Page 3 of 4 FirstFirst 1234 LastLast
Results 21 to 30 of 35

Thread: Fedora 20 Goes For No Default Sendmail, Syslog

  1. #21
    Join Date
    Nov 2011
    Posts
    271

    Default

    Quote Originally Posted by Rallos Zek View Post
    Amen! Thank g-d there are a few distros staying way from the SystemD crapfest.
    Sysv isn't? In a matter of times, they will end up using systemd or remain irrelevent. If SUN, Windows, Apple have their own init, why not a Linux distribution to cover a common ground to achieve better performance and efficiency.

  2. #22
    Join Date
    Jan 2013
    Posts
    1,116

    Default

    Quote Originally Posted by finalzone View Post
    Sysv isn't? In a matter of times, they will end up using systemd or remain irrelevent.
    If a distro is forced to use a specific software or they will become irrelevant then Linux will become irrelevant, since it will loose one of its major advantages.

  3. #23
    Join Date
    May 2012
    Posts
    435

    Default

    Quote Originally Posted by finalzone View Post
    Sysv isn't? In a matter of times, they will end up using systemd or remain irrelevent. If SUN, Windows, Apple have their own init, why not a Linux distribution to cover a common ground to achieve better performance and efficiency.
    yes, it isn't
    init currently uses 100k ram (64bit, 32bit probably less)
    syslogd uses 360k
    klogd uses 280k
    ntpd 700k
    acpid 124k
    dbus-deamon 1400k
    crond 452k
    atd 132k

    all of them together have used ~5 sec of cpu time (comp on for 14 hours)
    out of that 2 sec ntpd and 1.63 sec dbus (dbus was made for gnome and kde things, dont know how many non-that programs actually use it)

    and i can turn them on/off or enable/disable with one simple command, like systemd can too

    better performance and efficiency is not the goal of systemd, integration is


    what i'm trying to say:
    have a healthy argument, dont just spread FUD (like Lennart does)
    Last edited by gens; 09-16-2013 at 04:32 PM.

  4. #24
    Join Date
    Jan 2009
    Posts
    1,331

    Default

    Quote Originally Posted by gens View Post
    yes, it isn't
    init currently uses 100k ram (64bit, 32bit probably less)
    syslogd uses 360k
    klogd uses 280k
    ntpd 700k
    acpid 124k
    dbus-deamon 1400k
    crond 452k
    atd 132k

    all of them together have used ~5 sec of cpu time (comp on for 14 hours)
    out of that 2 sec ntpd and 1.63 sec dbus (dbus was made for gnome and kde things, dont know how many non-that programs actually use it)

    and i can turn them on/off or enable/disable with one simple command, like systemd can too

    better performance and efficiency is not the goal of systemd, integration is


    what i'm trying to say:
    have a healthy argument, dont just spread FUD (like Lennart does)

    First, why don't you go to lennart's blog and tell him what he is missing.
    Second, I would argue the motivating factor behind systemd was the inability of the various init systems to reliably track processes in terms of resources, dependencies, and forking/exec'ing. That was the reason, as I understand it, behind the very early decision to use cgroups in systemd.
    Third, the barrier for new sysadmins is much lower when learning systemd practices than sysv + sysconfig + syslog + whateverelse.
    Fourth, from what I've heard of the systemd testing, sysadmins really like it. It is well documented, and consistent.

  5. #25
    Join Date
    May 2012
    Posts
    435

    Default

    Quote Originally Posted by liam View Post
    Third, the barrier for new sysadmins is much lower when learning systemd practices than sysv + sysconfig + syslog + whateverelse.
    Fourth, from what I've heard of the systemd testing, sysadmins really like it. It is well documented, and consistent.
    barrier to init is shell scripting
    cgroups are well documented
    shell scripting language is well documented
    sysadmins should have a choice, not an ultimatum

    also binary logs do not make things easier, id think (am no admin)

    as i said in another thread;
    this documentation starts by talking bad about the former system, in parts with semi lies - lies
    i also said i dont wanna fight on what is "better", as it depends on goals

    with udev being integrated into systemd there is less choice
    and with linux being about choice.. wait what


    what bothers me, and why i write here is "its 30 years old so OMFG IT SUX"
    thats not a good view on anything
    if you look you will find no valid reason for saying systemd is better then init+couple daemons
    (faster boot is also not a reason as it can be done in shell too)

    think about it
    its 30 years old and there were many smart people looking over it, and none seen a problem
    why is that ?
    what is that problem ?


