Change to semantics of display power management
by Geoffrey Leach
This appears to have happened recently, but just when I can't say. Is it a bug or a feature?
Settings->Screensaver->Advanced->Display Power Management->Standby After
When Standby is activated, in addition to blanking the screen the code now (this is new) reduces the power draw without actually shutting off power (as it would for a suspend). I know this, because (a) the power-on light remains lit, (b) touching a key restores the display and (c) my Tripp-Lite power controller sees the power drop and shuts off connected devices.
Fedora 19 up-to-date
Any thoughts?
10 years, 3 months
suspend or hibernate
by Richard Vickery
Hi there:
I just called up the gnome-tweak-tool: what's the difference between
suspend and hibernate? It gives these, among other sleeping actions when
folding the computer up.
Just curious - hibernate doesn't have a man page.
Thanks
--
Richard
10 years, 3 months
high CPU use by bijiben-shell-search-provider after returning from "Activities"
by Andre Robatino
On F20, frequently, after returning from "Activities", I see
bijiben-shell-search-provider using near 100% CPU, and I have to kill it
(-TERM suffices). I am using the nvidia driver, not nouveau. Has anyone else
seen this? Something similar happens if I try to launch quadrapassel, it
also uses near 100% CPU but the app never appears.
I found the existing bug https://bugzilla.redhat.com/show_bug.cgi?id=1017978
which was closed as NOTABUG by the original reporter. I reopened it. The
correct Component is probably not bijiben but I have no idea what it should be.
10 years, 3 months
Re: F20 - Unintended consequences of no default MTA - How best to fix
by Rick Walker
OK.
Can we make journald have another configurable option? If it were
possible to simulate this this:
echo "Problem in the pit, Boss" | mail my_foreman
With something like
echo "Problem in the pit, Boss" | sendtojournald -also_mail my_foreman
Then you'd be able to get the people who like to send mail to
start sending mail in a way that gets them also using journald.
In addition, you could make it so that EITHER an MTA is installed
and /usr/bin/mail actuallly sends mail OR NO MTA is installed and
/usr/bin/mail is a symlink to journald which acts in many ways
like /usr/bin/mail, but which can optionally forward messages to
external mailservers.
That would be what I would call the traditional way of making such an
extension to linux. All new code strives to work with the simplified
old methods and understanding, and a more sophisticated user just gets
more functionality (albeit at a steeper learning curve). In some cases,
one would hope that the "new" way is actually a simpler or more
powerful way of doing things too.
What has been a bit scarey is seeing a trend towards making
things "harder" or more complex to learn and use without any obvious
benefit to the legacy user. We should strive to let a "pure text
console" user to at least be able to do the things they always used to do
at at least the same level that they used to work at.
Since many of the new generation don't use the command line very
much and don't really know what a good coder can do with command
line recall and scripting, they might be humble enough to at
least preserve those semantics as sacred until such time that
their chops are sufficient to understand the richness of the
environment.
A particularly funny circumstance is the program that
autoprobes any RS-232 device at a regular basis for the presence of a
modem. My god. This had some funny side-effects for anyone using
RS-232 and linux to control servo systems, 100KWatt-crystal-furnaces,
robots, and other such stuff. Sending binary characters out a generic
interface port to probe for the latest socially cool device probably
shouldn't have been put in the mainline code as a default.
If we progress forward in a way that honors the past, then we have a
chance of building a computing culture that has a chance of improving
not only on the short time scale, but the long time scale and which
works brilliantly as a GUI, but also as a text-based problem-solving and
programming platform.
OK. I'll stop now. This is perilously close to ranting. My apologies.
kind regards,
--
Rick Walker
10 years, 3 months
dnf update -
by Bob Goodwin
This is a Fedora-20 64 bit XFCE system.
Only once has this produced an update for me:
[root@box10 bobg]# dnf update
Resolving dependencies
--> Starting dependency resolution
--> Finished dependency resolution
Dependencies resolved.
Nothing to do.
