[ale] please bow your head for a moment of silence...

Discussion:

Chuck Payne

2017-09-06 06:02:20 UTC

Not sure if anyone saw the post, but Oracle laidoff all the Solaris people,
a lot of the articles are calling it the final nail in Sun's coffin.

As a Solaris admin, I am a bit sad, I remember when the user group went
away after the sale of Sun to Oracle. I also remember the end of
openSolaris (please note there were three distro of openSolaris at the time
of it''s end).

You know more to me Oracle is a technology killer than technology innovator
it's really sad I'm going to miss Solaris. Plus, Sparc computers. I used
Ultra25 for three yeara, great machine.

SunOS is dead, long live SunOS (Solaris).

Jim Kinney

2017-09-06 11:39:08 UTC

Permalink

What's left besides AIX? Carly pulled the plug on HPUX a long time back. IRIX (zombie)?

Oh. That never gonna amount to much more than an experiment little upstart called Linux.

I won't miss slowos but I feel for the newly unemployed.

Post by Chuck Payne
Not sure if anyone saw the post, but Oracle laidoff all the Solaris people,
a lot of the articles are calling it the final nail in Sun's coffin.
As a Solaris admin, I am a bit sad, I remember when the user group went
away after the sale of Sun to Oracle. I also remember the end of
openSolaris (please note there were three distro of openSolaris at the time
of it''s end).
You know more to me Oracle is a technology killer than technology innovator
it's really sad I'm going to miss Solaris. Plus, Sparc computers. I used
Ultra25 for three yeara, great machine.
SunOS is dead, long live SunOS (Solaris).

--
Sent from my Android device with K-9 Mail. All tyopes are thumb related and reflect authenticity.

leam hall

2017-09-06 11:53:36 UTC

Permalink

Given the Linux adoption of systemd the only options seem to be "Winderz
with a GUI", "Winderz with a command line", "Expensive Fruit", or some
flavor of BSD.

Post by Jim Kinney
What's left besides AIX? Carly pulled the plug on HPUX a long time back. IRIX (zombie)?
Oh. That never gonna amount to much more than an experiment little upstart called Linux.
I won't miss slowos but I feel for the newly unemployed.

Post by Chuck Payne
Not sure if anyone saw the post, but Oracle laidoff all the Solaris
people, a lot of the articles are calling it the final nail in Sun's coffin.
As a Solaris admin, I am a bit sad, I remember when the user group went
away after the sale of Sun to Oracle. I also remember the end of
openSolaris (please note there were three distro of openSolaris at the time
of it''s end).
You know more to me Oracle is a technology killer than technology
innovator it's really sad I'm going to miss Solaris. Plus, Sparc computers.
I used Ultra25 for three yeara, great machine.
SunOS is dead, long live SunOS (Solaris).

--
Sent from my Android device with K-9 Mail. All tyopes are thumb related
and reflect authenticity.
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Chuck Payne

2017-09-06 15:08:57 UTC

Permalink

Post by leam hall
Given the Linux adoption of systemd the only options seem to be "Winderz
with a GUI", "Winderz with a command line", "Expensive Fruit", or some
flavor of BSD.

Post by Chuck Payne
Not sure if anyone saw the post, but Oracle laidoff all the Solaris
people, a lot of the articles are calling it the final nail in Sun's coffin.
As a Solaris admin, I am a bit sad, I remember when the user group went
away after the sale of Sun to Oracle. I also remember the end of
openSolaris (please note there were three distro of openSolaris at the time
of it''s end).
You know more to me Oracle is a technology killer than technology
innovator it's really sad I'm going to miss Solaris. Plus, Sparc computers.
I used Ultra25 for three yeara, great machine.
SunOS is dead, long live SunOS (Solaris).

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

--
Terror PUP a.k.a
Chuck "PUP" Payne
-----------------------------------------
Discover it! Enjoy it! Share it! openSUSE Linux.
-----------------------------------------
openSUSE -- Terrorpup
openSUSE Ambassador/openSUSE Member
skype,twiiter,identica,friendfeed -- terrorpup
freenode(irc) --terrorpup/lupinstein
Register Linux Userid: 155363

Have you tried SUSE Studio? Need to create a Live CD, an app you want to
package and distribute , or create your own linux distro. Give SUSE Studio
a try.

Jerald Sheets

2017-09-06 15:15:58 UTC

Permalink

A lot of folks have seen the writing on the wall up at LOPSA. We had some hallway meetings trying to figure out what the next trajectory should be for âbig machineâ needs.

A lot of talk around HPC and a lot of talk around SSI Linux-based clusters and a BOATLOAD of testing.

Weâre at an inflection point, folks. Curious to see whatâs next. Iâll just keep using Linux. :D

âj

I was surprised, in my job hunt, over the last month I have five jobs, in the Atlanta area looking for a Solaris Admin.
I guess Solaris is used by a lot of banking, like AS/400 machines ( sorry I know that the same OS to Hardware ).
Given the Linux adoption of systemd the only options seem to be "Winderz with a GUI", "Winderz with a command line", "Expensive Fruit", or some flavor of BSD.
What's left besides AIX? Carly pulled the plug on HPUX a long time back. IRIX (zombie)?
Oh. That never gonna amount to much more than an experiment little upstart called Linux.
I won't miss slowos but I feel for the newly unemployed.
Not sure if anyone saw the post, but Oracle laidoff all the Solaris people, a lot of the articles are calling it the final nail in Sun's coffin.
As a Solaris admin, I am a bit sad, I remember when the user group went away after the sale of Sun to Oracle. I also remember the end of openSolaris (please note there were three distro of openSolaris at the time of it''s end).
You know more to me Oracle is a technology killer than technology innovator it's really sad I'm going to miss Solaris. Plus, Sparc computers. I used Ultra25 for three yeara, great machine.
SunOS is dead, long live SunOS (Solaris).
--
Sent from my Android device with K-9 Mail. All tyopes are thumb related and reflect authenticity.
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale <http://mail.ale.org/mailman/listinfo/ale>
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo <http://mail.ale.org/mailman/listinfo>
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale <http://mail.ale.org/mailman/listinfo/ale>
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo <http://mail.ale.org/mailman/listinfo>
--
Terror PUP a.k.a
Chuck "PUP" Payne
-----------------------------------------
Discover it! Enjoy it! Share it! openSUSE Linux.
-----------------------------------------
openSUSE -- Terrorpup
openSUSE Ambassador/openSUSE Member
skype,twiiter,identica,friendfeed -- terrorpup
freenode(irc) --terrorpup/lupinstein
Register Linux Userid: 155363
Have you tried SUSE Studio? Need to create a Live CD, an app you want to package and distribute , or create your own linux distro. Give SUSE Studio a try.
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Lightner, Jeffrey

2017-09-06 15:35:35 UTC

Permalink

The article about the layoffs:
https://www.theregister.co.uk/2017/09/04/oracle_layoffs_solaris_sparc_teams/

Carly didnât kill HP-UX â it was alive and (mostly) well after she left. What they did kill is the PA-RISC chip in favor of Itanium. I recently saw a job posting requesting HP-UX but it was the first I saw in some time but truth be told HP-UX was always a niche admin position based on all my job searches over the years. Oddly enough it was Oracle that tried to kill HP-UX after they bought Sun by announcing they wouldnât support Itanium any longer. I *think* HP successfully sued them over that and they were required to continue support for another 10 years but donât recall how long ago that was.

I too see a Solaris being requested frequently. I see more for AIX.

A few years back I turned down a friends offer to present me to his company for another job because I didnât want to do âonlyâ Linux. Now Iâd not consider any job that isnât primarily Linux. I liked UNIX (HP-UX most of all) but most people have been migrating to Linux off of UNIX for several years. As of last year we quit using our last UNIX systems when we retired the payroll software that required it.

