Broadcasting Annoyances

[Janina Sajka]

Am I the only person who's sick and tired of all manner of error
messages that are being splattered across all open console logins just
because some programmer wants to do that?

Example: Is it really necessary for ordinary users to get the following
from cdparanoia:

cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the actual device will
also suggest you mig


As if the ordinary user (who, in my instance didn't even launch the
process) could do anything about such a condition. Heck, I'd warrant
most users wouldn't even understand it. Yet, in this example, this
messages comes up every minute or so.

I could come up with other examples, but I think my point is
sufficiently illustrated.

Clearly such messages should go to logs. But, isn't this broadcast
abuse? I think so and dearly wish we had a mechanism to block such
abuse.

				Janina

[Michael Tiemann]

Perhaps you want to be using a terminal window, not a console, for
command-line activities.  I know the distinction may come as a surprise,
but terminals will give you the peace you seek.  I never open console
windows (well, almost never).

M

On Wed, 2004-08-11 at 09:05, Janina Sajka wrote:
>  Am I the only person who's sick and tired of all manner of error
>  messages that are being splattered across all open console logins just
>  because some programmer wants to do that?
>  
>  Example: Is it really necessary for ordinary users to get the following
>  from cdparanoia:
>  
>  cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the actual device will
> also suggest you mig
>  
>  
>  As if the ordinary user (who, in my instance didn't even launch the
>  process) could do anything about such a condition. Heck, I'd warrant
>  most users wouldn't even understand it. Yet, in this example, this
>  messages comes up every minute or so.
>  
>  I could come up with other examples, but I think my point is
>  sufficiently illustrated.
>  
>  Clearly such messages should go to logs. But, isn't this broadcast
>  abuse? I think so and dearly wish we had a mechanism to block such
>  abuse.
>  
>  				Janina
>

[Janina Sajka]

Not an option for me (in my eyes-free environment).

And, even if it were, is this really a solution? Or just a workaround?

Michael Tiemann writes:
>  Perhaps you want to be using a terminal window, not a console, for
>  command-line activities.  I know the distinction may come as a surprise,
>  but terminals will give you the peace you seek.  I never open console
>  windows (well, almost never).
>  
>  M
>  
>  On Wed, 2004-08-11 at 09:05, Janina Sajka wrote:
>  > Am I the only person who's sick and tired of all manner of error
>  > messages that are being splattered across all open console logins just
>  > because some programmer wants to do that?
>  > 
>  > Example: Is it really necessary for ordinary users to get the following
>  > from cdparanoia:
>  > 
>  > cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the actual device
> will also suggest you mig
>  > 
>  > 
>  > As if the ordinary user (who, in my instance didn't even launch the
>  > process) could do anything about such a condition. Heck, I'd warrant
>  > most users wouldn't even understand it. Yet, in this example, this
>  > messages comes up every minute or so.
>  > 
>  > I could come up with other examples, but I think my point is
>  > sufficiently illustrated.
>  > 
>  > Clearly such messages should go to logs. But, isn't this broadcast
>  > abuse? I think so and dearly wish we had a mechanism to block such
>  > abuse.
>  > 
>  > 				Janina
>  > 
>  
>  
>  -- 
>  fedora-devel-list mailing list
>  fedora-devel-list(a)redhat.com
>  http://www.redhat.com/mailman/listinfo/fedora-devel-list 
-- 
	
				Janina Sajka, Chair
				Accessibility Workgroup
				Free Standards Group (FSG)

janina(a)freestandards.org	Phone: [phone]

[Michael Tiemann]

Perhaps I don't know enough about your environment, but there is a
definite distinction in my mind between 'console' and 'terminal'.  Here
are some messages that reinforce that difference:

http://www.ussg.iu.edu/hypermail/linux/kernel/9704.3/0467.html
http://lists.debian.org/debian-user/2002/06/msg00138.html

It could be that the way you have set up your login environment, the
user's only option is a 'console', but I think you should evaluate
making it possible for users who need terminal access (even if it's a
terminal providing eye-free interaction) to use a 'terminal' instead.

