[arch-general] /etc/dbus-1/systemd/cups.conf and /usr/lib/systemd/system/cups.socket files
After my research on sysvinit, the simple version, I'm working my way through systemd. Two things I noticed about the cups configuration: 1: The cups.conf file in /etc/dbus-1/systemd reads <!DOCTYPE busconfig PUBLIC "-//freedesktop//DTD D-BUS Bus Configuration 1.0//EN" "http://www.freedesktop.org/standards/dbus/1.0/busconfig.dtd"> <busconfig> <!-- Only root can send this message --> <policy user="root"> <allow send_interface="com.redhat.PrinterSpooler"/> </policy> <!-- Allow any connection to receive the message --> <policy context="default"> <allow receive_interface="com.redhat.PrinterSpooler"/> </policy> </busconfig> Do the allow send and allow receive interfaces need to be changed, or is this taken into account by the rest of systemd? 2. The /usr/lib/systemd/system/cups.socket reads [Unit] Description=CUPS Printing Service Sockets [Socket] ListenStream=/var/run/cups/cups.sock ListenStream=631 ListenDatagram=0.0.0.0:631 BindIPv6Only=ipv6-only [Install] WantedBy=sockets.target What is the recommended best practice for dealing with this when you have no access to IPv6? Our satellite console supports IPv4 only so previously everything has been set up on 192.168.???.??. With the newer version of cups depending on avahi and the differences in the avahi hosts file and the hosts file, avahi uses gandalf.local while hosts still uses .localdomain. https://lwn.net/Articles/485617/ I'm not trying to be dense here, but some of this seems to be at cross purposes. That or this is something sysadmins on large systems do all the time and those of us with small installations haven't had to deal with yet. As with some of my other posts, trashing systemd etc and those behind it accomplishes nothing. If you can't say anything constructive, STFU. I apologize for offending anyone, but this would be easier with out all the BS, and I don't mean bikeshed. Myra -- Life's fun when your sick and psychotic!
On Wed, Aug 22, 2012 at 7:31 PM, Myra Nelson <myra.nelson@hughes.net> wrote:
After my research on sysvinit, the simple version, I'm working my way through systemd. Two things I noticed about the cups configuration:
1: The cups.conf file in /etc/dbus-1/systemd reads
Notice that this should have been "system.d" and not "systemd". These are the dbus configuration files (for the system bus), and not the systemd ones.
<!DOCTYPE busconfig PUBLIC "-//freedesktop//DTD D-BUS Bus Configuration 1.0//EN" "http://www.freedesktop.org/standards/dbus/1.0/busconfig.dtd"> <busconfig> <!-- Only root can send this message --> <policy user="root"> <allow send_interface="com.redhat.PrinterSpooler"/> </policy>
<!-- Allow any connection to receive the message --> <policy context="default"> <allow receive_interface="com.redhat.PrinterSpooler"/> </policy> </busconfig>
Do the allow send and allow receive interfaces need to be changed, or is this taken into account by the rest of systemd?
Sorry that I can't help you with more details, don't know cups/dbus that well.
2. The /usr/lib/systemd/system/cups.socket reads [Unit] Description=CUPS Printing Service Sockets
[Socket] ListenStream=/var/run/cups/cups.sock ListenStream=631 ListenDatagram=0.0.0.0:631 BindIPv6Only=ipv6-only
[Install] WantedBy=sockets.target
Again, I don't know much about CUPS, so can't answer your question. However, just in case it was not clear, the above socket file gives you: # ss -l -u -t | grep ipp udp UNCONN 0 0 *:ipp *:* tcp LISTEN 0 128 :::ipp :::* Where the tcp socket only accepts ipv6 connections. Please see systemd.socket(5) and ipv6(7) for more details.
With the newer version of cups depending on avahi and the differences in the avahi hosts file and the hosts file, avahi uses gandalf.local while hosts still uses .localdomain.
Not sure I got this last bit. Avahi defaults to using the "local" domain, but this can be configured if you prefer something else (at least I seem to remember doing that at some point). HTH, Tom
On Wed, Aug 22, 2012 at 1:37 PM, Tom Gundersen <teg@jklm.no> wrote:
On Wed, Aug 22, 2012 at 7:31 PM, Myra Nelson <myra.nelson@hughes.net> wrote:
After my research on sysvinit, the simple version, I'm working my way through systemd. Two things I noticed about the cups configuration:
1: The cups.conf file in /etc/dbus-1/systemd reads
Notice that this should have been "system.d" and not "systemd". These are the dbus configuration files (for the system bus), and not the systemd ones.