From: ale-***@ale.org [mailto:ale-***@ale.org] On Behalf Of Chuck Payne
Sent: Wednesday, September 06, 2017 11:09 AM
Cc: Atlanta Linux Enthusiasts
Subject: Re: [ale] please bow your head for a moment of silence...

I was surprised, in my job hunt, over the last month I have five jobs, in the Atlanta area looking for a Solaris Admin.

I guess Solaris is used by a lot of banking, like AS/400 machines ( sorry I know that the same OS to Hardware ).

On Wed, Sep 6, 2017 at 7:53 AM, leam hall <***@gmail.com<mailto:***@gmail.com>> wrote:
Given the Linux adoption of systemd the only options seem to be "Winderz with a GUI", "Winderz with a command line", "Expensive Fruit", or some flavor of BSD.

On Wed, Sep 6, 2017 at 7:39 AM, Jim Kinney <***@gmail.com<mailto:***@gmail.com>> wrote:
What's left besides AIX? Carly pulled the plug on HPUX a long time back. IRIX (zombie)?

Oh. That never gonna amount to much more than an experiment little upstart called Linux.

I won't miss slowos but I feel for the newly unemployed.
On September 6, 2017 2:02:20 AM EDT, Chuck Payne <***@gmail.com<mailto:***@gmail.com>> wrote:
Not sure if anyone saw the post, but Oracle laidoff all the Solaris people, a lot of the articles are calling it the final nail in Sun's coffin.

As a Solaris admin, I am a bit sad, I remember when the user group went away after the sale of Sun to Oracle. I also remember the end of openSolaris (please note there were three distro of openSolaris at the time of it''s end).

You know more to me Oracle is a technology killer than technology innovator it's really sad I'm going to miss Solaris. Plus, Sparc computers. I used Ultra25 for three yeara, great machine.

SunOS is dead, long live SunOS (Solaris).
--
Sent from my Android device with K-9 Mail. All tyopes are thumb related and reflect authenticity.

_______________________________________________
Ale mailing list
***@ale.org<mailto:***@ale.org>
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

_______________________________________________
Ale mailing list
***@ale.org<mailto:***@ale.org>
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo
--
Terror PUP a.k.a
Chuck "PUP" Payne
-----------------------------------------
Discover it! Enjoy it! Share it! openSUSE Linux.
-----------------------------------------
openSUSE -- Terrorpup
openSUSE Ambassador/openSUSE Member
skype,twiiter,identica,friendfeed -- terrorpup
freenode(irc) --terrorpup/lupinstein
Register Linux Userid: 155363

Have you tried SUSE Studio? Need to create a Live CD, an app you want to package and distribute , or create your own linux distro. Give SUSE Studio a try.

Steve Litt

2017-09-07 05:46:37 UTC

Permalink

On Wed, 6 Sep 2017 07:53:36 -0400

Post by leam hall
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.

That being said, several Linux distros don't use systemd and are
committed to not using systemd:

* Devuan
* Void
* Funtoo

There are probably seven or eight more that currently don't use
systemd, or don't use it by default, but have not foresworn the
possibility.

Except for those who "simply must" have their Gnome, KDE or
NetworkManager, there are plenty of Linux systems out there they can
run without systemd.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Lightner, Jeffrey

2017-09-07 12:29:46 UTC

Permalink

Caveman conversation:
Ug: What that?
Zog: Wheel.
Ug: Why wheel? Drag work for years.
Zog: More fast to use wheel.
Ug: Wheel made by false god to trap draggers. It bad.
Ug then clubs Zog because Zog doesn't see the intrinsic "reason" of Ug's opinion.

Move ahead 10,000 years:
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...

:p

-----Original Message-----
From: ale-***@ale.org [mailto:ale-***@ale.org] On Behalf Of Steve Litt
Sent: Thursday, September 07, 2017 1:47 AM
To: ***@ale.org
Subject: Re: [ale] please bow your head for a moment of silence...

On Wed, 6 Sep 2017 07:53:36 -0400

Post by leam hall
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.

That being said, several Linux distros don't use systemd and are committed to not using systemd:

* Devuan
* Void
* Funtoo

There are probably seven or eight more that currently don't use systemd, or don't use it by default, but have not foresworn the possibility.

Except for those who "simply must" have their Gnome, KDE or NetworkManager, there are plenty of Linux systems out there they can run without systemd.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical Troubleshooting Brand new, second edition http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jim Kinney

2017-09-07 14:18:44 UTC

Permalink

BWAHAHAHAHAHAHAHA!!!!!!!!

Post by Lightner, Jeffrey
Ug: What that?
Zog: Wheel.
Ug: Why wheel? Drag work for years.
Zog: More fast to use wheel.
Ug: Wheel made by false god to trap draggers. It bad.
Ug then clubs Zog because Zog doesn't see the intrinsic "reason" of Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...
:p
-----Original Message-----
Sent: Thursday, September 07, 2017 1:47 AM
Subject: Re: [ale] please bow your head for a moment of silence...
On Wed, 6 Sep 2017 07:53:36 -0400

Post by leam hall
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.

That being said, several Linux distros don't use systemd and are
* Devuan
* Void
* Funtoo
There are probably seven or eight more that currently don't use
systemd, or don't use it by default, but have not foresworn the
possibility.
Except for those who "simply must" have their Gnome, KDE or
NetworkManager, there are plenty of Linux systems out there they can
run without systemd.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition http://www.troubleshooters.
com/mgr
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

James Sumners

2017-09-07 14:35:30 UTC

Permalink

Meanwhile...
https://utcc.utoronto.ca/~cks/space/blog/linux/SystemdUbuntuRebootFailure

"We just went through a periodic exercise of rebooting all of our
<https://support.cs.toronto.edu/> Ubuntu servers in order to get up to date
on kernels and so on. By now almost all of our servers are running Ubuntu
16.04, which means that they're using systemd. Unfortunately this gives us
a real problem, because *on Ubuntu 16.04, systemd won't reliably reboot
your system*. On some servers, usually the busiest and most important ones,
the system will just stop during the shutdown process and sit there. And
sit there. And sit there. Perhaps it would eventually recover after tens of
minutes, but as mentioned these are generally our busiest and most
important servers, so we're not exactly going to let them sit there to find
out what happens eventually."

Post by Jim Kinney
BWAHAHAHAHAHAHAHA!!!!!!!!
Ug: What that?
Zog: Wheel.
Ug: Why wheel? Drag work for years.
Zog: More fast to use wheel.
Ug: Wheel made by false god to trap draggers. It bad.
Ug then clubs Zog because Zog doesn't see the intrinsic "reason" of Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...
:p
-----Original Message-----
Sent: Thursday, September 07, 2017 1:47 AM
Subject: Re: [ale] please bow your head for a moment of silence...
On Wed, 6 Sep 2017 07:53:36 -0400
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.
* Devuan
* Void
* Funtoo
There are probably seven or eight more that currently don't use systemd, or don't use it by default, but have not foresworn the possibility.
Except for those who "simply must" have their Gnome, KDE or NetworkManager, there are plenty of Linux systems out there they can run without systemd.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical Troubleshooting Brand new, second edition http://www.troubleshooters.com/mgr
_______________________________________________
See JOBS, ANNOUNCE and SCHOOLS lists athttp://mail.ale.org/mailman/listinfo
_______________________________________________
See JOBS, ANNOUNCE and SCHOOLS lists athttp://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

--
James Sumners
http://james.sumners.info/ (technical profile)
http://jrfom.com/ (personal site)
http://haplo.bandcamp.com/ (music)

Jim Kinney

2017-09-07 15:03:06 UTC

Permalink

And that is not so much a failure of systemd as it is a failure of the
distro.
The shutdown process requires that filesystems get unmounted. If a
process is trying to write to a log file while that log space is trying
to get unmounted, hang! Sucks using that in a server environment
without an addressable PDU.
Is it the fault of the shutdown command or the fault of the people who
wrote the logic that's failing? That logic layer is hinted at by
systemd but is supposed to be fleshed out by the distro writers.
So, um, yeah, Ubuntu dropped the ball. Fedora has some hangs as well in
the same manner. CentOS 7.3 does not. I don't count my Fedora systems
as "Server OS" although it could work with some tweaks. But Ubuntu
16.04 is supposed to be the LTS release and thus "server class".
That's (another) FAIL in the Ubuntu scorecard for me.
(I'm happy to throw gasoline on a fire of why some distros are better
suited than others for certain needs :-)
Just because someone misuses an adjustable wrench and claims that makes
it a lousy screwdriver doesn't make it a bad adjustable wrench.