    personally;
    not using systemd only limits me from using gnome, that i dont use

    even more personally;
    i would tell him
    but i dont care about him, i care about the knowledge and mentality of advanced computer users


    TO REPEAT; THERE'S (ALMOST) NOTHING WRONG WITH SYSTEMD, BUT IT SHOULD NOT BE FORCED UPON ANYBODY
    good thing it cant be

  6. #26
    Join Date
    Jan 2009
    Posts
    1,331

    Default

    Quote Originally Posted by gens View Post
    barrier to init is shell scripting
    cgroups are well documented
    shell scripting language is well documented
    sysadmins should have a choice, not an ultimatum

    also binary logs do not make things easier, id think (am no admin)

    as i said in another thread;
    this documentation starts by talking bad about the former system, in parts with semi lies - lies
    i also said i dont wanna fight on what is "better", as it depends on goals

    with udev being integrated into systemd there is less choice
    and with linux being about choice.. wait what


    what bothers me, and why i write here is "its 30 years old so OMFG IT SUX"
    thats not a good view on anything
    if you look you will find no valid reason for saying systemd is better then init+couple daemons
    (faster boot is also not a reason as it can be done in shell too)

    think about it
    its 30 years old and there were many smart people looking over it, and none seen a problem
    why is that ?
    what is that problem ?


    personally;
    not using systemd only limits me from using gnome, that i dont use

    even more personally;
    i would tell him
    but i dont care about him, i care about the knowledge and mentality of advanced computer users


    TO REPEAT; THERE'S (ALMOST) NOTHING WRONG WITH SYSTEMD, BUT IT SHOULD NOT BE FORCED UPON ANYBODY
    good thing it cant be
    You seem like you're unwilling to be swayed (you didn't address my points, you just spoke of how things should be) so I don't see the point in continuning.

  7. #27
    Join Date
    Nov 2011
    Posts
    271

    Default

    Quote Originally Posted by Vim_User View Post
    If a distro is forced to use a specific software or they will become irrelevant then Linux will become irrelevant, since it will loose one of its major advantages.
    In this case, that specific software is systemd designed to take further advantage of Linux kernel and adopted by an increasing amount of major distributions meaning its benefits outweighs its shortcoming.

  8. #28
    Join Date
    Jan 2013
    Posts
    525

    Default

    Quote Originally Posted by finalzone View Post
    In this case, that specific software is systemd designed to take further advantage of Linux kernel and adopted by an increasing amount of major distributions meaning its benefits outweighs its shortcoming.
    I take solace in the fact that it will be a cold day in hell before some distros like Gentoo, Slackware, Puppy Linux, any of the BSDs, Solaris, Minix and Ubuntu will adopt systemd.
    Competition is always a good thing.

  9. #29
    Join Date
    Jan 2013
    Posts
    1,116

    Default

    Quote Originally Posted by finalzone View Post
    In this case, that specific software is systemd designed to take further advantage of Linux kernel and adopted by an increasing amount of major distributions meaning its benefits outweighs its shortcoming for those distributions, but not for those that refuse to adopt it, obviously..
    Fixed that for you. It does not matter at all if some distributions, even if it are major ones, adopt systemd. That does not mean at all that every distribution has to adopt it.

  10. #30

    Default

    Quote Originally Posted by Vim_User View Post
    Fixed that for you. It does not matter at all if some distributions, even if it are major ones, adopt systemd. That does not mean at all that every distribution has to adopt it.
    Yes but it does mean that distributions not adopting it should cope with the changes upstream one way or the other and they are.

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •