So I noticed that the minimal installation of CentOS 7 comes - in contrary to minimal CentOS 6.5 - preinstalled with the NetworkManager.
I somewhat understand its usefulness, especially for wlan / desktops, and its not like it really bothers me.. That being said, it seems to me with my naive linux understanding that it kinda goes against the definition of a "minimal" installation.
Does anyone know why this decision was made / a good idea?
On 8/22/2014 2:55 PM, Christof Stocker wrote:
So I noticed that the minimal installation of CentOS 7 comes - in contrary to minimal CentOS 6.5 - preinstalled with the NetworkManager.
I somewhat understand its usefulness, especially for wlan / desktops, and its not like it really bothers me.. That being said, it seems to me with my naive linux understanding that it kinda goes against the definition of a "minimal" installation.
Does anyone know why this decision was made / a good idea?
network manager is used by default to manage all sorts of connections now, including ethernet with DHCP.
On Fri, Aug 22, 2014 at 02:58:29PM -0700, John R Pierce wrote:
On 8/22/2014 2:55 PM, Christof Stocker wrote:
So I noticed that the minimal installation of CentOS 7 comes - in contrary to minimal CentOS 6.5 - preinstalled with the NetworkManager.
I somewhat understand its usefulness, especially for wlan / desktops, and its not like it really bothers me.. That being said, it seems to me with my naive linux understanding that it kinda goes against the definition of a "minimal" installation.
Does anyone know why this decision was made / a good idea?
network manager is used by default to manage all sorts of connections now, including ethernet with DHCP.
More important with regards to the minimal install set it matches what Red Hat is doing.
John
John R. Dennison wrote:
On Fri, Aug 22, 2014 at 02:58:29PM -0700, John R Pierce wrote:
On 8/22/2014 2:55 PM, Christof Stocker wrote:
So I noticed that the minimal installation of CentOS 7 comes - in contrary to minimal CentOS 6.5 - preinstalled with the NetworkManager.
I somewhat understand its usefulness, especially for wlan / desktops, and its not like it really bothers me.. That being said, it seems to
me
with my naive linux understanding that it kinda goes against the definition of a "minimal" installation.
Does anyone know why this decision was made / a good idea?
network manager is used by default to manage all sorts of connections now, including ethernet with DHCP.
More important with regards to the minimal install set it matches what Red Hat is doing.
And most of us *still* don't like it....
mark
John-- If there is an embarrassment equivalent of post-traumatic stress disorder, South Carolina has it.
-- Dick Harpootlian, former state Democratic chairman, on its recent politics, New York Times, 12 June 2010 _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Fri, Aug 22, 2014 at 06:13:56PM -0400, m.roth@5-cent.us wrote:
And most of us *still* don't like it....
And luckily there is a solution. Don't use CentOS-7.
John
On 22/08/14 06:13 PM, m.roth@5-cent.us wrote:
John R. Dennison wrote:
On Fri, Aug 22, 2014 at 02:58:29PM -0700, John R Pierce wrote:
On 8/22/2014 2:55 PM, Christof Stocker wrote:
So I noticed that the minimal installation of CentOS 7 comes - in contrary to minimal CentOS 6.5 - preinstalled with the NetworkManager.
I somewhat understand its usefulness, especially for wlan / desktops, and its not like it really bothers me.. That being said, it seems to
me
with my naive linux understanding that it kinda goes against the definition of a "minimal" installation.
Does anyone know why this decision was made / a good idea?
network manager is used by default to manage all sorts of connections now, including ethernet with DHCP.
More important with regards to the minimal install set it matches what Red Hat is doing.
And most of us *still* don't like it....
mark
Time is ticking on... The longer you avoid learning what is coming, the further behind your peers you will fall.
On Fri, Aug 22, 2014 at 5:18 PM, Digimer lists@alteeve.ca wrote:
More important with regards to the minimal install set it matches what
Red Hat is doing.
And most of us *still* don't like it....
markTime is ticking on... The longer you avoid learning what is coming, the further behind your peers you will fall.
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
On Fri, Aug 22, 2014 at 05:24:06PM -0500, Les Mikesell wrote:
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
This is a _major_ release. Things change during _major_ releases. Luckily you aren't being forced at gunpoint to use it.
Seriously. Your constant complaints against the Red Hat way of doing things got old a decade ago.