Meanwhile... https://utcc.utoronto.ca/~cks/space/blog/linux/SystemdUb
untuRebootFailure
"We just went through a periodic exercise of rebooting all
of our Ubuntu servers in order to get up to date on kernels and so
on. By now almost all of our servers are running Ubuntu 16.04, which
means that they're using systemd. Unfortunately this gives us a real
problem, because on Ubuntu 16.04, systemd won't reliably reboot your
system. On some servers, usually the busiest and most important ones,
the system will just stop during the shutdown process and sit there.
And sit there. And sit there. Perhaps it would eventually recover
after tens of minutes, but as mentioned these are generally our
busiest and most important servers, so we're not exactly going to let
them sit there to find out what happens eventually."

Post by Jim Kinney
BWAHAHAHAHAHAHAHA!!!!!!!!

Post by leam hall
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line",
"Expensive
Fruit", or some flavor of BSD.

That being said, several Linux distros don't use systemd and are
* Devuan
* Void
* Funtoo
There are probably seven or eight more that currently don't use
systemd, or don't use it by default, but have not foresworn the
possibility.
Except for those who "simply must" have their Gnome, KDE or
NetworkManager, there are plenty of Linux systems out there they
can run without systemd.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition http://www.troubleshoot
ers.com/mgr
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Adam Jimerson

2017-09-07 15:26:13 UTC

Permalink

I would also like to add on to Jim's comment there is/was a bug with the
Watchdog process (see https://bugzilla.redhat.com/show_bug.cgi?id=1365352 just
happened to be one of the first links I found in a quick Google search)
this is an issue with a kernel driver that is just being exposed by SystemD
(thus making it look like SystemD is causing things to hang).

Just my $0.02 from an Arch user and BSD developer that mostly lurks here,
back to your normally scheduled rants.

Post by Jim Kinney
And that is not so much a failure of systemd as it is a failure of the
distro.
The shutdown process requires that filesystems get unmounted. If a process
is trying to write to a log file while that log space is trying to get
unmounted, hang! Sucks using that in a server environment without an
addressable PDU.
Is it the fault of the shutdown command or the fault of the people who
wrote the logic that's failing? That logic layer is hinted at by systemd
but is supposed to be fleshed out by the distro writers.
So, um, yeah, Ubuntu dropped the ball. Fedora has some hangs as well in
the same manner. CentOS 7.3 does not. I don't count my Fedora systems as
"Server OS" although it could work with some tweaks. But Ubuntu 16.04 is
supposed to be the LTS release and thus "server class". That's (another)
FAIL in the Ubuntu scorecard for me.
(I'm happy to throw gasoline on a fire of why some distros are better
suited than others for certain needs :-)
Just because someone misuses an adjustable wrench and claims that makes it
a lousy screwdriver doesn't make it a bad adjustable wrench.
Meanwhile...
https://utcc.utoronto.ca/~cks/space/blog/linux/SystemdUbuntuRebootFailure
"We just went through a periodic exercise of rebooting all of our
<https://support.cs.toronto.edu/> Ubuntu servers in order to get up to
date on kernels and so on. By now almost all of our servers are running
Ubuntu 16.04, which means that they're using systemd. Unfortunately this
gives us a real problem, because *on Ubuntu 16.04, systemd won't reliably
reboot your system*. On some servers, usually the busiest and most
important ones, the system will just stop during the shutdown process and
sit there. And sit there. And sit there. Perhaps it would eventually
recover after tens of minutes, but as mentioned these are generally our
busiest and most important servers, so we're not exactly going to let them
sit there to find out what happens eventually."
BWAHAHAHAHAHAHAHA!!!!!!!!
Ug: What that?
Zog: Wheel.
Ug: Why wheel? Drag work for years.
Zog: More fast to use wheel.
Ug: Wheel made by false god to trap draggers. It bad.
Ug then clubs Zog because Zog doesn't see the intrinsic "reason" of Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...
:p
-----Original Message-----
Sent: Thursday, September 07, 2017 1:47 AM
Subject: Re: [ale] please bow your head for a moment of silence...
On Wed, 6 Sep 2017 07:53:36 -0400
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.
* Devuan
* Void
* Funtoo
There are probably seven or eight more that currently don't use systemd, or don't use it by default, but have not foresworn the possibility.
Except for those who "simply must" have their Gnome, KDE or NetworkManager, there are plenty of Linux systems out there they can run without systemd.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical Troubleshooting Brand new, second edition http://www.troubleshooters.com/mgr
_______________________________________________
See JOBS, ANNOUNCE and SCHOOLS lists athttp://mail.ale.org/mailman/listinfo
_______________________________________________
See JOBS, ANNOUNCE and SCHOOLS lists athttp://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Steve Litt

2017-09-08 07:10:59 UTC

Permalink

On Thu, 7 Sep 2017 12:29:46 +0000

Let us count the falacies:

* Appeal to novelty: Being new, in and of itself, doesn't make
something better (or worse) than what came before.
https://en.wikipedia.org/wiki/Appeal_to_novelty

* Ad hominem: Painting as superstitious change-haters those who don't
like systemd doesn't in any way prove systemd is good.
https://en.wikipedia.org/wiki/Ad_hominem

* False choice: Implying that the only alternative to systemd is
"init" (I think he means sysvinit) is wrong. I know of at least five
additional init systems that are excellent, and use one of them every
day (runit). https://en.wikipedia.org/wiki/False_dilemma

I'll give you this: The caveman analogy is funny, it's clever, and it
greatly appeals to those already on your side. But it actually says
nothing.

The real crackup is that these same falacies pop up in almost every
defense of systemd, starting with LP himself. Read PID EINS, you'll see
what I mean.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jerald Sheets

2017-09-08 11:28:45 UTC

Permalink

I avoid all the BS and just hate it for its design, layout, and intrusion into all sorts of things it shouldnât be fiddling around in, breaking a myriad of tenets of the âUNIX wayâ.

The âUNIX Wayâ is to have a tool that does one thing, does it very well, has clearly defined input and output and doesnât try to handle multiple responsibility domains.

"This is the Unix philosophy: Write programs that do one thing and do it well. Write programs to work together. Write programs to handle text streams, because that is a universal interface.â â Doug McIlroy, the inventor of UNIX pipes

This is why grep doesnât awk and vice-versa. In case weâve forgotten:

The Rule of Modularity:
Write simple parts connected by clean interfaces.
Rule of Clarity:
Clarity is better than cleverness.
Rule of Composition:
Design programs to be connected with other programs.
Rule of Separation:
Separate policy from mechanism; separate interfaces from engines.
Rule of Simplicity:
Design for simplicity; add complexity only where you must.
Rule of Parsimony:
Write a big program only when it is clear by demonstration that nothing else will do.
Rule of Transparency:
Design for visibility to make inspection and debugging easier.
Rule of Robustness:
Robustness is the child of transparency and simplicity.
Rule of Representation:
Fold knowledge into data, so program logic can be stupid and robust.
Rule of Least Surprise:
In interface design, always do the least surprising thing.
Rule of Silence:
When a program has nothing surprising to say, it should say nothing.
Rule of Repair:
Repair what you can â but when you must fail, fail noisily and as soon as possible.
Rule of Economy:
Programmer time is expensive; conserve it in preference to machine time.
Rule of Generation:
Avoid hand-hacking; write programs to write programs when you can.
Rule of Optimization:
Prototype before polishing, Get it working before you optimize it.
Rule of Diversity:
Distrust all claims for one true way.
Rule of Extensibility:
Design for the future, because it will be here sooner than you think.

