freedom of choice, and PERSONAL CONTROL have been the driving forces of the OS since
its inception. To deny that just makes you sound like a mindless hack.

It is only relatively recently that big business has run roughshot of the end user/hacker
driven opportunity that Linux and Linux culture has represented. And now it is spread. LUGs
have driven up and Systemd and Android both represent the same cooperate imperative. Gone are
the volumes of useful HOWTOs and FAQs that came with every distro. Now we are facing an increasingly
top down hierarchy with no ear to the end user.

Imagine this. You guys actually passed off to a bunch of stupid kids that shell scripting
is inherently bad, and even unpredictable, and instead shell scripts should be replaced by
bright shiny incomprehensible binary that sits in the middle of everything like a cop, telling
you what you are allowed and what you are not allowed to do. I'm frankly flabbergasted.

The arrogance is beyond belief. And taking advantage of the uninformed with misinformation
with the use of FUD is a tactic that is hard to fathom. One can hardly believe that they are
actually listening to people from the Linux community.

This is a mindset of a group of people who actually believe that it is common for Linux users
to be running containers and that is healthy to have multiple instances of daemons running
simultaneously (defeating the whole purpose of a daemon).

Post by D. Joe
So glad to see
we're keeping it real, here.

What is REAL. Real is that monday morning you run an update and find that your entire OS has
been replaced with a different one. Boy that can really make for a bad week.

Post by D. Joe
*Computing* is fundamentally about choice, though: Read a bit, choose
whether and which bit to write. Repeat...until the entire tapestry of our
increasingly computing-controlled world has been woven.
The extent to which Linux is about computing, now, that's the tricky
question.
Maybe it's about joining the effort to sell a fantasy, in which a service or
appliance does exactly and only what we want, when we want it, how we want
it, pay no attention to the Turing machine behind the curtain. (Who "we"
are also left as an exercise for the reader.)

Right a thesis because this is gibberish. The fact that it is a Turing machine, and
I'm not certain that is even the case any longer in the classical sense, has nothing
to do with individual control of a system. Individual control is a MUST. This is not
supposed to be slaveware. Having control of your OS is a bit like having control of
your bladder. When you lose control the consequences are rather nasty.

Post by D. Joe
We must also ask ourselves these fundamental and obviously intrinsically
interlinked questions: "Do you ... hate handicap people? Do you hate
people who do not speak English? Do you only care about your own use-case
for GDM?"

I don't work for a cooperate entity and I don't need to ask this at all, not that it
has any direct relationship to systemd.

Post by D. Joe
http://youtu.be/ZTdUmlGxVo0
But I see we've already been drawing from that question pool, so, carry on.
--
Joe On ceding power to tech companies: http://xkcd.com/1118/
man screen | grep -A1 weird
A weird imagination is most useful to gain full advantage of
all the features.
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

D. Joe

2016-05-26 03:22:56 UTC

Post by Ruben Safir
Imagine this. You guys

Imagine this: You can tell whether someone just might, possibly, agree with
some things you've said.

Post by D. Joe
So glad to see
we're keeping it real, here.

What is REAL. Real is that monday morning you run an update and find that your entire OS has
been replaced with a different one. Boy that can really make for a bad week.

Good thing we have backups, eh?

Which individual? The one that just fed your machine an exploit?

http://langsec.org/occupy/

I don't work for a cooperate entity and I don't need to ask this at all, not that it
has any direct relationship to systemd.

OK, then how about the question "can I recognize sarcasm?"?

Post by D. Joe
http://youtu.be/ZTdUmlGxVo0
But I see we've already been drawing from that question pool, so, carry on.

--
Joe On ceding power to tech companies: http://xkcd.com/1118/
man screen | grep -A2 weird
A weird imagination is most useful to gain full advantage of
all the features.

Ruben Safir

2016-05-26 05:14:43 UTC

Post by D. Joe

Post by Ruben Safir
I don't work for a cooperate entity and I don't need to ask this at all, not that it
has any direct relationship to systemd.

OK, then how about the question "can I recognize sarcasm?"?

No I can NOT especially via email. Sarcasm is childish. If I had recognized it
I would have ignored it

Matt Palmer

2015-08-01 05:59:02 UTC

You appear to be describing https://devuan.org/.

- Matt

carmen r

2015-08-02 14:24:20 UTC

You appear to be describing devuan.org/.

devuan's marketing-materials fail to address why a fork was necessary

on a typical userspace-software distribution, switching init-systems involves something like:

0) install your init-system of choice. s6, runit, openrc (etc)
1) change one line in /etc/default/grub
GRUB_CMDLINE_LINUX="init=/usr/bin/runit-init"
2) run update-grub (to generate a new grub.cfg)

so Devuan wanted to immortalize "We're pissed off about systemd" in a bit of branding?
why didnt they just become runit/s6/openrc maintainers for sid instead?

Ruben Safir

2015-08-02 15:38:02 UTC

Post by carmen r

You appear to be describing devuan.org/.

devuan's marketing-materials fail to address why a fork was necessary
0) install your init-system of choice. s6, runit, openrc (etc)
1) change one line in /etc/default/grub
GRUB_CMDLINE_LINUX="init=/usr/bin/runit-init"
2) run update-grub (to generate a new grub.cfg)
so Devuan wanted to immortalize "We're pissed off about systemd" in a bit of branding?
why didnt they just become runit/s6/openrc maintainers for sid instead?

because of the increasingly specific dependency of packages on systemd,
for example....

jist tinkin out a dah box heer

Post by carmen r
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Don Marti

2015-08-02 16:12:21 UTC

Post by carmen r
so Devuan wanted to immortalize "We're pissed off about systemd" in a bit of branding?
why didnt they just become runit/s6/openrc maintainers for sid instead?

This is something I don't understand either.
On any packaging system it's easy to make an
"antipackage" that just has conflicts. So why not
just make a "no-systemd" package and install it?

It should Just Work, because Debian has to have
alternate init systems in order to support non-Linux
kernels. At least until alternate kernels add systemd
compatibility, the way the major proprietary Unixes
(remember them?) added "Linux kernel personalities".

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Rick Moen

2015-08-03 00:27:46 UTC

Post by Don Marti
This is something I don't understand either.
On any packaging system it's easy to make an
"antipackage" that just has conflicts. So why not
just make a "no-systemd" package and install it?

And indeed those have long existed, in the form of some of the various
prevent-* packages by Debian developer Thorsten Glaser ('mirabilos').
They work fine, FWIW. https://eurynome.mirbsd.org/debs/debidx.htm
https://wiki.debian.org/ThorstenGlaser

--
Cheers, "The trouble ain't that there is too many fools,
Rick Moen but that the lightning ain't distributed right."
***@linuxmafia.com -- Mark Twain
McQ! (4x80)

Marc MERLIN

2015-08-03 02:03:34 UTC

It's not that simple.
Many new packages will just not install because even though they don't
depend on systemd, they pick up systemd as a dep through automatic
dependency generation.

So basically, that would half freeze your debian system by preventing
many new packages from ever installing.

Marc

--
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
.... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/ | PGP 1024R/763BE901

Rick Moen

2015-08-03 02:45:08 UTC

Post by Marc MERLIN
It's not that simple.
Many new packages will just not install because even though they don't
depend on systemd, they pick up systemd as a dep through automatic
dependency generation.
So basically, that would half freeze your debian system by preventing
many new packages from ever installing.

It's a valid point, but, _so far_, I've not felt a pressing need for any
package with such a dependency (servers, workstations and laptops sans
DEs). Things may, of course, change -- and experiences will obviously
differ.

I may in the medium term (once again) build some packages locally to
trim dependencies, depending on particulars and how badly I want that
software. Network managers for wireless is one of my test cases for
that: I'm doing a survey of which ones are maintained and usable, then
checking and documenting what dependencies they tend to be built with --
along with what dependencies are truly needed.

Marc MERLIN

2015-08-03 04:14:55 UTC

It's a valid point, but, _so far_, I've not felt a pressing need for any
package with such a dependency (servers, workstations and laptops sans
DEs). Things may, of course, change -- and experiences will obviously
differ.
I may in the medium term (once again) build some packages locally to
trim dependencies, depending on particulars and how badly I want that
software. Network managers for wireless is one of my test cases for
that: I'm doing a survey of which ones are maintained and usable, then
checking and documenting what dependencies they tend to be built with --
along with what dependencies are truly needed.

Funny you should mention that, gnome's network-manager is what caused a
dependency hell which installed systemd which then nicely also decided to
take over my acpi functions, conflicting with the ones I had in acpid.
I'm still booting with regular init though.