<!DOCTYPE busconfig PUBLIC "-//freedesktop//DTD D-BUS Bus Configuration 1.0//EN" "http://www.freedesktop.org/standards/dbus/1.0/busconfig.dtd"> <busconfig> <!-- Only root can send this message --> <policy user="root"> <allow send_interface="com.redhat.PrinterSpooler"/> </policy>
<!-- Allow any connection to receive the message --> <policy context="default"> <allow receive_interface="com.redhat.PrinterSpooler"/> </policy> </busconfig>
Do the allow send and allow receive interfaces need to be changed, or is this taken into account by the rest of systemd?
Sorry that I can't help you with more details, don't know cups/dbus that well.
2. The /usr/lib/systemd/system/cups.socket reads [Unit] Description=CUPS Printing Service Sockets
[Socket] ListenStream=/var/run/cups/cups.sock ListenStream=631 ListenDatagram=0.0.0.0:631 BindIPv6Only=ipv6-only
[Install] WantedBy=sockets.target
Again, I don't know much about CUPS, so can't answer your question. However, just in case it was not clear, the above socket file gives you:
# ss -l -u -t | grep ipp udp UNCONN 0 0 *:ipp *:* tcp LISTEN 0 128 :::ipp :::*
Where the tcp socket only accepts ipv6 connections. Please see systemd.socket(5) and ipv6(7) for more details.
With the newer version of cups depending on avahi and the differences in the avahi hosts file and the hosts file, avahi uses gandalf.local while hosts still uses .localdomain.
Not sure I got this last bit. Avahi defaults to using the "local" domain, but this can be configured if you prefer something else (at least I seem to remember doing that at some point).
HTH,
Tom
Cool. I was just trying to make both the domains come out the same. I've tried it both ways, but can't get cups to work between the two boxes. Let me rephrase that. I have the printer set up on my SO's box and working. Then sudo firefox localhost:631, find/discover printers, finds the printer, attempt to connect and test the printer with a test page, "error can't find printer at arwen.localdomain". The pemissions on her box are about as lenient as you can get "Allow all" everywhere. Since all my network connections are static, and all the documentation I've ever read pointed to setting up the hosts file at 127.0.0.1 localhost.localdomain localhost gandalf and ::01 localhost.localdomain, and after reading the LWN article, I thought that might be the problem hence my question. On to more research. Thanks for you time and assistance. Myra -- Life's fun when your sick and psychotic!
On Wed, Aug 22, 2012 at 4:04 PM, Myra Nelson <myra.nelson@hughes.net> wrote:
On Wed, Aug 22, 2012 at 1:37 PM, Tom Gundersen <teg@jklm.no> wrote:
On Wed, Aug 22, 2012 at 7:31 PM, Myra Nelson <myra.nelson@hughes.net> wrote:
After my research on sysvinit, the simple version, I'm working my way through systemd. Two things I noticed about the cups configuration:
1: The cups.conf file in /etc/dbus-1/systemd reads
Notice that this should have been "system.d" and not "systemd". These are the dbus configuration files (for the system bus), and not the systemd ones.
<!DOCTYPE busconfig PUBLIC "-//freedesktop//DTD D-BUS Bus Configuration 1.0//EN" "http://www.freedesktop.org/standards/dbus/1.0/busconfig.dtd"> <busconfig> <!-- Only root can send this message --> <policy user="root"> <allow send_interface="com.redhat.PrinterSpooler"/> </policy>
<!-- Allow any connection to receive the message --> <policy context="default"> <allow receive_interface="com.redhat.PrinterSpooler"/> </policy> </busconfig>
Do the allow send and allow receive interfaces need to be changed, or is this taken into account by the rest of systemd?
Sorry that I can't help you with more details, don't know cups/dbus that well.
2. The /usr/lib/systemd/system/cups.socket reads [Unit] Description=CUPS Printing Service Sockets
[Socket] ListenStream=/var/run/cups/cups.sock ListenStream=631 ListenDatagram=0.0.0.0:631 BindIPv6Only=ipv6-only
[Install] WantedBy=sockets.target
Again, I don't know much about CUPS, so can't answer your question. However, just in case it was not clear, the above socket file gives you:
# ss -l -u -t | grep ipp udp UNCONN 0 0 *:ipp *:* tcp LISTEN 0 128 :::ipp :::*
Where the tcp socket only accepts ipv6 connections. Please see systemd.socket(5) and ipv6(7) for more details.
With the newer version of cups depending on avahi and the differences in the avahi hosts file and the hosts file, avahi uses gandalf.local while hosts still uses .localdomain.
Not sure I got this last bit. Avahi defaults to using the "local" domain, but this can be configured if you prefer something else (at least I seem to remember doing that at some point).
HTH,
Tom
Cool. I was just trying to make both the domains come out the same. I've tried it both ways, but can't get cups to work between the two boxes. Let me rephrase that. I have the printer set up on my SO's box and working. Then sudo firefox localhost:631, find/discover printers, finds the printer, attempt to connect and test the printer with a test page, "error can't find printer at arwen.localdomain". The pemissions on her box are about as lenient as you can get "Allow all" everywhere.