Iâll stick with what has worked extremely well for almost 50 years.

âjms

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

* Appeal to novelty: Being new, in and of itself, doesn't make
something better (or worse) than what came before.
https://en.wikipedia.org/wiki/Appeal_to_novelty <https://en.wikipedia.org/wiki/Appeal_to_novelty>
* Ad hominem: Painting as superstitious change-haters those who don't
like systemd doesn't in any way prove systemd is good.
https://en.wikipedia.org/wiki/Ad_hominem <https://en.wikipedia.org/wiki/Ad_hominem>
* False choice: Implying that the only alternative to systemd is
"init" (I think he means sysvinit) is wrong. I know of at least five
additional init systems that are excellent, and use one of them every
day (runit). https://en.wikipedia.org/wiki/False_dilemma <https://en.wikipedia.org/wiki/False_dilemma>
I'll give you this: The caveman analogy is funny, it's clever, and it
greatly appeals to those already on your side. But it actually says
nothing.
The real crackup is that these same falacies pop up in almost every
defense of systemd, starting with LP himself. Read PID EINS, you'll see
what I mean.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr <http://www.troubleshooters.com/mgr>
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale <http://mail.ale.org/mailman/listinfo/ale>
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo <http://mail.ale.org/mailman/listinfo>

Jim Kinney

2017-09-08 11:58:13 UTC

Permalink

UNIX is dead. Linux isn't UNIX. It's only UNIX-like.

Emacs violates those rules. An editor that can read your email out loud is rather crossing domains.

Apache violates those rules. The proxy capabilities are beyond it's initial web server domain.

But at least systemd provides a topic other than vi vs emacs to squabble over. That was getting old.

Besides, systemd follows most of the rules that were listed about as well as any other PID 1 process responsible for a current system startup.

Post by Jerald Sheets
I avoid all the BS and just hate it for its design, layout, and
intrusion into all sorts of things it shouldnât be fiddling around in,
breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does it very
well, has clearly defined input and output and doesnât try to handle
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing and do
it well. Write programs to work together. Write programs to handle text
streams, because that is a universal interface.â â Doug McIlroy, the
inventor of UNIX pipes
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from engines.
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration that nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and robust.
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say nothing.
Repair what you can â but when you must fail, fail noisily and as soon as possible.
Programmer time is expensive; conserve it in preference to machine time.
Avoid hand-hacking; write programs to write programs when you can.
Prototype before polishing, Get it working before you optimize it.
Distrust all claims for one true way.
Design for the future, because it will be here sooner than you think.
Iâll stick with what has worked extremely well for almost 50 years.
âjms

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

* Appeal to novelty: Being new, in and of itself, doesn't make
something better (or worse) than what came before.
https://en.wikipedia.org/wiki/Appeal_to_novelty

<https://en.wikipedia.org/wiki/Appeal_to_novelty>

Post by Steve Litt
* Ad hominem: Painting as superstitious change-haters those who don't
like systemd doesn't in any way prove systemd is good.
https://en.wikipedia.org/wiki/Ad_hominem

<https://en.wikipedia.org/wiki/Ad_hominem>

Post by Steve Litt
* False choice: Implying that the only alternative to systemd is
"init" (I think he means sysvinit) is wrong. I know of at least five
additional init systems that are excellent, and use one of them

every

Post by Steve Litt
day (runit). https://en.wikipedia.org/wiki/False_dilemma

<https://en.wikipedia.org/wiki/False_dilemma>

Post by Steve Litt
I'll give you this: The caveman analogy is funny, it's clever, and it
greatly appeals to those already on your side. But it actually says
nothing.
The real crackup is that these same falacies pop up in almost every
defense of systemd, starting with LP himself. Read PID EINS, you'll

see

Post by Steve Litt
what I mean.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr

<http://www.troubleshooters.com/mgr>

Post by Steve Litt
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale

<http://mail.ale.org/mailman/listinfo/ale>

Post by Steve Litt
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

<http://mail.ale.org/mailman/listinfo>

--
Sent from my Android device with K-9 Mail. All tyopes are thumb related and reflect authenticity.

DJ-Pfulio

2017-09-08 13:22:13 UTC

Permalink

Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out loud
is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond it's
initial web server domain.
But at least systemd provides a topic other than vi vs emacs to squabble
over. That was getting old.
Besides, systemd follows most of the rules that were listed about as
well as any other PID 1 process responsible for a current system startup.
I avoid all the BS and just hate it for its design, layout, and
intrusion into all sorts of things it shouldn’t be fiddling around
in, breaking a myriad of tenets of the “UNIX way”.
The “UNIX Way” is to have a tool that does one thing, does it very
well, has clearly defined input and output and doesn’t try to handle
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing and
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.” — Doug
McIlroy, the inventor of UNIX pipes
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from engines.
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration that
nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and robust.
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say nothing.
Repair what you can — but when you must fail, fail noisily and as
soon as possible.
Programmer time is expensive; conserve it in preference to machine time.
Avoid hand-hacking; write programs to write programs when you can.
Prototype before polishing, Get it working before you optimize it.
Distrust all claims for one true way.
Design for the future, because it will be here sooner than you think.
I’ll stick with what has worked extremely well for almost 50 years.
—jms

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

Chuck Payne

2017-09-08 13:46:22 UTC

Permalink

Ok, I have to defend Unix, it not totally dead. Remember BSD was giving
away, we have openBSD, FreeBSD, and a few more.

I love FreeBSD, I love how a lot of security apps are built on. Like
pfSense, great firewall.

I think Vi vs Emacs fight is old. Emacs should just be re-label as a CLI
shell. I am surprised just has it come up like X when they start.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out loud
is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond it's
initial web server domain.
But at least systemd provides a topic other than vi vs emacs to squabble
over. That was getting old.
Besides, systemd follows most of the rules that were listed about as
well as any other PID 1 process responsible for a current system startup.
I avoid all the BS and just hate it for its design, layout, and
intrusion into all sorts of things it shouldnât be fiddling around
in, breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does it very
well, has clearly defined input and output and doesnât try to handle
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing and
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.â â Doug
McIlroy, the inventor of UNIX pipes
This is why grep doesnât awk and vice-versa. In case weâve
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from engines.
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration that
nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and robust.
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say nothing.
Repair what you can â but when you must fail, fail noisily and as
soon as possible.
Programmer time is expensive; conserve it in preference to machine

time.

Post by Jim Kinney
Avoid hand-hacking; write programs to write programs when you can.
Prototype before polishing, Get it working before you optimize it.
Distrust all claims for one true way.
Design for the future, because it will be here sooner than you think.
Iâll stick with what has worked extremely well for almost 50 years.
âjms

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Adam Jimerson

2017-09-08 14:01:42 UTC

Permalink

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Sorry but that is a pretty week statement unless you use something that can
only edit text/hex/whatever your working with files and/or only using a web
server only capable of service static files (html, css, js, misc files,
etc)/a dedicated proxy server/a dedicated caching server.

Nginx even violates the "Do one thing and do it well" as it is a web server
that works well as forward/reverse proxy and it has really good caching
support. Same for some of the smaller more obscure web servers like lighttpd
and Caddy to name a few.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out loud
is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond it's
initial web server domain.
But at least systemd provides a topic other than vi vs emacs to squabble
over. That was getting old.
Besides, systemd follows most of the rules that were listed about as
well as any other PID 1 process responsible for a current system startup.
I avoid all the BS and just hate it for its design, layout, and
intrusion into all sorts of things it shouldnât be fiddling around
in, breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does it very
well, has clearly defined input and output and doesnât try to handle
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing and
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.â â Doug
McIlroy, the inventor of UNIX pipes
This is why grep doesnât awk and vice-versa. In case weâve
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from engines.
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration that
nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and robust.
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say nothing.
Repair what you can â but when you must fail, fail noisily and as
soon as possible.
Programmer time is expensive; conserve it in preference to machine

time.