Of course, it's my fault for running network-manager, I just have so many
network configs in it, that it's been a pain to switch to something else,
but if you find one you can recommend, I would be happy to listen, because
I'm sure not happy with at least nm-applet (which I run inside some systray
inside enlightenment 19).
(I don't otherwise run gnome)

Marc

--
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
.... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/

Rick Moen

2015-08-03 04:41:22 UTC

Of course, it's my fault for running network-manager....

Yep. ;-> But I sympathise.

if you find one you can recommend, I would be happy to listen, because
I'm sure not happy with at least nm-applet (which I run inside some systray
inside enlightenment 19).

A huge number of people seem to like wicd.

On Debian, 'wicd' is a metapackage. You need core package wicd-daemon
and one or more of: wicd-gtk, wicd-curses, wicd-cli, wicd-client.

The clients are Python (+ respective rendering libs). The daemon
requires dbus, wpasupplicant, wireless-tools, and a few obvious network
things. Nothing surprising.

Don Marti

2015-08-03 03:49:48 UTC

Post by Marc MERLIN

It's not that simple.
Many new packages will just not install because even though they don't
depend on systemd, they pick up systemd as a dep through automatic
dependency generation.
So basically, that would half freeze your debian system by preventing
many new packages from ever installing.

So if exampleware 1.0 does not depend on systemd,
but exampleware 1.1 does, would the user who installs
both exampleware and prevent-systemd end up stuck
on exampleware 1.0 when exampleware 1.1 comes out,
or would that user be forced to remove either
prevent-systemd or exampleware?

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Rick Moen

2015-08-03 05:40:22 UTC

Post by Don Marti
So if exampleware 1.0 does not depend on systemd,
but exampleware 1.1 does, would the user who installs
both exampleware and prevent-systemd end up stuck
on exampleware 1.0 when exampleware 1.1 comes out,
or would that user be forced to remove either
prevent-systemd or exampleware?

The prevent-* packages substantively consist just of a Conflicts line.
Here's the one inside the 'control' file of the latest
prevent-systemd-running deb:

Package: prevent-systemd-running
Version: 28
Section: metapackages
Priority: extra
Architecture: all
Multi-Arch: foreign
Conflicts: live-config-systemd, systemd-cron, systemd-sysv, systemd-ui
Installed-Size: 44
Origin: WTF
Maintainer: Thorsten Glaser <***@mirbsd.de>
Bugs: mailto:***@mirbsd.org
Source: mirabilos-support
Description: mirabilos metapackage to avoid running systemd
This metapackage ensures that systemd is not used as the init
system for your computer. This is necessary to enable standard
Unix/GNU/Linux system administrators to operate your system.
.
This variant of the metapackage permits installation of systemd
to satisfy package dependencies, but not of the systemd-sysv
package which takes over the init system.
Note that, once installed, systemd can still be run by changing
the <E2><80><9C>init=<E2><80><9D> kernel parameter. Be careful.
important: yes

To answer your question as seems most informative, the presence of a
prevent-* package on the system merely ensures that the initial attempt
to install a package with a conflicting Depends will initially fail.
How the sysadmin then decides to proceed is up to him/her.

There have been systemd packages in Debian for something like five
years, now, and the problems remain localised to GNOME, gdm, and
NetworkManager. Could the sky fall tomorrow? Magic 8 ball unclear.

Rick Moen

2015-08-03 04:35:19 UTC

FWIW, in looking through /var/lib/apt/lists/*Packages files for direct
dependencies on package systemd, they seem extremely rare -- details
below. There are many direct dependencies on libsystemd0, but that
should not bother anyone but the rabid, concerned about infection of
their systems by systemd's homeopathic essence even in the absence of
systemd itself. ;->

I'm guessing you mean the sysadmin requested package A, A has a Depends
for package B, B has a Depends for package systemd (or some longer but
similar chain that reaches systemd). Correct?

Jessie x86_64 currently has:
Depends on 'systemd-shim' but not systemd: 1 (init-select)
Depends on 'systemd-shim | systemd-sysv' but not systemd: 1 (xfce4-session)
Depends on 'netbase | systemd-sysv' but not systemd: 1 (gpsd)
Depends on 'systemd-sysv' but not systemd: 1 (systemd-cron)
Depends on 'systemd-ui' but not systemd: 1 (systemd-gui)
Depends on 'systemd | consolekit': 2 (mate-power-manager, solaar)
Depends on 'systemd | sysvinit': 1 (libguestfs0)
Depends on 'tmpreaper | systemd': 1 (sogo)
Depends on 'lsb-base | systemd': 1 (ligthttpd)
Depends on 'consolekit | upower | systemd': 1 (lxsession)
Depends on 'systemd' (with or without version qualfiers): 14

The 14 package with a direct dependency on package systemd are:
debian-installer (and nine other debian-installer* variants),
live-config-systemd, libpam-systemd, systemd-dbg, systemd-sysv.

Anyway, please detail the 'automatic dependency generation' scenario
you're contemplating, if/when you have time -- because I'm having a
difficult time envisioning a chain of Depends involving any of those
14 on my non-GNOME, non-NetworkManager systems.[1] I trust you're not
talking about Recommends, as automatic installation of Recommends
packages can be readily suppressed via lines like these in
/etc/apt.con.d/* :

APT::Install-Recommends "0";
APT::Install-Suggests "0";

(Second of those hasn't been needed, but is just in case.)

[1] Closest to problematic in that regard, it seems to me, as
libpam-systemd. Here are the packages that depend on that package:
gdm3, gnome-bluetooth, gnome-settings-daemon, network-manager,
policykit-1, udisks2, wmshutdown. (Package lightdm requires either
libpam-systemd or consolekit.) No surprise that GNOME and
NetworkManager are horrendous dependency hairballs, but that's an
obvious cue for the 'Doctor, it _hurts_ when I do this!' joke.

Marc MERLIN

2015-08-03 15:04:18 UTC

I'm guessing you mean the sysadmin requested package A, A has a Depends
for package B, B has a Depends for package systemd (or some longer but
similar chain that reaches systemd). Correct?

Yes.

Post by Rick Moen
Depends on 'systemd-shim' but not systemd: 1 (init-select)
Depends on 'systemd-shim | systemd-sysv' but not systemd: 1 (xfce4-session)
Depends on 'netbase | systemd-sysv' but not systemd: 1 (gpsd)
Depends on 'systemd-sysv' but not systemd: 1 (systemd-cron)
Depends on 'systemd-ui' but not systemd: 1 (systemd-gui)
Depends on 'systemd | consolekit': 2 (mate-power-manager, solaar)
Depends on 'systemd | sysvinit': 1 (libguestfs0)
Depends on 'tmpreaper | systemd': 1 (sogo)
Depends on 'lsb-base | systemd': 1 (ligthttpd)
Depends on 'consolekit | upower | systemd': 1 (lxsession)
Depends on 'systemd' (with or without version qualfiers): 14

Well, my system currently says:
sudo apt-get remove systemd
The following packages will be REMOVED:
libpam-systemd network-manager network-manager-gnome systemd

Post by Rick Moen
14 on my non-GNOME, non-NetworkManager systems.[1] I trust you're not
talking about Recommends, as automatic installation of Recommends
packages can be readily suppressed via lines like these in
APT::Install-Recommends "0";
APT::Install-Suggests "0";

Indeed. I already turned that off:
/etc/apt/apt.conf:APT::Install-Recommends "false";

Post by Rick Moen
[1] Closest to problematic in that regard, it seems to me, as
gdm3, gnome-bluetooth, gnome-settings-daemon, network-manager,
policykit-1, udisks2, wmshutdown. (Package lightdm requires either
libpam-systemd or consolekit.) No surprise that GNOME and
NetworkManager are horrendous dependency hairballs, but that's an
obvious cue for the 'Doctor, it _hurts_ when I do this!' joke.

There you go.
Stuff depends on libpam-systemd and it depends on systemd.

I agree that networkmanager has "issues" as I mentioned in my previous post.

e19 recommends connman. I'm not sure what it's worth today.

wicd also didn't quite do what I needed (*) at the time last I looked at it,
but it's been several years since I last tried, so I can look again

(*) missing at the time:
- networking via bluetooth through my phone
- networking via usb tethering through my phone

I'm hopeful those work now.

And just to be clear, I don't actually have an agenda against systemd, I
just have very customized systems where upgrading to systemd is guaranteed
to break stuff and cost me a lot of time to switch, so I'm not in a hurry to
do this.
Having systemd on a new fresh install and migrate will be a lot easier, but
that again, is work for me and I'm not in a hurry since I have other things
to do.