John
On Fri, Aug 22, 2014 at 5:33 PM, John R. Dennison jrd@gerdesas.com wrote:
On Fri, Aug 22, 2014 at 05:24:06PM -0500, Les Mikesell wrote:
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
This is a _major_ release. Things change during _major_ releases. Luckily you aren't being forced at gunpoint to use it.
Seriously. Your constant complaints against the Red Hat way of doing things got old a decade ago.
It's not so much 'The' Red Hat way of doing things - although SysV mostly had it right in the first place. But the annoying part is the number of Red Hat "Ways' that are just arbitrarily different - like a car company swapping the brake and gas pedal locations for every new model. I suppose if you sell training courses you have to make a reason for people to come back.
On 8/22/2014 3:39 PM, Les Mikesell wrote:
It's not so much 'The' Red Hat way of doing things - although SysV mostly had it right in the first place. But the annoying part is the number of Red Hat "Ways' that are just arbitrarily different - like a car company swapping the brake and gas pedal locations for every new model. I suppose if you sell training courses you have to make a reason for people to come back.
in the early days of automobiles there was quite a lot of that.
ever drive a model T ?
On 22/08/14 06:39 PM, Les Mikesell wrote:
On Fri, Aug 22, 2014 at 5:33 PM, John R. Dennison jrd@gerdesas.com wrote:
On Fri, Aug 22, 2014 at 05:24:06PM -0500, Les Mikesell wrote:
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
This is a _major_ release. Things change during _major_ releases. Luckily you aren't being forced at gunpoint to use it.
Seriously. Your constant complaints against the Red Hat way of doing things got old a decade ago.
It's not so much 'The' Red Hat way of doing things - although SysV mostly had it right in the first place. But the annoying part is the number of Red Hat "Ways' that are just arbitrarily different - like a car company swapping the brake and gas pedal locations for every new model. I suppose if you sell training courses you have to make a reason for people to come back.
To continue your analogy, should car companies have stopped changing after the 20s? I mean, the cars then got you were you needed to go, right?
Things change. You are certainly free to deny that and stay on old releases, but the world *will* move forward, with or without you.
On Fri, Aug 22, 2014 at 5:46 PM, Digimer lists@alteeve.ca wrote:
To continue your analogy, should car companies have stopped changing after the 20s? I mean, the cars then got you were you needed to go, right?
The point is to abstract an interface so you can make changes behind it without breaking the things already built around it. You can always add things without breaking anything that already worked for your community of users. If you didn't care about that yourself, you'd be recompiling a gentoo weekly instead of being here.
Things change. You are certainly free to deny that and stay on old releases, but the world *will* move forward, with or without you.
Or around in circles...
On 22/08/14 07:07 PM, Les Mikesell wrote:
On Fri, Aug 22, 2014 at 5:46 PM, Digimer lists@alteeve.ca wrote:
To continue your analogy, should car companies have stopped changing after the 20s? I mean, the cars then got you were you needed to go, right?
The point is to abstract an interface so you can make changes behind it without breaking the things already built around it. You can always add things without breaking anything that already worked for your community of users. If you didn't care about that yourself, you'd be recompiling a gentoo weekly instead of being here.
To echo John, this is a major release. It's where, when needed, things can change and break backwards compatibility. If a change like this happened as a y-stream release, sure, I'll grab my pitch fork along with you.
It's not realistic to expect backwards compatibility to last forever. The sysv init stuff had a good long run, but it was time to change. Now, you're welcome to disagree with me (and the archives are littered already with this argument), but in the end, it changed. A major version was the right place to do it, and now it is done.
So this brings me back to my original point... Unless you plan to wage a war against things like Network Manager, systemd or what have you in the faint home of reverting in the next major release, you don't have a lot of viable long term options.
Learn the new ways or fade from relevance.
I say this without passing judgment on the merits of the new or old ways, simply as a fact of life. Even if you did hold out hope for, say, RHEL 8 to return to the old ways, you will have a hard time avoiding EL7. It will almost certainly be adopted wide-scale and that will provide inertia.
On 08/22/2014 07:42 PM, Digimer wrote:
On 22/08/14 07:07 PM, Les Mikesell wrote:
On Fri, Aug 22, 2014 at 5:46 PM, Digimer lists@alteeve.ca wrote:
To continue your analogy, should car companies have stopped changing after the 20s? I mean, the cars then got you were you needed to go, right?
The point is to abstract an interface so you can make changes behind it without breaking the things already built around it. You can always add things without breaking anything that already worked for your community of users. If you didn't care about that yourself, you'd be recompiling a gentoo weekly instead of being here.