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Adam Jimerson

2017-09-08 14:09:49 UTC

Permalink

As for the BSDs they are great (personal bias) but I would consider them
more of a UNIX clone/successor than anything at this point. Also even the
BSDs are not safe from breaking the "Do one thing..." philosophy there
have been multiple attempts to port/add something SystemD/LaunchD like to
at least FreeBSD, don't remember about any of the other BSDs.

Post by Adam Jimerson

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Sorry but that is a pretty week statement unless you use something that
can only edit text/hex/whatever your working with files and/or only using a
web server only capable of service static files (html, css, js, misc files,
etc)/a dedicated proxy server/a dedicated caching server.
Nginx even violates the "Do one thing and do it well" as it is a web
server that works well as forward/reverse proxy and it has really good
caching support. Same for some of the smaller more obscure web servers
like lighttpd and Caddy to name a few.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

startup.

Post by Jim Kinney
I avoid all the BS and just hate it for its design, layout, and
intrusion into all sorts of things it shouldnât be fiddling around
in, breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does it very
well, has clearly defined input and output and doesnât try to handle
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing and
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.â â Doug
McIlroy, the inventor of UNIX pipes
This is why grep doesnât awk and vice-versa. In case weâve
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from engines.
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration that
nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and robust.
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say nothing.
Repair what you can â but when you must fail, fail noisily and as
soon as possible.
Programmer time is expensive; conserve it in preference to machine

time.

think.

Post by Jim Kinney
Iâll stick with what has worked extremely well for almost 50 years.
âjms

Post by Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

Post by Jim Kinney

Post by Steve Litt

Post by Lightner, Jeffrey
Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jim Kinney

2017-09-08 14:23:50 UTC

Permalink

Summary: Those rules were guidelines at best that were promptly
violated by nearly everything ever written of any size or complexity
and always with a decent reason at the time.
Styles change. Wide ties. Skinny ties. Bow ties. No ties. Some people
see systemd as bondage ropes. Some see it as a tool that automates
stuff so they don't have to. Who's right? Depends on where the rope
burns came from.
<snark> Gnome _had_ to use systemd so some of it's backend crap that is
long-time flakey will get restarted when it dies suddenly </snark>
(said the long-time gnome user :-)
Hmm. I've never used any of the BSDs. The initials were always
associated with Blue Screen of Death for me.

Post by Adam Jimerson
As for the BSDs they are great (personal bias) but I would consider
them more of a UNIX clone/successor than anything at this point. Also
even the BSDs are not safe from breaking the "Do one thing..."
philosophy there have been multiple attempts to port/add something
SystemD/LaunchD like to at least FreeBSD, don't remember about any of
the other BSDs.

Post by DJ-Pfulio

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or

Apache.
Sorry but that is a pretty week statement unless you use something
that can only edit text/hex/whatever your working with files and/or
only using a web server only capable of service static files (html,
css, js, misc files, etc)/a dedicated proxy server/a dedicated
caching server.
Nginx even violates the "Do one thing and do it well" as it is a
web server that works well as forward/reverse proxy and it has
really good caching support. Same for some of the smaller more
obscure web servers like lighttpd and Caddy to name a few.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.

Preston

2017-09-08 14:34:45 UTC

Permalink

On 9/8/2017 9:23 AM, Jim Kinney wrote:
<snipped>

Post by Jim Kinney
Hmm. I've never used any of the BSDs. The initials were always
associated with Blue Screen of Death for me.

I have the same hangup about Fedora using "dnf" for package management.

I associate DNF with "Did Not Finish"

Preston

--
Be who you are and say what you feel, because those who mind don’t
matter and those who matter don’t mind.
-Dr. Seuss

Kyle Brieden

2017-09-08 15:27:53 UTC

Permalink

Oh.... was that not the joke? :P

---
Very respectfully,
Kyle Brieden

Post by Preston
<snipped>

Post by Jim Kinney
Hmm. I've never used any of the BSDs. The initials were always
associated with Blue Screen of Death for me.

I have the same hangup about Fedora using "dnf" for package management.
I associate DNF with "Did Not Finish"
Preston

Kyle Brieden

2017-09-08 15:34:16 UTC

Permalink

I would argue that Nginx is MUCH closer to the "Do one thing" philosophy
than apache is. Nginx essentially only serves up static content and
caches some of that static content in memory to serve it faster. I
think that falls well within the "do one thing" of serving static
content. Nginx doesn't do any server side processing at all; you need
php-fpm for php, gunicorn/uwsgi for python, etc... Nginx passes the
request back to those, those do the processing and hand rendered pages
to nginx, and as far as nginx is concerned, that's a static asset that
it fires out the front. All HTML/CSS/JS assets that it serves are
static content, rendered in browser. One could argue that nginx isn't a
"webserver" at all, although that's what it made it's name on. It
basically only does proxy to other things that do processing and then
serves up static objects.

On that same token, the fact that it can loadbalance and have many
backends to proxy to falls well within the purview of "being good at
proxying".

Just my 2 cents.

---
Very respectfully,
Kyle Brieden

Post by Adam Jimerson

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Sorry but that is a pretty week statement unless you use something
that can only edit text/hex/whatever your working with files and/or
only using a web server only capable of service static files (html,
css, js, misc files, etc)/a dedicated proxy server/a dedicated caching
server.
Nginx even violates the "Do one thing and do it well" as it is a web
server that works well as forward/reverse proxy and it has really good
caching support. Same for some of the smaller more obscure web
servers like lighttpd and Caddy to name a few.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out

loud

Post by Jim Kinney
is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond

it's

Post by Jim Kinney
initial web server domain.
But at least systemd provides a topic other than vi vs emacs to

squabble

Post by Jim Kinney
over. That was getting old.
Besides, systemd follows most of the rules that were listed about

Post by Jim Kinney
well as any other PID 1 process responsible for a current system

startup.

Post by Jim Kinney
On September 8, 2017 7:28:45 AM EDT, Jerald Sheets
I avoid all the BS and just hate it for its design, layout,

and

Post by Jim Kinney
intrusion into all sorts of things it shouldnât be fiddling

around

Post by Jim Kinney
in, breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does

it very

Post by Jim Kinney
well, has clearly defined input and output and doesnât try

to handle

Post by Jim Kinney
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing

and

Post by Jim Kinney
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.â

â Doug

Post by Jim Kinney
McIlroy, the inventor of UNIX pipes
This is why grep doesnât awk and vice-versa. In case
Write simple parts connected by clean interfaces.
Clarity is better than cleverness.
Design programs to be connected with other programs.
Separate policy from mechanism; separate interfaces from

engines.

Post by Jim Kinney
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration

that

Post by Jim Kinney
nothing else will do.
Design for visibility to make inspection and debugging easier.
Robustness is the child of transparency and simplicity.
Fold knowledge into data, so program logic can be stupid and

robust.

Post by Jim Kinney
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say

nothing.

Post by Jim Kinney
Repair what you can â but when you must fail, fail noisily

and as

Post by Jim Kinney
soon as possible.
Programmer time is expensive; conserve it in preference to

machine time.

Post by Jim Kinney
Avoid hand-hacking; write programs to write programs when you

can.

Post by Jim Kinney
Prototype before polishing, Get it working before you optimize

it.

Post by Jim Kinney
Distrust all claims for one true way.
Design for the future, because it will be here sooner than you

think.

Post by Jim Kinney
Iâll stick with what has worked extremely well for almost 50

years.

Post by Jim Kinney
âjms

On Sep 8, 2017, at 3:10 AM, Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

"reason" of

Post by Jim Kinney

Post by Lightner, Jeffrey
Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jerald Sheets

2017-09-08 15:47:24 UTC

Permalink

I think that at some level, weâre dabbling in philosophy here.