Marc

--
"A mouse is a device used to point at the xterm you want to type in" - A.S.R.
Microsoft is to operating systems ....
.... what McDonalds is to gourmet cooking
Home page: http://marc.merlins.org/

Ruben Safir

2015-08-03 03:19:11 UTC

Post by carmen r
so Devuan wanted to immortalize "We're pissed off about systemd" in a bit of branding?
why didnt they just become runit/s6/openrc maintainers for sid instead?

This is something I don't understand either.
On any packaging system it's easy to make an
"antipackage" that just has conflicts. So why not
just make a "no-systemd" package and install it?

I'm not a debian user so i can't speak for the system specifically but
in general, it is likely easier to make packages for different kernels
than to make non-systemd packages from a systemd centric platform, and
it will be this way in the future. It is a stated goal of systemd
developers that all linux distros will look and act exactly the same
way.

Systemd is NOT an init system. It can best be described, IMO, as an OS
that sits on the Linux Kernel.

and if you think ifconfig is depecated, just see how bad rsyslog and
family will look in a few years.

Can someone tell me how many debian packages are no
depend on systemd as a requesist?

Quote:
Other parts include a logging daemon, utilities to control basic system
configuration like the hostname, date, locale, maintain a list of
logged-in users and running containers and virtual machines, system
accounts, runtime directories and settings, and daemons to manage simple
network configuration, network time synchronization, log forwarding, and
name resolution.

http://www.freedesktop.org/wiki/Software/systemd/writing-display-managers/
just for one complication

Post by Don Marti
It should Just Work, because Debian has to have
alternate init systems in order to support non-Linux
kernels. At least until alternate kernels add systemd
compatibility, the way the major proprietary Unixes
(remember them?) added "Linux kernel personalities".
--
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Greg KH

2015-08-03 04:26:03 UTC

Post by carmen r
so Devuan wanted to immortalize "We're pissed off about systemd" in a bit of branding?
why didnt they just become runit/s6/openrc maintainers for sid instead?

This is something I don't understand either.
On any packaging system it's easy to make an
"antipackage" that just has conflicts. So why not
just make a "no-systemd" package and install it?

No, it's not that at all, it provides the same services and interfaces
that all the distros had to create on their own because no one else had
done so. And by now unifying it across all distros, it provides an
interfaces that application developers can rely on being there to solve
real problems.

The "fun" thing about the whole Debian mess I always found was that most
of the interfaces, and standards that systemd implemented came directly
from Debian, so in fact, all old-time Debian admins should feel more at
home on Fedora, openSUSE, and everyone else.

Post by Ruben Safir
and if you think ifconfig is depecated, just see how bad rsyslog and
family will look in a few years.

If an operating system does not change, it is dead.

good luck!

greg k-h

Rick Moen

2015-08-03 07:27:03 UTC

Quoting Greg KH (***@kroah.com):

[unifying, blah blah]

You mean documented, stable interfaces as in uselessd, right? ;->
Because, systemd itself has been actually a prime example of churn,
and a really dreadful example of stability.

Post by Greg KH
The "fun" thing about the whole Debian mess I always found was that most
of the interfaces, and standards that systemd implemented came directly
from Debian, so in fact, all old-time Debian admins should feel more at
home on Fedora, openSUSE, and everyone else.

ISTR that the crisis arose because of a build decision for the gdm3
package, where the Debian developers suddenly decided that the gdm3
package was going to have a Depends declaration requring systemd to
ensure that the systemd-logind daemon would be present.

And why would systemd-logind need be present just so an X11 display
manager could function? Because the GNOME Project, in its great wisdom,
had decided that logging in shall require not _just_ logging in but also
'seat' state data to support a weird Freedesktop.org-specific user model
-- and Consolekit had heretofore provided that API but, in accordance
with the frequent churn of typical Freedesktop.org projects, was now
essentially orphaned and deprecated (and ConsoleKit2 might eventually
get created but didn't exist yet).

Thus, gdm on Debian would start requiring systemd, and because GNOME
was/is the standard default DE on Debian, that meant Debian had to start
requiring systemd. Everyone got dragged along by the code churn of some
GNOME jackasses.

Except, of course, the many of us whose reaction was 'Hmm, graduitous
dependency drama or no GNOME? Right then, Column B it is.'

Post by Greg KH
If an operating system does not change, it is dead.

And that's why everyone loved HAL and devfs, you may remember. ;->

(Good luck getting sysadmins to consider systemd-journald binary
logfiles a good idea just because an OS 'needs to change'.)

--
Cheers, Atque memento, nulli adsunt Romanorum
Rick Moen qui locutionem tuam corrigant.
***@linuxmafia.com
McQ! (4x80)

Teh Entar-Nick

2015-08-03 11:00:57 UTC

Post by Rick Moen
[unifying, blah blah]

Remember this scene from Tron?

Master Control Program: All Programs have a desire to be useful.
But in moments, you will no longer seek
communication with each other, or your
superfluous Users. You will each be a part
of me. And together, we will be complete.

Don Marti

2015-08-03 16:03:47 UTC

Post by Rick Moen
And why would systemd-logind need be present just so an X11 display
manager could function? Because the GNOME Project, in its great wisdom,
had decided that logging in shall require not _just_ logging in but also
'seat' state data to support a weird Freedesktop.org-specific user model

So you can build a school computer lab with one $600
computer per 4 students instead of 1 $300 computer
per student.

https://cedarandthistle.wordpress.com/2013/11/30/multiseat-in-fedora-19/

https://bluehatrecord.wordpress.com/2015/01/15/multiseat-configuration-in-fedora-21/

Sometimes 1/4 of a decent computer is better than a
thin client or a whole cheap computer.

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Rick Moen

2015-08-03 05:21:53 UTC

Post by Ruben Safir
I'm not a debian user so i can't speak for the system specifically but
in general, it is likely easier to make packages for different kernels
than to make non-systemd packages from a systemd centric platform, and
it will be this way in the future.

You are wildly hand-waving without any data whatsoever. Admittedly,
this is a lot easier than thinking, but please see my posting in this
thread in response to Marc where I did the needed homework about the
current state of dependencies on package 'systemd' within Debian 8.0
Jessie (x86_64).

Clue: Those data fail to support your claim. The Linneus Bottomus:

1. Several GNOME packages depend on libpam-systemd, and thus systemd.
2. gdm3 (came from GNOME) depends on libpam-systemd, and thus systemd.
3. lightdm (came from GNOME) depends on libpam-systemd or consolekit.
4. NetworkManager (came from GNOME) depends on libpam-systemd, and thus systemd.

And that's just about it. Avoid dropping the GNOME or NetworkManager
hammers on your foot, and your Debian Jessie system does not become a
'systemd-centric platform'.

Yes, the default Jessie installer supplies systemd. Two obvious apt-get
commands -- install a different init, remove systemd -- and it's gone.
Other '...and stay out!' measures like Thorsten Glazer's prevent-*
packages and this /etc/apt/preferences.d/systemd file can be added, but
aren't really necessary:

Package: systemd
Pin: origin ""
Pin-Priority: -1

Post by Ruben Safir
It is a stated goal of systemd developers that all linux distros will
look and act exactly the same way.

And that plus $2.25 will get you a ride on SF Muni. I wish them joy of
their occasional Bond-villain ranticles, but that has nothing to do with
real systems.

Post by Ruben Safir
and if you think ifconfig is depecated, just see how bad rsyslog and
family will look in a few years.

Right, let's make this interesting: I am offering to wager US $100 that
packages for _both_ rsyslog (or a similar successor) _and_ syslog-ng (or
a similar successor) will still be trivially installable on Debian,
usable on Debian, and developer-maintained on August 2, 2018. If you're
confident of what you say, this will be easy money for you. For my
part, on the basis of professional experience as a senior sysadmin, I'm
confident you're profoundly mistaken, and am willing to take your money.

Wager offer is extended for 48 hours from the timestamp on this posting,
and you'll need to indicate acceptance on this mailing list. (In the
event of dispute about the state of affairs on August 2, 2018, we'll
defer to Mr. Don Marti's judgement if he's willing.)

Post by Ruben Safir
Can someone tell me how many debian packages are no depend on systemd
as a requesist?

¿Qué significa esta palabra 'requesist'? No creo que significa lo que
usted piensa que significa.

All of that rhetoric, and you explicitly admit you have _no_ data.
Tsk-tsk.

See my posting to this thread for actual data. And there are currently
about 44000 packages in Debian 8.0 Jessie for x86_64 (main + contrib +
non-free).

Ruben Safir

2015-08-03 13:28:03 UTC

Post by Ruben Safir
family will look in a few years.