To echo John, this is a major release. It's where, when needed, things can change and break backwards compatibility. If a change like this happened as a y-stream release, sure, I'll grab my pitch fork along with you.
It's not realistic to expect backwards compatibility to last forever. The sysv init stuff had a good long run, but it was time to change. Now, you're welcome to disagree with me (and the archives are littered already with this argument), but in the end, it changed. A major version was the right place to do it, and now it is done.
So this brings me back to my original point... Unless you plan to wage a war against things like Network Manager, systemd or what have you in the faint home of reverting in the next major release, you don't have a lot of viable long term options.
Learn the new ways or fade from relevance.
I say this without passing judgment on the merits of the new or old ways, simply as a fact of life. Even if you did hold out hope for, say, RHEL 8 to return to the old ways, you will have a hard time avoiding EL7. It will almost certainly be adopted wide-scale and that will provide inertia.
NetworkManager is the window's world way of doing things for people that don't really understand what is going on. I see no use for it immediately disable it. But it pains me to have to take the time.
On 08/23/2014 12:24 AM, Les Mikesell wrote:
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
I totally agree, Les, I think RH and CentOS are really going the wrong way since the release of that ugly version 7 !!!
On 2014-08-22, Bernard Lheureux bernard.lheureux@bbsoft4.org wrote:
I totally agree, Les, I think RH and CentOS are really going the wrong way since the release of that ugly version 7 !!!
You can always start a NoNetworkManager SIG.
--keith
On Fri, 22 Aug 2014, Les Mikesell wrote:
On Fri, Aug 22, 2014 at 5:18 PM, Digimer lists@alteeve.ca wrote:
More important with regards to the minimal install set it matches what
Red Hat is doing.
And most of us *still* don't like it....
markTime is ticking on... The longer you avoid learning what is coming, the further behind your peers you will fall.
Except that wasting time re-learning a new and strange way to do something that already worked - or how to disable the new thing so it doesn't break your working setup - doesn't really put you ahead of anything.
I hate network mangler as much as the next guy but is it really worth all of the whining when all it takes to disable it is:
systemctl disable NetworkManager systemctl enable network systemctl stop NetworkManager systemctl start network
And now you are back to the old behavior. Red Hat even went to the trouble of documenting it for you at https://access.redhat.com/documentation/en-US/Red_Hat_Enterprise_Linux/7/htm...
Regards,
On Sat, August 23, 2014 5:00 am, me@tdiehl.org wrote:
I hate network mangler as much as the next guy but is it really worth all of the whining when all it takes to disable it is:
It would be worth "whining about it" if anybody of decision makers ever listened to these complaints. As some day "reverting to old behavior" option will be gone. But most likely no one will listen to all our "whining", and all the decisions are already made at least a year ago... so you probably are 100% right: all our whining serves is just to let our own steam out. Once we realize it we start looking for alternatives, - for the servers at least.
Valeri
++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++
You are whining about something FREE…don’t like it, don’t use it….if you had a PAID RHEL sub, upstream to Cent, on then bitch…..but whining about something free, well
On Aug 23, 2014, at 8:38 AM, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On Sat, August 23, 2014 5:00 am, me@tdiehl.org wrote:
I hate network mangler as much as the next guy but is it really worth all of the whining when all it takes to disable it is:
It would be worth "whining about it" if anybody of decision makers ever listened to these complaints. As some day "reverting to old behavior" option will be gone. But most likely no one will listen to all our "whining", and all the decisions are already made at least a year ago... so you probably are 100% right: all our whining serves is just to let our own steam out. Once we realize it we start looking for alternatives, - for the servers at least.
Valeri
++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
On Sat, August 23, 2014 8:42 am, William Woods wrote:
You are whining about something FREE dont like it, dont use it .if you had a PAID RHEL sub, upstream to Cent, on then bitch ..but whining about something free, well
Was I that unclear that I sounded like the one who keeps whining? I tried to say that the moment we could affect anything has past a year or two ago. That was the time the systemd introduction into all Linuxes was made. It is done deal now, and the last one of the major distros - debian (and its clones) - goes systemd in next release. So, it is not RH, it is all of them built on Linux kernel...
And yes, I did start using something else (FreeBSD) for servers a while ago. Also free. Also open source. Better suited for servers in my book (your mileage may differ ;-)
Alas, not all of the decisions that are made in/by open source programmer (steering) teams can be affected by us. They are achieved in the battles, and there are arguments "on our side" that are made then. But. As I said to one of my users: KDE-3 person, who hates KDE-4, stays with KDE-3 while it lasts. Brilliant programmers who create this software need to make progress as _they_ see it. And this (making these fundamental for us changes) often is their only reward for the great programming job they are doing. Let's be grateful to them.