For instance, one could argue that without any plugins, additions, and most simply configured that Apache is similarly âbare bonesâ. (http_core.c and core.c only) certainly falls within this purview.

Apache at its simplest satisfies the description.
Nginx at its simplest satisfies the description.

Itâs when vendors begin âfront-loadingâ the applications with all sorts of modules and add-ons things start to get weird.

âjms

I would argue that Nginx is MUCH closer to the "Do one thing" philosophy than apache is. Nginx essentially only serves up static content and caches some of that static content in memory to serve it faster. I think that falls well within the "do one thing" of serving static content. Nginx doesn't do any server side processing at all; you need php-fpm for php, gunicorn/uwsgi for python, etc... Nginx passes the request back to those, those do the processing and hand rendered pages to nginx, and as far as nginx is concerned, that's a static asset that it fires out the front. All HTML/CSS/JS assets that it serves are static content, rendered in browser. One could argue that nginx isn't a "webserver" at all, although that's what it made it's name on. It basically only does proxy to other things that do processing and then serves up static objects.

Adam Jimerson

2017-09-08 16:04:10 UTC

Permalink

I would say for Nginx to truly fall under the "do one thing and do it well"
It would have to lose its proxy support and caching support. This means
instead of having Ngnix talk to php-fpm/gunicorn/uwsgi/passenger/etc and it
happen to cache things either to disk or RAM to speed things up you would
need to have something more like, HAProxy that talks to your
php-fpm/gunicorn/uwsgi/passenger/etc
and maybe Nginx for any static assets and if you wanted caching you would
need to throw in something like Varnish/Memcache. That way each moving
wheel is truly only responsible for doing one thing. That is a very
contrived example I'll admit and may actually be dumb to try and deploy
which is why Nginx does all the things that it out of the box.

This is just my view on it and was just meant as an example, not to say
that any of this is right or not as at the end of the day we are talking
about a philosophy. I'm not against Nginx or Apache for that matter, both
tools get the job done although I do prefer Nginx over Apache as I find it
nicer to work with.

Post by Kyle Brieden
I would argue that Nginx is MUCH closer to the "Do one thing" philosophy
than apache is. Nginx essentially only serves up static content and
caches some of that static content in memory to serve it faster. I
think that falls well within the "do one thing" of serving static
content. Nginx doesn't do any server side processing at all; you need
php-fpm for php, gunicorn/uwsgi for python, etc... Nginx passes the
request back to those, those do the processing and hand rendered pages
to nginx, and as far as nginx is concerned, that's a static asset that
it fires out the front. All HTML/CSS/JS assets that it serves are
static content, rendered in browser. One could argue that nginx isn't a
"webserver" at all, although that's what it made it's name on. It
basically only does proxy to other things that do processing and then
serves up static objects.
On that same token, the fact that it can loadbalance and have many
backends to proxy to falls well within the purview of "being good at
proxying".
Just my 2 cents.
---
Very respectfully,
Kyle Brieden

Post by Adam Jimerson

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Sorry but that is a pretty week statement unless you use something
that can only edit text/hex/whatever your working with files and/or
only using a web server only capable of service static files (html,
css, js, misc files, etc)/a dedicated proxy server/a dedicated caching
server.
Nginx even violates the "Do one thing and do it well" as it is a web
server that works well as forward/reverse proxy and it has really good
caching support. Same for some of the smaller more obscure web
servers like lighttpd and Caddy to name a few.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out

loud

Post by Jim Kinney
is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond

it's

Post by Jim Kinney
initial web server domain.
But at least systemd provides a topic other than vi vs emacs to

squabble

Post by Jim Kinney
over. That was getting old.
Besides, systemd follows most of the rules that were listed about

Post by Jim Kinney
well as any other PID 1 process responsible for a current system

startup.

Post by Jim Kinney
On September 8, 2017 7:28:45 AM EDT, Jerald Sheets
I avoid all the BS and just hate it for its design, layout,

and

Post by Jim Kinney
intrusion into all sorts of things it shouldnât be fiddling

around

Post by Jim Kinney
in, breaking a myriad of tenets of the âUNIX wayâ.
The âUNIX Wayâ is to have a tool that does one thing, does

it very

Post by Jim Kinney
well, has clearly defined input and output and doesnât try

to handle

Post by Jim Kinney
multiple responsibility domains.
"This is the Unix philosophy: Write programs that do one thing

and

Post by Jim Kinney
do it well. Write programs to work together. Write programs to
handle text streams, because that is a universal interface.â

â Doug

engines.

Post by Jim Kinney
Design for simplicity; add complexity only where you must.
Write a big program only when it is clear by demonstration

that

robust.

Post by Jim Kinney
In interface design, always do the least surprising thing.
When a program has nothing surprising to say, it should say

nothing.

Post by Jim Kinney
Repair what you can â but when you must fail, fail noisily

and as

Post by Jim Kinney
soon as possible.
Programmer time is expensive; conserve it in preference to

machine time.

Post by Jim Kinney
Avoid hand-hacking; write programs to write programs when you

can.

Post by Jim Kinney
Prototype before polishing, Get it working before you optimize

it.

Post by Jim Kinney
Distrust all claims for one true way.
Design for the future, because it will be here sooner than you

think.

Post by Jim Kinney
Iâll stick with what has worked extremely well for almost 50

years.

Post by Jim Kinney
âjms

On Sep 8, 2017, at 3:10 AM, Steve Litt
On Thu, 7 Sep 2017 12:29:46 +0000

"reason" of

Post by Jim Kinney

Post by Lightner, Jeffrey
Ug's opinion.
Ug: What that?
Zog: Systemd.
Ug: Why systemd. Init work for years...

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jim Kinney

2017-09-08 14:12:38 UTC

Permalink

Ha! I use vim(vi) because it's (almost) always installed and available
from runlevel 1/ system_repair. Emacs almost never was.
My son is writing an application in node.js. It is it's own web server.

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Post by Jim Kinney
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.

_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

James Sumners

2017-09-08 14:23:34 UTC

Permalink

Node is not "its own web server." It has a standard library that
facilitates writing HTTP application servers.

Ha! I use vim(vi) because it's (almost) always installed and available
from runlevel 1/ system_repair. Emacs almost never was.
My son is writing an application in node.js. It is it's own web server.
Thanks for the reminder for why I don't use either Emacs or Apache.
UNIX is dead. Linux isn't UNIX. It's only UNIX-like.
_______________________________________________
Ale mailing list
See JOBS, ANNOUNCE and SCHOOLS lists athttp://mail.ale.org/mailman/listinfo
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

--
James Sumners
http://james.sumners.info/ (technical profile)
http://jrfom.com/ (personal site)
http://haplo.bandcamp.com/ (music)

Solomon Peachy

2017-09-09 04:02:47 UTC

Permalink

Post by DJ-Pfulio
Thanks for the reminder for why I don't use either Emacs or Apache.

Yet you wrote this email using Thunderbird...

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Delray Beach, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Steve Litt

2017-09-09 02:21:10 UTC

Permalink

On Fri, 08 Sep 2017 07:58:13 -0400

Post by Jim Kinney
UNIX is dead.

Not unless somehow you can say BSD isn't Linux.

Post by Jim Kinney
Linux isn't UNIX. It's only UNIX-like.
Emacs violates those rules. An editor that can read your email out
loud is rather crossing domains.
Apache violates those rules. The proxy capabilities are beyond it's
initial web server domain.

This is a new fallacy I haven't seen used by systemd defenders before:
Argumentum ad populum. Emacs does it, Apache does it, so it must be the
right way. https://en.wikipedia.org/wiki/Argumentum_ad_populum

Post by Jim Kinney
But at least systemd provides a topic other than vi vs emacs to
squabble over. That was getting old.