Right, let's make this interesting: I am offering to wager US $100 that
packages for _both_ rsyslog (or a similar successor) _and_ syslog-ng (or
a similar successor) will still be trivially installable on Debian,
usable on Debian, and developer-maintained on August 2, 2018. If you're
confident of what you say, this will be easy money for you. For my
part, on the basis of professional experience as a senior sysadmin, I'm
confident you're profoundly mistaken, and am willing to take your money.
Wager offer is extended for 48 hours from the timestamp on this posting,
and you'll need to indicate acceptance on this mailing list. (In the
event of dispute about the state of affairs on August 2, 2018, we'll
defer to Mr. Don Marti's judgement if he's willing.)

I've already had trouble getting logging to function on Suse and Manjaro
from its nicely integrated package until 6 months ago, but if you wish
to change from $100 to the typical betting vehical of choice on the
internet which is Junior Cheese Cakes, then I'm willing to make a go of
this

Post by Ruben Safir
Can someone tell me how many debian packages are no depend on systemd
as a requesist?

¿Qué significa esta palabra 'requesist'? No creo que significa lo que
usted piensa que significa.
All of that rhetoric, and you explicitly admit you have _no_ data.
Tsk-tsk.

How many current debian packages have dependency on systemd ?

The Gnome interface, btw, is not a small matter, and the wifi networking
tools are not simple to drop in and replace.

How do you install, now, X11 without systemd? Those would be two
different install approaches altogehter, and would require a split in
the packages if one is using the rootless systemd setup.

Post by Rick Moen
See my posting to this thread for actual data. And there are currently
about 44000 packages in Debian 8.0 Jessie for x86_64 (main + contrib +
non-free).
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Rick Moen

2015-08-03 18:29:15 UTC

Post by Ruben Safir
I've already had trouble getting logging to function on Suse and Manjaro
from its nicely integrated package until 6 months ago, but if you wish
to change from $100 to the typical betting vehical of choice on the
internet which is Junior Cheese Cakes, then I'm willing to make a go of
this.

Junior's Cheesecakes of Brooklyn looks to do fabulous work;
unfortunately, lactose and I long ago has a parting of ways.

Fortunately, just across the East River sits Second Avenue Deli,
purveyor of fine pastrami sandwiches with potato salad and mixed
pickles, one dish of which I'm glad to have as my wager payoff.
But, sorry, I have no ability to predict how often OpenSUSE or Manjaro
breaks logging, or how important functional rsyslog and syslog-ng
logging is to their developers -- but I do know Debian cares quite a bit
about that, which is why, for example, Debian's systemd package disables
systemd's systemd-journald feature and redirects system logging output
to rsyslog.

Wager offer as amended to a Junior's cheesecake and the best pastrami in
Midtown East is still outstanding -- but still limited to Debian rsyslog
/ syslog-ng, because Debian's serious about functional system logging.

Post by Ruben Safir
How many current debian packages have dependency on systemd ?

Asked and answered. To recap:

10 debian-installer* packages that depend directly on systemd because
the Debian 8.0 default installer provides it.

2 GNOME packages that depend on systemd or its PAM auth library
(gnome-bluetooth, gnome-settings-daemon).

1 WindowMaker dock applet for shutting down a machine by clicking a
button (wmshutdown).

5 packages that depend on systemd because they're systemd-related:
(live-config-systemd, libpam-systemd, systemd-dbg, systemd-sysv,
libpam-systemd).

1 GNOME display manager (gdm3).

1 GNOME-affiliated display manager that requires either libpam-system or
consolekit (lightdm).

6 assorted other packages that require that systemd _or_ something else be
present (mate-power-manager, solaar, libguestfs0, sogo, ligthttpd,
lxsession).

2 packages from the core Freedesktop.org stack -- the guys responsible
for most of the furious code churn in GNOME -- that depend on
libpam-systemd (policykit-1, udisks2).

1 network manager from GNOME that depends on libpam-systemd
(network-manager).

For completeness, one would also want ot chase down what depends on
policykit-1 ('framework for managing administrative policies and
privileges') or udisks2 ('D-Bus service to access and manipulate storage
devices'). I'm sure they're becoming a bit of a problem among the sorts
of desktop software that has been wooed over by the Freedesktop.org
infrastructure. At a glance, looks like the most annoying of those is
hplip, which depends on policykit-1. (Why on Earth? So, that one
useful package might need local rebuilding.) Mostly, otherwise, it's a
dozen or so libs, gnome-core, gnome-disk-utility, and things like the
pcmanfm lightweight graphical file manger, and daisy-player ('player for
DAISY Digital Talking Books'), both of which I think I can live without.

Post by Ruben Safir
The Gnome interface, btw, is not a small matter...

Bloated, in fact. ;->

Yeah, sorry, GNOME is deeply in bed with the Freedesktop.org framework,
which means 'seat' support is required in the display manager, thus
requiring either libpam-systemd or ConsoleKit, and the latter is
unmaintained and suffering bitrot like a typical Freedesktop.org 'We
redesign everything every six months' project. Also, you get
dependencies via policykit-1 and udisks.

So, write it off, or learn to love the whole Freedesktop.org code
hairball, including systemd (which actually should be the least of your
worries, but maybe you've been oblivious about what's gone on with
GNOME).

Post by Ruben Safir
....and the wifi networking tools are not simple to drop in and
replace.

Are you saying you were yet another person caught napping by
NetworkManager wanting to pull in the whole code hairball? Well, there
at least you have good company, as Marc got caught by the same thing.
However, you both basically _should_ have been forewarned, because
NetworkManager says GNOME right on the tin.

Time for you to look at wicd, connman, etc. Inconvenient? Shed a
bitter tear into your egg cream, and deal.

Post by Ruben Safir
How do you install, now, X11 without systemd?

On Debian, using apt-get or tasksel. You seem to be assuming problems
not present. If you're saying OpenSUSE presented you with some tale of
woe, sorry to hear that.

Post by Ruben Safir
Those would be two different install approaches altogehter, and would
require a split in the packages if one is using the rootless systemd
setup.

You might need to unpack that. I'm guessing you encountered some trauma
on distros that try to abstract session management via systemd-logind or
ConsoleKit and suspend/hibernate via upower (all of that
Freesdesktop.org stack's stuff), and where, thus, the X server cannot
open /dev/ttyN at startup because it lacks root rights unless the whole
Freedesktop.org thundering herd of systemd-logind | ConsoleKit, upower,
PolicyKit, udisk2 (etc.) being present, because otherwise lacks
root-window rights and isn't being helped by the Freedesktop.org stuff
to run 'rootless'. Generally, this is a GNOME problem.

(I should stress that I don't really know this stuff, having gone to
some pains to avoid needing to know it.)

The obstinate take measures to restore the X server's root-window rights
xorg-xwrapper (if you omitted it from the commands above) or by creating
/etc/X11/Xwrapper.config (containing 'echo "needs_root_rights = yes"')
and, basically write off GNOME as hopeless.

The lazy (like me) take the more prudent path of sticking to distros
that haven't fallen for the whole Freedesktop.org tangleware trap --
like Debian, for example. And also write off GNOME as hopeless.

My view, yours for a small fee and waiver of reverse-engineering rights.

Ruben Safir

2015-08-04 01:58:05 UTC

FWIW, the 2nd avenue Deli that we holed up in waaaay back then is long
gone. An imitation of it sprung up in Midtown but you might be better
off with Jay and Lloyds.

Regardless, I take this offer and might be willing to the Bay Area to
join in any festivities that involves all my westcoast loved ones

Ruben

\@/
|
/ \

Junior's Cheesecakes of Brooklyn looks to do fabulous work;
unfortunately, lactose and I long ago has a parting of ways.
Fortunately, just across the East River sits Second Avenue Deli,
purveyor of fine pastrami sandwiches with potato salad and mixed
pickles, one dish of which I'm glad to have as my wager payoff.
But, sorry, I have no ability to predict how often OpenSUSE or Manjaro
breaks logging, or how important functional rsyslog and syslog-ng
logging is to their developers -- but I do know Debian cares quite a bit
about that, which is why, for example, Debian's systemd package disables
systemd's systemd-journald feature and redirects system logging output
to rsyslog.
Wager offer as amended to a Junior's cheesecake and the best pastrami in
Midtown East is still outstanding -- but still limited to Debian rsyslog
/ syslog-ng, because Debian's serious about functional system logging.

Post by Ruben Safir
How many current debian packages have dependency on systemd ?