M

On Wed, 2004-08-11 at 09:18, Janina Sajka wrote:
>  Not an option for me (in my eyes-free environment).
>  
>  And, even if it were, is this really a solution? Or just a workaround?
>  
>  Michael Tiemann writes:
>  > Perhaps you want to be using a terminal window, not a console, for
>  > command-line activities.  I know the distinction may come as a surprise,
>  > but terminals will give you the peace you seek.  I never open console
>  > windows (well, almost never).
>  > 
>  > M
>  > 
>  > On Wed, 2004-08-11 at 09:05, Janina Sajka wrote:
>  > > Am I the only person who's sick and tired of all manner of error
>  > > messages that are being splattered across all open console logins just
>  > > because some programmer wants to do that?
>  > > 
>  > > Example: Is it really necessary for ordinary users to get the following
>  > > from cdparanoia:
>  > > 
>  > > cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the actual
> device will also suggest you mig
>  > > 
>  > > 
>  > > As if the ordinary user (who, in my instance didn't even launch the
>  > > process) could do anything about such a condition. Heck, I'd warrant
>  > > most users wouldn't even understand it. Yet, in this example, this
>  > > messages comes up every minute or so.
>  > > 
>  > > I could come up with other examples, but I think my point is
>  > > sufficiently illustrated.
>  > > 
>  > > Clearly such messages should go to logs. But, isn't this broadcast
>  > > abuse? I think so and dearly wish we had a mechanism to block such
>  > > abuse.
>  > > 
>  > > 				Janina
>  > > 
>  > 
>  > 
>  > -- 
>  > fedora-devel-list mailing list
>  > fedora-devel-list(a)redhat.com
>  > http://www.redhat.com/mailman/listinfo/fedora-devel-list
>  
>  -- 
>  	
>  				Janina Sajka, Chair
>  				Accessibility Workgroup
>  				Free Standards Group (FSG)
>  
>  janina(a)freestandards.org	Phone: [phone]
>

[Janina Sajka]

OK. I think I did misunderstand you--you're not speaking of  Xterm (or
gnome terminal, etc).
I'll be happy to clean up my act on this, if that will do the trick.
Thanks for the pointer. Guess I need to understand better the
distinction in the context of Fedora--the various ~/.bashrc,
~/.bash_profile, etc. 

Michael Tiemann writes:
>  Perhaps I don't know enough about your environment, but there is
> a
>  definite distinction in my mind between 'console' and 'terminal'.  Here
>  are some messages that reinforce that difference:
>  
>  http://www.ussg.iu.edu/hypermail/linux/kernel/9704.3/0467.html
>  http://lists.debian.org/debian-user/2002/06/msg00138.html
>  
>  It could be that the way you have set up your login environment, the
>  user's only option is a 'console', but I think you should evaluate
>  making it possible for users who need terminal access (even if it's a
>  terminal providing eye-free interaction) to use a 'terminal' instead.
>  
>  M
>  
>  On Wed, 2004-08-11 at 09:18, Janina Sajka wrote:
>  > Not an option for me (in my eyes-free environment).
>  > 
>  > And, even if it were, is this really a solution? Or just a workaround?
>  > 
>  > Michael Tiemann writes:
>  > > Perhaps you want to be using a terminal window, not a console, for
>  > > command-line activities.  I know the distinction may come as a surprise,
>  > > but terminals will give you the peace you seek.  I never open console
>  > > windows (well, almost never).
>  > > 
>  > > M
>  > > 
>  > > On Wed, 2004-08-11 at 09:05, Janina Sajka wrote:
>  > > > Am I the only person who's sick and tired of all manner of error
>  > > > messages that are being splattered across all open console logins just
>  > > > because some programmer wants to do that?
>  > > > 
>  > > > Example: Is it really necessary for ordinary users to get the following
>  > > > from cdparanoia:
>  > > > 
>  > > > cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the
> actual device will also suggest you mig
>  > > > 
>  > > > 
>  > > > As if the ordinary user (who, in my instance didn't even launch the
>  > > > process) could do anything about such a condition. Heck, I'd warrant
>  > > > most users wouldn't even understand it. Yet, in this example, this
>  > > > messages comes up every minute or so.
>  > > > 
>  > > > I could come up with other examples, but I think my point is
>  > > > sufficiently illustrated.
>  > > > 
>  > > > Clearly such messages should go to logs. But, isn't this broadcast
>  > > > abuse? I think so and dearly wish we had a mechanism to block such
>  > > > abuse.
>  > > > 
>  > > > 				Janina
>  > > > 
>  > > 
>  > > 
>  > > -- 
>  > > fedora-devel-list mailing list
>  > > fedora-devel-list(a)redhat.com
>  > > http://www.redhat.com/mailman/listinfo/fedora-devel-list
>  > 
>  > -- 
>  > 	
>  > 				Janina Sajka, Chair
>  > 				Accessibility Workgroup
>  > 				Free Standards Group (FSG)
>  > 
>  > janina(a)freestandards.org	Phone: [phone]
>  > 
>  
>  
>  -- 
>  fedora-devel-list mailing list
>  fedora-devel-list(a)redhat.com
>  http://www.redhat.com/mailman/listinfo/fedora-devel-list 
-- 
	