Systemd vs sane inits is nothing like vi vs emacs. You can switch
between them any time you want (unless you're using very esoteric
capabilities or addons). But systemd manages to get its tenticles into
every last bit of the software ecosphere, making replacing it a
gruesome task of finding and pulling out all those tenticles, before
finally rebuilding the whole system in a sane way.

Post by Jim Kinney
Besides, systemd follows most of the rules that were listed about as
well as any other PID 1 process responsible for a current system
startup.

Where'd you pull that statement out of?

If your statement is true then why is it so darn hard to replace
systemd's (kidnapped) udev with another device whateverer, if systemd's
software is simple, with simple parts and clean interfaces? It should
be trivial to bold on a similarly working device whateverer.

Try initting with s6, runit or Epoch and come back to me and tell me
systemd follows most of the rules as well as other init systems.

Here's how simple runit is: I built my own imitation of runit from the
83 line of C Suckless Init program, the daemontools-encore supervisor
suite, and a few shellscripts I wrote (called LittKit) that impose
startup order.

As far as daemontools-encore, it's very simple: Loop
through the /service (or /var/service or whatever) directory looking
for symlinks to other directories, and if they're not running and
should be running, run the script called "run" in those directories,
and that script sets up environment etc and then just execs the daemon.
State is kept in that directory's ./supervise subdirectory.

Runit, (and my Suckless Init/daemontools-encore/LittKit combo) conform
to the rules. Systemd: Not so much.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Solomon Peachy

2017-09-09 04:01:42 UTC

Permalink

Post by Steve Litt
Here's how simple runit is: I built my own imitation of runit from the
83 line of C Suckless Init program, the daemontools-encore supervisor
suite, and a few shellscripts I wrote (called LittKit) that impose
startup order.

Congragulations; Now that you've implemented init, the next assignment
in CS3411 is to write a coalescing malloc implementation.

Bonus points will be awarded for showing why an implementation
consistent with The UNIX Way (tm) doesn't exist outside a textbook. [1]

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Delray Beach, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Steve Litt

2017-09-09 22:36:22 UTC

Permalink

On Sat, 9 Sep 2017 00:01:42 -0400

Post by Solomon Peachy

Post by Steve Litt
Here's how simple runit is: I built my own imitation of runit from
the 83 line of C Suckless Init program, the daemontools-encore
supervisor suite, and a few shellscripts I wrote (called LittKit)
that impose startup order.

Congragulations; Now that you've implemented init, the next
assignment in CS3411 is to write a coalescing malloc implementation.

Fascinating: A strawman fallacy
(https://en.wikipedia.org/wiki/Straw_man) with an ad-hominem flavor.
Strawman because nobody was discussing "coalescing malloc
implementations" nor any kind of CS course. The discussion (so neatly
snipped from the quote) was about how simple runit is, so simple in
fact that I basically bolted together a similar init out of Suckless
Init, daemontools-encore, and a few home-grown shellscripts.

Post by Solomon Peachy
Bonus points will be awarded for showing why an implementation
consistent with The UNIX Way (tm) doesn't exist outside a textbook.
[1]
- Solomon

LOL, you have the [1] for reference, but there is no reference. Anyway,
the fallacy involved in your preceding sentence is the Alternative
Facts Fallacy (https://en.wikipedia.org/wiki/Alternative_facts). Runit's
consistent with The UNIX Way, as is s6, as was my homegrown init. So
are most shellscripts written by intelligent and efficient admins. From
what I've seen from using it, Dovecot is a great example of The UNIX
Way, as is fetchmail, procmail, and tools like ls, cat, sed, awk, cut,
grep, sort, make, and the like.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Solomon Peachy

2017-09-09 23:31:37 UTC

Permalink

From what I've seen from using it, Dovecot is a great example of The
UNIX Way

*snerk* IMAP is about as far from The UNIX way as it gets, which doesn't
bode well for any implementations of it, including Dovecot, which does a
lot more than serve up IMAP (and its myriad extensions) which is just
one of its many horrible transgressions aganist the UNIX way.

as is fetchmail,

Fetchmail is a flaming train wreck from start to finish. It's an even
worse example of the UNIX way than Dovecot is, but that doesn't make it
any less useful.

procmail

That violates the "do one thing" principle, but I'll concede it does
what it does admirably well.

ls, cat, cut, sort

Single-purpose commands

sed, awk, grep

Highly useful, but there is a very strong case to be made for combining
them due to their overlap. (Resulting in Perl)

make, and the like.

Make is awesome, but it's scope is so limited that it's rarely used on
its own. It will outlive all of its would-be replacements.

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Delray Beach, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Solomon Peachy

2017-09-10 00:56:23 UTC

Permalink

Post by Steve Litt
Strawman because nobody was discussing "coalescing malloc
implementations" nor any kind of CS course. The discussion (so neatly
snipped from the quote) was about how simple runit is, so simple in
fact that I basically bolted together a similar init out of Suckless
Init, daemontools-encore, and a few home-grown shellscripts.

Since you're also incapable of understanding subtelties, let me be more
direct. Writing an init and a malloc implementation were just two of the
many exercises expected of all students enrolled in CS3411 in the late
'90s at Georgia Tech.

So, again, congratulations on your college-junior-level accomplishment,
but your appeal to self-authority doesn't really stand for much when
what you've described is documented in the 20-year old textbook
("Operating System Concepts" by Galvin et al) used by that class, to say
nothing about the excellent texts about UNIX internals by W. Richard
Stephens and literally decades worth of UNIX and BSD source code that's
out there for anyone to peruse should they feel inclined.

So, why is this point relevant? It's simple -- you seem to be
completely incapable of grasping the essential premise for systemd;
Namely that traditional init is _too_ simple for the demands of modern
systems. How, pray tell, does presenting an even _simpler_ option
provide a counterpoint to that argument?

Here's a hint: It doesn't, and in doing so, you committed a strawman
fallacy of your very own in the process.

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Delray Beach, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Steve Litt

2017-09-10 09:53:47 UTC

Permalink

On Sat, 9 Sep 2017 20:56:23 -0400

Post by Solomon Peachy
So, why is this point relevant? It's simple -- you seem to be
completely incapable of grasping the essential premise for systemd;
Namely that traditional init is _too_ simple for the demands of
modern systems.

Says you. The preceding sentence is your opinion. That does not make it a fact.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Solomon Peachy

2017-09-20 02:07:48 UTC

Permalink

Post by Steve Litt

Says you. The preceding sentence is your opinion. That does not make it a fact.

Not I -- The systemd authors. Or are you seriously asserting that the
folks who wrote systemd are actively lying about why they wrote (and
use, because they eat their own dogfood) it?

You're entitled to your own opinions, but not your own facts -- And as
you just aptly pointed out, those words are not synonmous.

Cheers,

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Coconut Creek, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Solomon Peachy

2017-09-09 04:17:53 UTC

Permalink

Post by Jerald Sheets
Clarity is better than cleverness.

This is actually a gross understatement -- UNIX prioritized simplicity
of implementation over all other considerations, including simplicity of
interface, consistency, perforamnce, and even correctness (ie handling
all corner cases properly)

In the real world (and, I might add, that includes actual UNIXes)
"cleverness" trumps "clarity" most of the time because all of those
other considerations matter, including and especially performance.

Meanwhile, Reading down your list, I can't help but marvel that systemd
manages to hit the overwhelming majority of them. Not all, mnind you,
but that's not terribly surprising as many of those properties
contracict with some of the others. It is impossible to satisfy them
all.

Post by Jerald Sheets
Distrust all claims for one true way.

...Judge not, lest ye be judged.

Post by Jerald Sheets
Iâll stick with what has worked extremely well for almost 50 years.

By that same token we should still have pay phones on every street
corner. They worked great for much longer, after all.

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Delray Beach, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Steve Litt

2017-09-09 22:41:40 UTC

Permalink

On Sat, 9 Sep 2017 00:17:53 -0400

Post by Solomon Peachy

Post by Jim Kinney
I’ll stick with what has worked extremely well for almost 50
years.

By that same token we should still have pay phones on every street
corner. They worked great for much longer, after all.
- Solomon

You systemd guys just can't help using the Appeal to Novelty Fallacy,
can you? (https://en.wikipedia.org/wiki/Appeal_to_novelty)

Just because newer phones are better (in most regards) than older
phones doesn't mean that newer inits are better than older inits.

And of course, systemd is about the same age as superior inits like
Runit and s6.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr

Solomon Peachy

2017-09-10 00:25:49 UTC

Permalink

Post by Steve Litt
You systemd guys just can't help using the Appeal to Novelty Fallacy,
can you? (https://en.wikipedia.org/wiki/Appeal_to_novelty)

Who said I'm a systemd guy? And since you're such a fan of logical
fallacies, I shouldn't have to point out that you're committing appeals
to tradition, with an unhealthy dose of appealing to authority
sandwiching moralistic fallacies.

Post by Steve Litt
Just because newer phones are better (in most regards) than older
phones doesn't mean that newer inits are better than older inits.

But that also doesn't mean that newer inits _aren't_ better than older
inits -- Congratulations, you just committed a another logical fallacy.
Now its precise name escapes me, but I'm sure that you and your superior
grasp of logical fallacies can educate us.

Post by Steve Litt
And of course, systemd is about the same age as superior inits like
Runit and s6.

Now who's appealing to novelty? Meanwhile, "superior" is a subjective
term, and if you're being remotely intellectually honest, you will
concede that others might use different evaluation criteria and as such,
might come to different conclusions. Otherwise, you're just begging
the question and/or commiting circular reasoning.

If you want to live by the fallacy, be prepared to die by the fallacy.

Meanwhile, I don't have anything better to do than keep this thread
going until Irma rips the roof off my building. C'est la vie.

- Solomon

--
Solomon Peachy pizza at shaftnet dot org
Coconut Creek, FL ^^ (email/xmpp) ^^
Quidquid latine dictum sit, altum videtur.

Steve Litt

2017-09-10 04:41:32 UTC

Permalink

On Sat, 9 Sep 2017 20:25:49 -0400

Post by Solomon Peachy

Post by Steve Litt
You systemd guys just can't help using the Appeal to Novelty
Fallacy, can you?
(https://en.wikipedia.org/wiki/Appeal_to_novelty)

Who said I'm a systemd guy?

Your past performance on this mailing list.

Post by Solomon Peachy
And since you're such a fan of logical
fallacies, I shouldn't have to point out that you're committing
appeals to tradition,

No. I didn't say runit and s6 are better because they're older. In many
other postings on this list I gave my reasons why they were better (or
actually, why systemd is worse).

Post by Solomon Peachy
with an unhealthy dose of appealing to
authority

What authority? I could bring my Rich Felker, and you could bring your
Lennart Poettering. I don't think either of us did that so far.

sandwiching moralistic fallacies.

I had to look that up at
https://en.wikipedia.org/wiki/Moralistic_fallacy , and I don't recall,
**in this thread**, doing anything resembling that.

Post by Solomon Peachy

Post by Steve Litt
Just because newer phones are better (in most regards) than older
phones doesn't mean that newer inits are better than older inits.

But that also doesn't mean that newer inits _aren't_ better than
older inits

Pre-cisely! Age of something doesn't indicate its betterness or
worseness, or even betterness or worseness for a given usage.

Post by Solomon Peachy
-- Congratulations, you just committed a another logical
fallacy. Now its precise name escapes me, but I'm sure that you and
your superior grasp of logical fallacies can educate us.

OK, let me rephrase to get rid of the slightest hint of any fallacy,
although logically this sentence is the same as my original:

Just because newer phones are better (in most regards) than older
phones doesn't *NECESSARILY* mean that newer inits are better than older
inits.

Let's leave age out of the discussion: It contributes nothing.

Post by Solomon Peachy

Post by Steve Litt
And of course, systemd is about the same age as superior inits like
Runit and s6.

Now who's appealing to novelty? Meanwhile, "superior" is a
subjective term,

You're right, it's subjective. Let me substitute the following text for
"superior":

=========================================================
An init system that's 1 to 3 orders of magnitude more complex, meaning
more parts, more interactions (especially circular ones), unless one of
systemd's special features is essential and there's no other way to
provide it.
=========================================================

Three or four block diagrams with less than 10 boxes apiece, all fitted
with interaction lines, describe the architecture of runit. S6 would
require a similar number and style of block diagrams. But systemd is so
complex that I've never seen a complete block diagram of it **complete
with interaction lines**.

Post by Solomon Peachy
and if you're being remotely intellectually honest,
you will concede that others might use different evaluation criteria
and as such, might come to different conclusions.

This is true. I believe that complexity comes with a significant cost.
Anyone not believing that would have no reason to eschew systemd. I
believe that the only time I'm willing to pay the price of
massive complexity is if I can't get along without what it gives me.
This includes the smog control in my car engine and the Linux kernel.
For those who really, truly need features like cgroups, per-seat
computing and can't use Linux Terminal Server Project, or need to start
daemons on an as-needed basis (superserver) and for some reason can't
use xinetd, the complexity cost of systemd might be acceptable. For the
vast majority who don't fit that description and do believe complexity
has a significant cost, systemd doesn't make sense.

One might say "well then just change out the init system." That is VERY
difficult to do on a systemd distro, because systemd gets its roots and
tenticles into so much software you can't chisel it out without busting
a lot of software that's hard to put back together. So realistically,
you have to go with a sans-systemd distro to use something like runit
or s6. Understand, it's pretty easy to swap in runit or s6 for
sysvinit. The *only* popular Linux init that makes itself irreplacable
is systemd.

Post by Solomon Peachy
Otherwise, you're
just begging the question and/or commiting circular reasoning.
If you want to live by the fallacy, be prepared to die by the fallacy.

Wait! It's not *me* who is living by the fallacy.

SteveT

Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Joey Kelly

2017-09-08 17:51:34 UTC

Permalink

Post by Steve Litt
Except for those who "simply must" have their Gnome, KDE or
NetworkManager, there are plenty of Linux systems out there they can
run without systemd.

Oh yes, I simply /must/ have NetworkMangler.

--Joey
_______________________________________________
Ale mailing list
***@ale.org
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo

Jeff Hubbs

2017-09-07 13:18:01 UTC

Permalink

-------- Forwarded Message --------
Subject: Re: [ale] please bow your head for a moment of silence...
Date: Thu, 7 Sep 2017 09:17:40 -0400
From: Jeff Hubbs <***@att.net>
To: Steve Litt <***@troubleshooters.com>

As of now systemd remains an optional path in the Gentoo Handbook (the
official document that guides you through a fresh Gentoo Linux
installation). My only attempt at it was several years ago when it was
still quite new and my experience was not good. Have been holding off
until the day comes when I think Gnome etc. is worth having on Gentoo.
The only *desktop* Gentoo machine I've set up in the last ten years or
so was as usable as it needed to be with xfce.

Post by Steve Litt
On Wed, 6 Sep 2017 07:53:36 -0400

Post by leam hall
Given the Linux adoption of systemd the only options seem to be
"Winderz with a GUI", "Winderz with a command line", "Expensive
Fruit", or some flavor of BSD.

That being said, several Linux distros don't use systemd and are
* Devuan
* Void
* Funtoo
There are probably seven or eight more that currently don't use
systemd, or don't use it by default, but have not foresworn the
possibility.
Except for those who "simply must" have their Gnome, KDE or
NetworkManager, there are plenty of Linux systems out there they can
run without systemd.
SteveT
Steve Litt
September 2017 featured book: Manager's Guide to Technical
Troubleshooting Brand new, second edition
http://www.troubleshooters.com/mgr
_______________________________________________
Ale mailing list
http://mail.ale.org/mailman/listinfo/ale
See JOBS, ANNOUNCE and SCHOOLS lists at
http://mail.ale.org/mailman/listinfo