10 debian-installer* packages that depend directly on systemd because
the Debian 8.0 default installer provides it.
2 GNOME packages that depend on systemd or its PAM auth library
(gnome-bluetooth, gnome-settings-daemon).
1 WindowMaker dock applet for shutting down a machine by clicking a
button (wmshutdown).
(live-config-systemd, libpam-systemd, systemd-dbg, systemd-sysv,
libpam-systemd).
1 GNOME display manager (gdm3).
1 GNOME-affiliated display manager that requires either libpam-system or
consolekit (lightdm).
6 assorted other packages that require that systemd _or_ something else be
present (mate-power-manager, solaar, libguestfs0, sogo, ligthttpd,
lxsession).
2 packages from the core Freedesktop.org stack -- the guys responsible
for most of the furious code churn in GNOME -- that depend on
libpam-systemd (policykit-1, udisks2).
1 network manager from GNOME that depends on libpam-systemd
(network-manager).
For completeness, one would also want ot chase down what depends on
policykit-1 ('framework for managing administrative policies and
privileges') or udisks2 ('D-Bus service to access and manipulate storage
devices'). I'm sure they're becoming a bit of a problem among the sorts
of desktop software that has been wooed over by the Freedesktop.org
infrastructure. At a glance, looks like the most annoying of those is
hplip, which depends on policykit-1. (Why on Earth? So, that one
useful package might need local rebuilding.) Mostly, otherwise, it's a
dozen or so libs, gnome-core, gnome-disk-utility, and things like the
pcmanfm lightweight graphical file manger, and daisy-player ('player for
DAISY Digital Talking Books'), both of which I think I can live without.

Post by Ruben Safir
The Gnome interface, btw, is not a small matter...

Bloated, in fact. ;->
Yeah, sorry, GNOME is deeply in bed with the Freedesktop.org framework,
which means 'seat' support is required in the display manager, thus
requiring either libpam-systemd or ConsoleKit, and the latter is
unmaintained and suffering bitrot like a typical Freedesktop.org 'We
redesign everything every six months' project. Also, you get
dependencies via policykit-1 and udisks.
So, write it off, or learn to love the whole Freedesktop.org code
hairball, including systemd (which actually should be the least of your
worries, but maybe you've been oblivious about what's gone on with
GNOME).

Post by Ruben Safir
....and the wifi networking tools are not simple to drop in and
replace.

Are you saying you were yet another person caught napping by
NetworkManager wanting to pull in the whole code hairball? Well, there
at least you have good company, as Marc got caught by the same thing.
However, you both basically _should_ have been forewarned, because
NetworkManager says GNOME right on the tin.
Time for you to look at wicd, connman, etc. Inconvenient? Shed a
bitter tear into your egg cream, and deal.

Post by Ruben Safir
How do you install, now, X11 without systemd?

On Debian, using apt-get or tasksel. You seem to be assuming problems
not present. If you're saying OpenSUSE presented you with some tale of
woe, sorry to hear that.

Post by Ruben Safir
Those would be two different install approaches altogehter, and would
require a split in the packages if one is using the rootless systemd
setup.

You might need to unpack that. I'm guessing you encountered some trauma
on distros that try to abstract session management via systemd-logind or
ConsoleKit and suspend/hibernate via upower (all of that
Freesdesktop.org stack's stuff), and where, thus, the X server cannot
open /dev/ttyN at startup because it lacks root rights unless the whole
Freedesktop.org thundering herd of systemd-logind | ConsoleKit, upower,
PolicyKit, udisk2 (etc.) being present, because otherwise lacks
root-window rights and isn't being helped by the Freedesktop.org stuff
to run 'rootless'. Generally, this is a GNOME problem.
(I should stress that I don't really know this stuff, having gone to
some pains to avoid needing to know it.)
The obstinate take measures to restore the X server's root-window rights
xorg-xwrapper (if you omitted it from the commands above) or by creating
/etc/X11/Xwrapper.config (containing 'echo "needs_root_rights = yes"')
and, basically write off GNOME as hopeless.
The lazy (like me) take the more prudent path of sticking to distros
that haven't fallen for the whole Freedesktop.org tangleware trap --
like Debian, for example. And also write off GNOME as hopeless.
My view, yours for a small fee and waiver of reverse-engineering rights.
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Rick Moen

2015-08-04 04:40:26 UTC

Post by Ruben Safir
FWIW, the 2nd avenue Deli that we holed up in waaaay back then is long
gone. An imitation of it sprung up in Midtown but you might be better
off with Jay and Lloyd's.

FYI, same firm, same owner, same great food -- lacks only the original
location's atmosphere and half-century of history. They had to move
in '06-'07 because of a stratospheric rent increase.
http://www.nytimes.com/2007/10/21/magazine/21deli-t.html

Post by Ruben Safir
Regardless, I take this offer and might be willing to the Bay Area to
join in any festivities that involves all my westcoast loved ones

Splendid. My Debian systems will log it all. ;->

--
Cheers, Mayn shveb-shif iz ful mit vengers.
Rick Moen
***@linuxmafia.com
McQ! (4x80)

Bob Bernstein

2015-08-03 15:24:58 UTC

Post by Rick Moen
1. Several GNOME packages depend on libpam-systemd,
and thus systemd. 2. gdm3 (came from GNOME) depends
on libpam-systemd, and thus systemd. 3. lightdm
(came from GNOME) depends on libpam-systemd or
consolekit. 4. NetworkManager (came from GNOME)
depends on libpam-systemd, and thus systemd.

You are in good form today. Thanks for making sense of
this clusterf**k.

Post by Rick Moen
Avoid dropping the GNOME or NetworkManager hammers
on your foot, and your Debian Jessie system does not
become a 'systemd-centric platform'.

And _that_ is my "take-away!"

--
"Natasha! First we shoot jet, then we go after moose and squirrel."

Eugen Leitl

2015-08-04 12:50:20 UTC

Post by Bob Bernstein

Avoid dropping the GNOME or NetworkManager hammers on your foot,
and your Debian Jessie system does not become a 'systemd-centric
platform'.

And _that_ is my "take-away!"

But of course this is exactly the scenario that they told us
wouldn't happen. And these are just the dependencies as of mid-2015.

Rick Moen

2015-08-04 21:32:52 UTC

Post by Eugen Leitl

Post by Bob Bernstein

Avoid dropping the GNOME or NetworkManager hammers on your foot,
and your Debian Jessie system does not become a 'systemd-centric
platform'.

And _that_ is my "take-away!"

But of course this is exactly the scenario that they told us
wouldn't happen. And these are just the dependencies as of mid-2015.

(Just to stress: I don't speak for the Debian Project in any way. I
just use the stuff, and am otherwise a project outsider.)

To the extent there are things in Debian that now in 2015 require systemd
-- by my count, 13 assorted regular packages (mostly GNOME things), 10
Debian installer packages, and 5 that require systemd only because
they're systemd-related, out of ~44,000 Debian x86_64 packages total --
they were largely imposed onto the Debian Project by upstream GNOME
decisions related to the GNOME 'login' API (seat support) and
Freedesktop.org developers' orphaning of ConsoleKit.

Even if the Debian Project had been militantly opposed to that upstream
development and undivided in its members' views, there isn't a lot they
could have done without dropping GNOME as a supported and default DE.
Even the more strongly source code-oriented distros such as
Gentoo/Funtoo and Sorcerer/Lunar Linux have been caught up in the same
mess, even though they have greater overall ease of controlling
dependencies through custom build options, and less incentive to throw
in features to create one-size-fits-all binary packages.

Personally, I wish Debian Project _had_ opted to drop GNOME as a
supported and default DE. However, from my perspective it's just as
good and functionally equivalent to just deliberately omit GNOME locally
(including NetworkManager, which is a GNOME codebase).

Over half a decade, accumulation of 28 packages out of ~44,000 with
dependencies on systemd strikes me as negligible -- unless of course you
want GNOME or NetworkManager, in which case good luck with that, and
your beef should logically be with the GNOME / Freedesktop.org people,
not with the Debian Project -- irrespective of what someone (who? Russ
Allbery? the Technical Committee? The Debian Project Leader?) some
years ago told you wouldn't happen.

Of course, for the folks who've never met a tangled software dependency
they didn't like, things are fine and getting better all the time. Back
in the 90s, such things were called 'integration' in the IT press in
tones of great approval. I think 'unifying' might have been in there,
too.

Ruben Safir

2015-08-05 02:51:55 UTC

Post by Rick Moen
(Just to stress: I don't speak for the Debian Project in any way. I
just use the stuff, and am otherwise a project outsider.)
To the extent there are things in Debian that now in 2015 require systemd
-- by my count, 13 assorted regular packages (mostly GNOME things), 10
Debian installer packages, and 5 that require systemd only because
they're systemd-related, out of ~44,000 Debian x86_64 packages total --
they were largely imposed onto the Debian Project by upstream GNOME
decisions related to the GNOME 'login' API (seat support) and
Freedesktop.org developers' orphaning of ConsoleKit.

It is hard to understand how it can be double digits. Gnome depedent
packages alone is gigabytes of packages. Then there is is the affects to
X and logging. This is a fork. Your looking over your shoulder and
saying, I can run around this corner pretty quick right now. yeah. But
not in the future. Not as things seem to be progressing.