				Janina Sajka, Chair
				Accessibility Workgroup
				Free Standards Group (FSG)

janina(a)freestandards.org	Phone: [phone]

[Chris Geddings]

If you are launching things from the command line, you could redirect 
STDERR into a log file or /dev/null.
Depending on exactly which output you are referring to, you may also 
wish to remove any lines in syslog.conf that refer to
  /dev/console.

--Chris

On Aug 11, 2004, at 9:18 AM, Janina Sajka wrote:

>  Not an option for me (in my eyes-free environment).
>
>  And, even if it were, is this really a solution? Or just a workaround?

[Peter Jones]

On Wed, 2004-08-11 at 09:50 -0400, Chris Geddings wrote:
>  If you are launching things from the command line, you could redirect
>
>  STDERR into a log file or /dev/null.
>  Depending on exactly which output you are referring to, you may also 
>  wish to remove any lines in syslog.conf that refer to
>    /dev/console. 
The error messages being discussed aren't stderr -- they're kernel
warnings when applications do Bad Things[1].

[1] This particular bad thing I happen to be responsible for -- there is
a test version of cdparanoia at
http://people.redhat.com/pjones/cdparanoia .   .sgio3 is known not to
work perfectly on USB drives, but I think it works ok for everything
else.  I'll probably release another version there sometime later this
week which should work with USB drives as well.

-- 
        Peter

[Chris Geddings]

Are these having a "wall" like effect, or being relayed through syslog,
or something else?

--Chris, just curious

On Wed, 2004-08-11 at 15:09, Peter Jones wrote:

>  The error messages being discussed aren't stderr -- they're
> kernel
>  warnings when applications do Bad Things[1].

[Arjan van de Ven]

On Wed, 2004-08-11 at 15:05, Janina Sajka wrote:
>  Am I the only person who's sick and tired of all manner of error
>  messages that are being splattered across all open console logins just
>  because some programmer wants to do that? 

you can control which level of severity this happens for. It seems for
you it happens for non-fatal kernel warnings already...

[Leonard den Ottolander]

Hello Janina,

On Wed, 2004-08-11 at 15:05, Janina Sajka wrote:
>  Am I the only person who's sick and tired of all manner of error
>  messages that are being splattered across all open console logins just
>  because some programmer wants to do that? 
/etc/syslog.conf can be edited to change syslogs behaviour. Also the
sysctl.conf kernel.printk values can be changed to set klogd's
verbosity.

On my firewall I use:
in /etc/sysctl.conf:
kernel.printk = 4 4 1 7
so only warnings or worse are logged to the console

in /etc/syslog.conf:
kern.warning				/dev/console
kern.notice				/dev/tty9
*.debug					/dev/tty10
(not sure if the first line is redundant, check for yourself)

Leonard.

-- 
mount -t life -o ro /dev/dna /genetic/research

[Stephen Smoogen]

Janina Sajka wrote:
>  Am I the only person who's sick and tired of all manner of error
>  messages that are being splattered across all open console logins just
>  because some programmer wants to do that?
>  
>  Example: Is it really necessary for ordinary users to get the following
>  from cdparanoia:
>  
>  cdparanoia: Using deprecated /dev/sg mechanism instead of SG_IO on the actual device will
> also suggest you mig
>   
For most kernel errors, one can edit /etc/sysconfig/syslog and in the 
klog section

# Options to syslogd
# -m 0 disables 'MARK' messages.
# -r enables logging from remote machines
# -x disables DNS lookups on messages recieved with -r
# See syslogd(8) for more details
SYSLOGD_OPTIONS="-m 0"
# Options to klogd
# -2 prints all kernel oops messages twice; once for klogd to decode, and
#    once for processing with 'ksymoops'
# -x disables all klogd processing of oops messages entirely
# See klogd(8) for more details
KLOGD_OPTIONS="-x -c 1"

That should quiet out a lot of cruft that kernel sometimes prints.

>   

-- 
Stephen John Smoogen		smoogen(a)lanl.gov
Los Alamos National Lab  CCN-5 Sched 5/40  PH: 4-0645
Ta-03 SM-1498 MailStop B255 DP 10S  Los Alamos, NM 87545