Since all my network connections are static, and all the documentation I've ever read pointed to setting up the hosts file at 127.0.0.1 localhost.localdomain localhost gandalf and ::01 localhost.localdomain, and after reading the LWN article, I thought that might be the problem hence my question. On to more research.
Thanks for you time and assistance.
Myra
-- Life's fun when your sick and psychotic!
Tom: I got cups fixed with a little voodoo. Now for the bad news. I've been working on the systemd stuff and moving along, that is until I read news feed on Planet Arch by Dave Reisner about using startx and xinit. Personally, I don't care if xinit is a "stupid program", and "startx is a glorified compiled shell script with some extra signal handling", they, IMO -- nothing humble about it, work and give me a great deal of flexiblity in my system. Since we seem to use broken shell scripts for the old init system, and that's bad, here we are adding to that by scripting the behaviour of xinit and startx to fit with systemd. Something about that just doesn't seem kosher. I want to reiterate something. I don't know Lennart Poettering, Dave Reisner, or you for that matter, nor do I have any opinions on your successes for failures. But based on the derogatory nature of most of Mr. Reisner's comments, I would have to say he is one of the few people in the world with a bigger ego than mine, but as one of my former employers stated "at least you go out and back it up". I thought, as did many people I've worked for and with, I had a lock on the worst attitude in the world towards other, references available, but I think I found my match. I've also developed an aversion to Dave's approach, here are some examples:
do you appreciate the system requiring an unknown amount of your (limited) time each day you decide to update? don't you ever wish you could just say "hey computer 'ol pal, aggressively follow upstream > source for package X and merge remote user Y's with the local configuration, unless either requires changes to package Z -- then ask me first, cuz i run the show here"?
Yes I do appreciate the system requiring an amount of my time each day. The day the system is in control they can bury me face down so the world can X@(W (* Q)Y.
SYSVINIT HAS NO POWER, NO FUNCTIONALITY, AND ABSOLUTELY ZERO USEFULNESS ON IT'S OWN.
Neither does systemd!!!!!!
the "unix philosophy" (debatable in itself...) of doing one thing doesn't usually translate to LITERALLY DOING ONE THING.
Only because "modern programmers" want things so tightly bound together they break this principle!!!!
please please please, please, read that several times until it sinks in nice and deep; every single argument about sysvinit's "simplicity", "maturity", blah blah, woka woka, etc etc, and anything else related to it's stability is complete nonsense because ... YOUR BOOT PROCESS IS A MEGAMATRON SHELL SCRIPT, AND HAS PEANUTS TO DO WITH INIT; IN FACT, INIT IS NOT EVEN NEEDED.
I have and my IQ of 17, " we grant these applications the most awesomest superpowers then give it the IQ of 17 or so." my 113 says they seem to work fine and I might be able to fix them right here instead of sending a patch upstream for those who aren't really interested. I may be branded a luddite, old fashioned, stubborn, backwards, stupid, ignorant, etc, and as Mr. Reisner so eloqently stated "I'm not sure most of the everyday arch users count" As Tom Gunderson pointed out, this is probably true because the devs build arch for themselves. They are intersted in positive helpful comments and feedback. Some thing Mr. Reisner seems incapable of. Since I'm sure this will be the last post I'm allowed on this list, a little about me. I seem to be able to ask questions that are considered incomprehensible or irrelavant. I realize this is way too much information for this mailing list, but here's why. I'm bipolar, suffer from adult ADHD, I have dain bramage from CO2 poisoning (when I went to ICU my O2 sats were 35% and my CO2 stats were 112), by blood oxygen saturation stays below the 95% recommended for a healthy person (lung damage from H2S and SO2 poisoning and I'm miswired and forget to breathe), I don't leave the house more than once every two weeks so I don't have a panic attack, I don't play well with others, and I suffer from fits of rage which can be uncontrolable. This means I have trouble deciphering documentation without reading it 5 or 6 times, then there is no guarantee. When I work on a program, especially learning a new language, I get to start over everyday becuase what I did the day before it gone - most of the time, When I write something it has to be read and edited by someone or it sounds like I'm an idiot, but I do my best. Maybe I should be using a Mac or Windows, but hey "At least I'm trying", that should count for something. As for the so often mentioned "business benefits" of systemd, I could care less. It was once said at Woodstock "A little capitalism isn't all that bad", but that doesn't mean I want someone in cubicle somewhere trying to tell me how I should use my computer, or that I'm stupid because I don't want to use my computer the way he thinks I should. Home desktop computer != business decision. I would like to apologize for anyone I've offended, not ticked off, offended. Tom: Keep up the good work. Ciao for Now and to all a good night! Myra Nelson -- Life's fun when your sick and psychotic!
participants (2)
-
Myra Nelson
-
Tom Gundersen