You can call it hand waiving if you want, and I'll admit to being quite
passionate on this subject, but if systemd is the linux future, I want
no part of it.

Ruben

Greg KH

2015-08-05 04:58:03 UTC

Post by Ruben Safir
You can call it hand waiving if you want, and I'll admit to being quite
passionate on this subject, but if systemd is the linux future, I want
no part of it.

Ok, I'll take the bait.

Please explain _why_ you don't like systemd.

Rick Moen

2015-08-05 06:24:26 UTC

Post by Ruben Safir
It is hard to understand how it can be double digits. Gnome depedent
packages alone is gigabytes of packages. Then there is is the affects to
X and logging. This is a fork. Your looking over your shoulder and
saying, I can run around this corner pretty quick right now. yeah. But
not in the future. Not as things seem to be progressing.

Because I need GNOME? I clearly do not, so that can't be why. You
don't say what 'progressing' means, here. Cound that be because, I
dunno, just spitballing here, the facts aren't on your side?

It would take a massive uptake _outside_ GNOME of secondary
Freedesktop.org 'desktop' software infrastructure with systemd
dependencies -- PolKit, udisks2 -- before I would even perceive a
problem of desirable packages I cannot get without the code hairball.
And that's been predicted by you and some others for several years to be
imminent, but continues to just not happen.

There have always been code hairballs in the Linux software ecosystem,
and people have always bitched about excessive package dependency trees.
Stupid ideas have frequently been lobbed at use from desktop space.
Remember when everything was going to need to work as object software in
coordination with CORBA brokers? Remember when everything was going to
run with root authority because users can't handle the concepts of
ownership and permissions?

Things have not been 'progressing' over the years you've been making bad
and poorly informed predictions.

The sum and substance of this is, a couple of years ago, some drama
happened with you on OpenSUSE 12.x, and part of it involved inability to
boot because of some startup task problem and systemd wouldn't let you
get into the problem to recover the system. And then there was some
claim about how X.org moving to running X servers as a non-privileged
users was horrible because doing so required Freedesktop.org tangleware
pieces PolKit, systemd-logind, and upower[1], some of which require
systemd -- when it simply not the case that X needs to require those.

You cursed a blue streak, moved sideways from OpenSUSE to Manjaro, noted
that your various problems went away, blamed them all on systemd, and
started asserting that everyone else would soon be equally plagued.

And you keep saying that -- but it keeps not happening.

One of us isn't really following the details of what's been going on.
You can think it's me. I will take that view under advisement.

In case you want to understand the X server matter, here's a page by
some Arch Linux people who've documented how to do systemd-free X.org
even using MATE or Cinnamon. Note in particular the section about
'Rootless X': http://systemd-free.org/config.php

[1] You didn't clarify that it was these pieces, because your account
was incoherent, but I've since then figured out what you must have been
referring to.

Ruben Safir

2015-08-05 16:46:42 UTC

Because I need GNOME?

Well, Gnome does that applications which are useful and the sound
integration was decentbefore it went off the rails. I have had compains
adopt gnome screadsheets, their quicken like finance program, and video
editing programs.

Additionallym the network manager seems to succeed at the universities
wifi in a way that wicd just won't.

don't ask me why.

But forget Gnome. X11 needs an entire backport configuration to reroot
it and make it work without systemd. Is this hard? It is for me. It
is definetely a fork in the package design.

Post by Rick Moen
I clearly do not, so that can't be why. You
don't say what 'progressing' means, here. Cound that be because, I
dunno, just spitballing here, the facts aren't on your side?
It would take a massive uptake _outside_ GNOME of secondary
Freedesktop.org 'desktop' software infrastructure with systemd
dependencies -- PolKit, udisks2 -- before I would even perceive a
problem of desirable packages I cannot get without the code hairball.
And that's been predicted by you and some others for several years to be
imminent, but continues to just not happen.

I'm not sure what you mean with this paragraph, but there is an
additional program of insane UIID name lables et al for flash drives and
CDROMs that frankly make me crazy. I haven't had to ask where my fash
drive went for years. Now I have to pour over dmesg and with systemd
binary logs to find them

Post by Rick Moen
There have always been code hairballs in the Linux software ecosystem,
and people have always bitched about excessive package dependency trees.
Stupid ideas have frequently been lobbed at use from desktop space.
Remember when everything was going to need to work as object software in
coordination with CORBA brokers? Remember when everything was going to
run with root authority because users can't handle the concepts of
ownership and permissions?
Things have not been 'progressing' over the years you've been making bad
and poorly informed predictions.

I never made predictions. The only prediction I ever made was that the
GNU/Linux desktop as early as the late 1990'swas already superior to MS
Windows

and it still is

Post by Rick Moen
The sum and substance of this is, a couple of years ago, some drama
happened with you on OpenSUSE 12.x, and part of it involved inability to
boot because of some startup task problem and systemd wouldn't let you
get into the problem to recover the system. And then there was some
claim about how X.org moving to running X servers as a non-privileged
users was horrible because doing so required Freedesktop.org tangleware
pieces PolKit, systemd-logind, and upower[1], some of which require
systemd -- when it simply not the case that X needs to require those.
You cursed a blue streak, moved sideways from OpenSUSE to Manjaro, noted
that your various problems went away, blamed them all on systemd, and
started asserting that everyone else would soon be equally plagued.

I love it when you write about me in the third party.

You left out some essentle details. Like researching the problems and
discovering that there was a courus of anger at systemd for this cludge
of a mess they made by unifying all the subsystems under one umbrella
and lecturing everyone to get with the program or get out of the linux
world.

I. personally, have become very adapt at altering the configure and init
files for my systems, and created a fairly customized and working
environment for myself that has functioned for 15 years, and those tools
have been pulled out from under me with systemd. Can I adapt them to
systemd. Maybe? I don't want to learn the damn thing. It is a monster
and contantaly does what I don't expect. I don't want to be forced into
change for its own sake. I like systems KISS simple. I don't want go
to the suse mailing list and be ridiculed for running init scripts that
have functioned for DECADES. My email for help from lists had been
recieved by systemd proponents with childish ridicule and posturing.

Leanard Pottering and his whole Pancho Dia caverly can scratch my balls

Post by Rick Moen
And you keep saying that -- but it keeps not happening.
One of us isn't really following the details of what's been going on.
You can think it's me. I will take that view under advisement.
In case you want to understand the X server matter, here's a page by
some Arch Linux people who've documented how to do systemd-free X.org
even using MATE or Cinnamon. Note in particular the section about
'Rootless X': http://systemd-free.org/config.php
[1] You didn't clarify that it was these pieces, because your account
was incoherent, but I've since then figured out what you must have been
referring to.
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Rick Moen

2015-08-05 17:49:25 UTC

Post by Ruben Safir
Well, Gnome does that applications which are useful and the sound
integration was decent before it went off the rails.

Well, luckily for you, the problem isn't with GNOME applicatiosn but
rather with GNOME. Remember the list? gnome-bluetooth,
gnome-settings-daemon, gdm3, gnome-core, gnome-disk-utility, pcmanfm,
daisy-player, and a couple of other obscure apps that require
policykit-1 that in turn requires systemd.

Do you see Abiword in that list? Gnumeric? No, you don't. Like me,
you like to run those GNOME applications under Window Maker. And
there's nothing preventing you from running those and a dozen or two
other GNOME applications, because 'GNOME applicaition' means only that
it depends on gnome-libs and gtk, not that it necessarily requires
anything else. _GNOME_ as a framework has a problem, because things
like gnome-settings-daemon, gdm3, gnome-core, gnome-disk-utility have
dependency problems, but individual GNOME applications do not.

Honestly, was that actually difficult to figure out? Surely you've
understood for many years that each DE is merely a suite of X11
applications with a common look and feel on account of some libs and a
graphics toolkit they all use, and nothing prevents you from using those
applications on an a la carte basis without needing the whole DE.

Post by Ruben Safir
I have had compains adopt gnome screadsheets, their quicken like
finance program, and video editing programs.

If you had complaints about _them_, that has nothing to do with systemd.
But I suspect you're confused, and failing to understnad the difference
between GNOME as a whole and GNOME applications.

Fix that.

Post by Ruben Safir
Additionallym the network manager seems to succeed at the universities
wifi in a way that wicd just won't.
don't ask me why.

Once again, we get to remedial details:

NetworkManager is merely and purely a front-end to Wireless Tools for
Linux and wpa_supplicant. So is wicd. So is connman. (Well, as Marc
points out, NetworkManager's feature bloat means it also calls Bluetooth
network code.)

So, any WiFi networking connection that can be initiated by one can
be initiated by any of the others _or_ with no wireless network manager
at all. Yeah, _no_ wireless network manager: Remember iwconfig /
ifrename / iwgetid / iwlist / iwpriv / iwspy? They still work fine.

Not able to make iwconfig work? Figure it out. It's not significantly
more complex than iwconfig, and there's a perfeclty suitable man page.

Post by Ruben Safir
But forget Gnome. X11 needs an entire backport configuration to reroot
it and make it work without systemd.

No, it actually doesn't.

I really don't know what exact idiocy you got into with OpenSUSE, but I
have a running instance of Debian Jessie in front of me with package
xserver-common package version 1.16.4 , which is to say, current. It
looks to me that _if_ there _were_ a lack of root user privilege to
start the Xorg server -- which there is not -- then 'man 5
Xwrapper.config' would tell you wnat to put on the
/etc/X11/Xwrapper.config line to fix that.

Which is what I said upthread, except I'm saying it in more detail this
time.

I didn't need any existing expertise to know that, which is lucky
because I actually didn't know that -- until I did 30 seconds of Web
searching a couple of days ago.

Welcome to the Internet. Occasionally something needs 30 seconds of Web
searching.

Post by Ruben Safir
Is this hard? It is for me. It is definetely a fork in the package
design.

You keep using this phrase. I do not think it means what you think it
means.

Again, not certain what exact Xorg problem you ran into on OpenSUSE that
proximately resulted (along with your 'init scripts' thing) in your
running to Manjaro telling tales of terror about systemd, but all I'm
_aware_ of happening with 'rootless X' is that some Xorg guys figured
out about eight/ten years ago how Xorg _could_ be started without the
need to run that massive graphics engine with root-user authority --
which is A Good Thing, because the X server binary running as a big
root-privilege process has always been a security Sword of Damocles.
Nothing about that requires systemd, nor even PolKit, ConsoleKit,
upower, udisk2, and the rest of the Freedesktop.org brass band.

Maybe next time you have a technical problem, you should not overreact,
change distributions, and go around yelling on mailing lists, but rather
do what the rest of us do: Spend a little quality time with a search
engine, grep some logfiles, tinker a little, fix it.

Post by Ruben Safir
I'm not sure what you mean with this paragraph....

So, you're going to ignore what I said, talk past me, and change the
subject to something wildly different about use of UUIDs in partition
mounts -- which if you don't like them (as I don't), don't use them (as
I don't). Towards which end, you might want to turn off DE automounters
(as I do).

If you run DEs with their whole marching bands of processes that each
have their own ideas about how to run your system, don't act surprised
when your system is no longer being administered your way. Again, this
is Linux 101. We should not be needing to have that conversation on
linux-elitists in 2015.

Post by Ruben Safir
I never made predictions.

That bit about if things go the way they've been progressing was some
_other_ Ruben Safir?

Post by Ruben Safir
I love it when you write about me in the third party.

THe correct phrase is 'third person' (e.g., 'he'), but what I wrote
there was entirely second person (e.g., 'you'). As the saying goes, you
could look it up.

Post by Ruben Safir
You left out some essentle details.

Rick Moen

2015-08-05 18:01:23 UTC

Post by Rick Moen
Not able to make iwconfig work? Figure it out. It's not significantly
more complex than iwconfig, and there's a perfeclty suitable man page.

^^^^^^^^ ifconfig

(And now I'm off to make myself some coffee.)

Ruben Safir

2015-08-05 23:18:59 UTC

I need tpo print this to keep up with this :)

Post by Ruben Safir
Well, Gnome does that applications which are useful and the sound
integration was decent before it went off the rails.

Well, luckily for you, the problem isn't with GNOME applicatiosn but
rather with GNOME. Remember the list? gnome-bluetooth,
gnome-settings-daemon, gdm3, gnome-core, gnome-disk-utility, pcmanfm,
daisy-player, and a couple of other obscure apps that require
policykit-1 that in turn requires systemd.
Do you see Abiword in that list? Gnumeric? No, you don't. Like me,
you like to run those GNOME applications under Window Maker. And
there's nothing preventing you from running those and a dozen or two
other GNOME applications, because 'GNOME applicaition' means only that
it depends on gnome-libs and gtk, not that it necessarily requires
anything else. _GNOME_ as a framework has a problem, because things
like gnome-settings-daemon, gdm3, gnome-core, gnome-disk-utility have
dependency problems, but individual GNOME applications do not.
Honestly, was that actually difficult to figure out? Surely you've
understood for many years that each DE is merely a suite of X11
applications with a common look and feel on account of some libs and a
graphics toolkit they all use, and nothing prevents you from using those
applications on an a la carte basis without needing the whole DE.

Post by Ruben Safir
I have had compains adopt gnome screadsheets, their quicken like
finance program, and video editing programs.

Post by Ruben Safir
Additionallym the network manager seems to succeed at the universities
wifi in a way that wicd just won't.
don't ask me why.

NetworkManager is merely and purely a front-end to Wireless Tools for
Linux and wpa_supplicant. So is wicd. So is connman. (Well, as Marc
points out, NetworkManager's feature bloat means it also calls Bluetooth
network code.)
So, any WiFi networking connection that can be initiated by one can
be initiated by any of the others _or_ with no wireless network manager
at all. Yeah, _no_ wireless network manager: Remember iwconfig /
ifrename / iwgetid / iwlist / iwpriv / iwspy? They still work fine.
Not able to make iwconfig work? Figure it out. It's not significantly
more complex than iwconfig, and there's a perfeclty suitable man page.

Post by Ruben Safir
But forget Gnome. X11 needs an entire backport configuration to reroot
it and make it work without systemd.

No, it actually doesn't.
I really don't know what exact idiocy you got into with OpenSUSE, but I
have a running instance of Debian Jessie in front of me with package
xserver-common package version 1.16.4 , which is to say, current. It
looks to me that _if_ there _were_ a lack of root user privilege to
start the Xorg server -- which there is not -- then 'man 5
Xwrapper.config' would tell you wnat to put on the
/etc/X11/Xwrapper.config line to fix that.
Which is what I said upthread, except I'm saying it in more detail this
time.
I didn't need any existing expertise to know that, which is lucky
because I actually didn't know that -- until I did 30 seconds of Web
searching a couple of days ago.
Welcome to the Internet. Occasionally something needs 30 seconds of Web
searching.

Post by Ruben Safir
Is this hard? It is for me. It is definetely a fork in the package
design.

You keep using this phrase. I do not think it means what you think it
means.
Again, not certain what exact Xorg problem you ran into on OpenSUSE that
proximately resulted (along with your 'init scripts' thing) in your
running to Manjaro telling tales of terror about systemd, but all I'm
_aware_ of happening with 'rootless X' is that some Xorg guys figured
out about eight/ten years ago how Xorg _could_ be started without the
need to run that massive graphics engine with root-user authority --
which is A Good Thing, because the X server binary running as a big
root-privilege process has always been a security Sword of Damocles.
Nothing about that requires systemd, nor even PolKit, ConsoleKit,
upower, udisk2, and the rest of the Freedesktop.org brass band.
Maybe next time you have a technical problem, you should not overreact,
change distributions, and go around yelling on mailing lists, but rather
do what the rest of us do: Spend a little quality time with a search
engine, grep some logfiles, tinker a little, fix it.

Post by Ruben Safir
I'm not sure what you mean with this paragraph....

So, you're going to ignore what I said, talk past me, and change the
subject to something wildly different about use of UUIDs in partition
mounts -- which if you don't like them (as I don't), don't use them (as
I don't). Towards which end, you might want to turn off DE automounters
(as I do).
If you run DEs with their whole marching bands of processes that each
have their own ideas about how to run your system, don't act surprised
when your system is no longer being administered your way. Again, this
is Linux 101. We should not be needing to have that conversation on
linux-elitists in 2015.

Post by Ruben Safir
I never made predictions.

That bit about if things go the way they've been progressing was some
_other_ Ruben Safir?

Post by Ruben Safir
I love it when you write about me in the third party.

there was entirely second person (e.g., 'you'). As the saying goes, you
could look it up.

Post by Ruben Safir
You left out some essentle details.

Quite probably because you never provided a coherent factual account --
and you still haven't.
And, by the way, the bit about how you could no longer run tried and
tested init scripts doesn't make sense, either, because systemd-based
distributions include the ability to parse and run SysVInit scripts.
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

James Morris

2015-08-12 04:26:31 UTC

Post by Ruben Safir
I need tpo print this to keep up with this :)

ARE YOU FROM THE PAST?