And as we know, not all of the changes is really a progress, even if they give you very fast boot as systemd does, or pretend to give you more security as SELinux advertizes in its name. I was displeased by introduction of SELinux into mainstream kernel back then. As, it is not a good defense in a first place (can it be if you can switch it off on the fly? and after that things are as if it is not there). On the other hand it is extra dozens of thousands of lines of code in the kernel, which may have bugs with security implications. Which down the road proved to be true - search for SELinux security patch. Still, even disagreeing with something I kept living with it for quire some time. But one day the time came to switch servers to better (in my book; your mileage may be different ;-) alternative. Oh, yes, I should have mentioned SELinux competitive security solution. it was LIDS (Linux Intrusion Detection System). The name is a bit confusing. In three words: It was sort of kernel patch that after boot demotes root to user nobody. So after boot you can not administer the system at all. On the fly the system is locked. Dead locked. Makes more sense to me (security wise) than SELinux, but SELinux made it into mainstream kernel instead of LIDS...
The suggestion you made to switch to commercial system [sorry I brought your suggestion one step further in the same direction, oh I'm really tricky person] is quite in line with what commercial vendors would like to happen to free (as free beer) competitive software: users, feel this free software is as nasty as our commercial alternative is. So you may look at better sides of commercial software, and come back to us. This may be strategic thought behind such events as acquisition of widest used database mysql by most famous database company oracle. Another example may be proving an opposite (I mean cups acquired by Apple, the reason here could be mere survival of cups that Apple is going to keep using themselves).
So, for good or for bad, after letting all of our steam out about bad decisions in the system we love or used to love (and I was happy with Linux, - RedHat and CentOS in particular, - for much longer than decade) we can bite the bullet, realize that the life is such, and Linux from now on is such, and start continuing our life with Linux (while the enterprise life cycle lasts ;-) or with alternatives, - those of us who found them more adequate.
One way or another whining of all of us who is displeased only serves to let our own steam out.
Valeri
On Aug 23, 2014, at 8:38 AM, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
On Sat, August 23, 2014 5:00 am, me@tdiehl.org wrote:
I hate network mangler as much as the next guy but is it really worth all of the whining when all it takes to disable it is:
It would be worth "whining about it" if anybody of decision makers ever listened to these complaints. As some day "reverting to old behavior" option will be gone. But most likely no one will listen to all our "whining", and all the decisions are already made at least a year ago... so you probably are 100% right: all our whining serves is just to let our own steam out. Once we realize it we start looking for alternatives, - for the servers at least.
Valeri
++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++ _______________________________________________ CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
CentOS mailing list CentOS@centos.org http://lists.centos.org/mailman/listinfo/centos
++++++++++++++++++++++++++++++++++++++++ Valeri Galtsev Sr System Administrator Department of Astronomy and Astrophysics Kavli Institute for Cosmological Physics University of Chicago Phone: 773-702-4247 ++++++++++++++++++++++++++++++++++++++++
On 2014-08-23, Valeri Galtsev galtsev@kicp.uchicago.edu wrote:
The suggestion you made to switch to commercial system [sorry I brought your suggestion one step further in the same direction, oh I'm really tricky person] is quite in line with what commercial vendors would like to happen to free (as free beer) competitive software: users, feel this free software is as nasty as our commercial alternative is.
I don't think that's precisely the issue. The issue (to me anyway) is that people are complaining about free software *whose explicitly stated goal is to remain as closely as possible to the commercial upstream*. If this were a base distro like Debian or Slackware, then people could legitimately complain that Debian was moving to systemd, because the Debian maintainers made that decision. The CentOS maintainers did not!
So it's not really about free vs. nonfree, it's about who the deciders are.
Since I mentioned it, Slackware might be a reasonable compromise for those of you who prefer a more ''purist'' (whatever that means) environment but don't want to completely break away from linux. When I was an active Slackware user I heard the comparison that Slackware was the most *BSD-like of the linux distros.
--keith
-
Bernard Lheureux -
Christof Stocker -
Digimer -
John R Pierce -
John R. Dennison -
Keith Keller -
Les Mikesell -
m.roth＠5-cent.us -
me＠tdiehl.org -
Steve Clark -
Valeri Galtsev -
William Woods