--
James Morris
<***@namei.org>

Don Marti

2015-08-03 17:32:15 UTC

Post by Rick Moen
Right, let's make this interesting: I am offering to wager US $100 that
packages for _both_ rsyslog (or a similar successor) _and_ syslog-ng (or
a similar successor) will still be trivially installable on Debian,
usable on Debian, and developer-maintained on August 2, 2018. If you're
confident of what you say, this will be easy money for you. For my
part, on the basis of professional experience as a senior sysadmin, I'm
confident you're profoundly mistaken, and am willing to take your money.
Wager offer is extended for 48 hours from the timestamp on this posting,
and you'll need to indicate acceptance on this mailing list. (In the
event of dispute about the state of affairs on August 2, 2018, we'll
defer to Mr. Don Marti's judgement if he's willing.)

I'm willing to act as referee. If there's any
question about whether these two packages actually
work, I will install a Debian VM, use apt-get (or
the documented successor tool) to install rsyslog,
run a few commands to show that it works, and then
do the same for syslog-ng.

--
Don Marti <***@zgp.org>
http://zgp.org/~dmarti/
Are you safe from 3rd-party web tracking? http://www.aloodo.org/test/

Rick Moen

2015-08-03 20:10:20 UTC

Post by Don Marti
I'm willing to act as referee. If there's any
question about whether these two packages actually
work, I will install a Debian VM, use apt-get (or
the documented successor tool) to install rsyslog,
run a few commands to show that it works, and then
do the same for syslog-ng.

If you like pastrami, I can save you some. And a pickle.

--
Cheers, "The trouble ain't that there is too many fools,
Rick Moen but that the lightning ain't distributed right."
***@linuxmafia.com -- Mark Twain
McQ! (4x80)

Eugen Leitl

2015-08-01 12:53:38 UTC

These days, the elitist's Linux is *BSD.

At least on the desktop. I still use Debian server-side, and
if I'm going to have to learn OpenStack I'll be still sticking
there for a while.

But ceph on top of zfs does have a nice ring to it.

Bob Bernstein

2015-08-02 05:15:56 UTC

Post by Eugen Leitl
These days, the elitist's Linux is *BSD.

Here is the lamest (most lame?) story any one of you will hear um
this weekend:

I have been a devout fan of bsd's for years, and settled down some
time ago to tithe, as it were, at NetBSD. A few months back I ran
out of the old i686ish boxes I used as bsd playgrounds, so I gave
Staples $100 for a refurb HP. It worked fine but would not boot a
bsd kernel, period. I tried Net-, Open-, and FreeBSD. I lerned this
perversion is a wrinkle in the HP bios, and yes, it can be
circumvented, some say, with a fair amount of tricky business but I
am not up to that sort of thing anymore. Since the gods still love
a good joke the HP refurb booted and installed Debian like old
friends meeting in front of their once-favorite bar.

Now I sit here next to a Jessie machine that runs like a clock but I
have to say the thrill is gone. I don't know how to do anything on
this machine the "right way." On debian-user I learn things such as
"ifconfig is obsolete." So is ntpdate-debian, according to that
venue. They both appear to my untutored eye to work just fine. (Do
you think I should tell them that over there on debian-user?)

So at my earliest convenience I will have the new NetBSD-7.0 running
here on something besides the VMplayer on my old $300 Walmart
eMachine screamer running Windows 7 (which also has run like a
clock).

Thanks all!

--
Bob Bernstein

"No matter how big the problem is, you can always run away from it."

Dom Irrera

Ruben Safir

2015-08-02 06:14:18 UTC

Post by Bob Bernstein

Post by Eugen Leitl
These days, the elitist's Linux is *BSD.

Here is the lamest (most lame?) story any one of you will hear um
I have been a devout fan of bsd's for years, and settled down some
time ago to tithe, as it were, at NetBSD. A few months back I ran
out of the old i686ish boxes I used as bsd playgrounds, so I gave
Staples $100 for a refurb HP. It worked fine but would not boot a
bsd kernel, period. I tried Net-, Open-, and FreeBSD. I lerned this
perversion is a wrinkle in the HP bios, and yes, it can be
circumvented, some say, with a fair amount of tricky business but I
am not up to that sort of thing anymore. Since the gods still love
a good joke the HP refurb booted and installed Debian like old
friends meeting in front of their once-favorite bar.
Now I sit here next to a Jessie machine that runs like a clock but I
have to say the thrill is gone. I don't know how to do anything on
this machine the "right way." On debian-user I learn things such as
"ifconfig is obsolete." So is ntpdate-debian, according to that
venue. They both appear to my untutored eye to work just fine. (Do
you think I should tell them that over there on debian-user?)
So at my earliest convenience I will have the new NetBSD-7.0 running
here on something besides the VMplayer on my old $300 Walmart
eMachine screamer running Windows 7 (which also has run like a
clock).
Thanks all!

Nothing is obsolete if it is useful and there is nothing useful about
changing how a system works...just to say, oh it is changed and it must
be better if it is new. Putting experienced users through the torture
of having to relearn everything is just stupid.

try manjaro's openrc port

Post by Bob Bernstein
--
Bob Bernstein
"No matter how big the problem is, you can always run away from it."
Dom Irrera
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
http://zgp.org/cgi-bin/mailman/listinfo/linux-elitists

Rick Moen

2015-08-02 07:15:10 UTC

Post by Bob Bernstein
On debian-user I learn things such as
"ifconfig is obsolete."

$ alias ifconfig='echo /sbin/ifconfig is deprecated, use /bin/ip instead'

There. I fixed it for you.

net-tools for Linux (furnishing ifconfig) was last updated 2001.
More such problems here: http://inai.de/2008/02/19

Example usage, /bin/ip's equivalent of '/sbin/ifconfig -a', where the
'addr' subcommand has been abbreviated to 'a':

$ ip a show
1: lo: <LOOPBACK,UP,LOWER_UP> mtu 16436 qdisc noqueue state UNKNOWN
link/loopback 00:00:00:00:00:00 brd 00:00:00:00:00:00
inet 127.0.0.1/8 scope host lo
inet6 ::1/128 scope host
valid_lft forever preferred_lft forever
2: eth0: <BROADCAST,MULTICAST,UP,LOWER_UP> mtu 1500 qdisc pfifo_fast state UP qlen 1000
link/ether 00:d0:b7:a7:a4:ec brd ff:ff:ff:ff:ff:ff
inet 198.144.195.186/29 brd 198.144.195.191 scope global eth0
inet6 fe80::2d0:b7ff:fea7:a4ec/64 scope link
valid_lft forever preferred_lft forever
$

But sure, you won't be alone in using ifconfig, route, arp, and netstat
out of habit, especially if you admin other Unices, too. (It took me
a while to stop using nslookup and consistently use dig, instead, too.)

Post by Bob Bernstein
So is ntpdate-debian, according to that venue.

It'll probably vanish upstream eventually, so start getting used to
invoking ntpd (see options -p -n -q), and definitely stop using ntpdate
in new scripts and start updating old ones.

Post by Bob Bernstein
They both appear to my untutored eye to work just fine.

$ dict deprecated
[...]
From The Free On-line Dictionary of Computing (20 July 2014) [foldoc]:

deprecated

Said of a program or feature that is considered obsolescent
and in the process of being phased out, usually in favour of a
specified replacement. Deprecated features can,
unfortunately, linger on for many years.
[...]

--
Cheers, "I know you believe you understood what you think I said,
Rick Moen but I'm not sure you realize that what you heard is not
***@linuxmafia.com what I meant." -- S.I. Hayakawa
McQ! (4x80)

Greg KH

2015-08-01 16:47:09 UTC

Why would you want such a thing?

Rick Moen

2015-08-01 18:08:57 UTC

Why would you want such a thing?

To insult you personally, presumably. That's the only answer that could
possibly make sense.

http://www.linux.com/news/featured-blogs/200-libby-clark/771055-fun-photo-greg-kroah-hartman-crowned-at-the-systemd-hack-fest/

About GNOME Network Manager, I used to have a list of the various
less-baroque and less dependency-ridden alternative wireless network
managers (starting with wicd, connman, netifd, wpa_supplicant/wpa_cli,
wifi-wiz, getwifi, wlassistant, wifi-radar, GTKWifi, ceni, netctl), but
most often (personally) I'd rather just use iwconfig / ifrename /
iwgetid / iwlist / iwpriv / iwspy ('Wireless Tools for Linux').

Marcel Gagné once made a good argument for Wireless Assistant and netGo
(neither yet tried) to manage multiple wireless profiles, so those merit
a try, too: http://www.linuxjournal.com/node/8355/print
_______________________________________________
Do not Cc: anyone else on mail sent to this list. The list server is set for maximum one recipient.
linux-elitists mailing list
linux-***@zgp.org
http://zgp.org/cgi-bin/mail

Ruben Safir

2015-08-01 23:27:44 UTC