But yum update a moment later:
Transaction Summary
=========================================================================================================================================================================================
Install 2 Packages
Upgrade 23 Packages
Total download size: 45 M
Is this ok [y/d/N]: y
And it went through to complete this morning's update.
What is wrong? DNF man page shows an update command ,,,
Bob
--
http://www.qrz.com/db/w2bod
Box10 Fedora-20/64bit Linux
10 years, 3 months
Re: F20 - Unintended consequences of no default MTA - How best to fix
by Rick Walker
> Chris Murphy writes:
> Now there will be some additional users talking to upstream
> developers to incorporate better means of notification rather
> than depending on emails that obviously aren't being seen by most
> peopel anyway,
Was there actually a poll taken about who is using mail on their
desktops? I didn't see one, so I don't know the percentage of MTA
users, but as a long-time technical user, I have always run an MTA and
use it often as a lightweight way to communicate results of jobs such as
backups, SPICE simulations, general cron cleanup scripts and so forth.
All it takes is to add something like:
echo "Memory error in cleanup script" | mail staff
Boom. I get timely notification right where I need it - in the
one place that I check multiple times per day - my inbox.
I guess I'll just add one more thing to my post-install checklist.
One of the nice things about UNIX is the clean and minimal ethic that it
was born with. I think that we need to work to keep it as simple as
possible - especially in the core system. Otherwise it becomes nearly
impossible to learn and less maintainable over time.
kind regards,
--
Rick Walker
10 years, 3 months
upgrade to F19: now dhcpd works then fails
by sean darcy
Just upgraded from 18 > 19. dhcpd works, for a while and then stops
recognizing DCHP requests. On restart, it works again.
Here I start tailing /var/log/messages:
Jan 2 11:57:00 new-gateway systemd-logind[316]: New session 6 of user root.
No dhcp log messages. But tcpdump shows lots received:
11:57:01.210319 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:02.210316 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:13.205292 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:14.205304 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:15.205282 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:18.897303 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:0b:82:40:0e:93, length 548
11:57:21.983290 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:0b:82:40:0e:93, length 548
11:57:25.061279 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:0b:82:40:0e:93, length 548
11:57:26.200279 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:27.200284 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:28.199258 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:28.848254 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from cc:6d:a0:11:22:a0, length 548
11:57:39.198233 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:40.195231 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
11:57:41.194227 IP 0.0.0.0.bootpc > 255.255.255.255.bootps: BOOTP/DHCP,
Request from 00:15:99:6a:9c:94, length 554
,,,,,,,,,
Now I systemctl restart dhcpd, and it recognizes all the DHCP requests:
Jan 2 11:59:36 new-gateway systemd[1]: Started DHCPv4 Server Daemon.
Jan 2 11:59:36 new-gateway dhcpd: Internet Systems Consortium DHCP
Server 4.2.5
Jan 2 11:59:37 new-gateway dhcpd: Copyright 2004-2013 Internet Systems
Consortium.
Jan 2 11:59:37 new-gateway dhcpd: All rights reserved.
Jan 2 11:59:37 new-gateway dhcpd: For info, please visit
https://www.isc.org/software/dhcp/
Jan 2 11:59:37 new-gateway dhcpd: Not searching LDAP since ldap-server,
ldap-port and ldap-base-dn were not specified in the config file
Jan 2 11:59:37 new-gateway dhcpd: Wrote 0 deleted host decls to leases
file.
Jan 2 11:59:37 new-gateway dhcpd: Wrote 0 new dynamic host decls to
leases file.
Jan 2 11:59:37 new-gateway dhcpd: Wrote 12 leases to leases file.
Jan 2 11:59:37 new-gateway dhcpd:
Jan 2 11:59:37 new-gateway dhcpd: No subnet declaration for wlan0 (no
IPv4 addresses).
Jan 2 11:59:37 new-gateway dhcpd: ** Ignoring requests on wlan0. If
this is not what
Jan 2 11:59:37 new-gateway dhcpd: you want, please write a subnet
declaration
Jan 2 11:59:37 new-gateway dhcpd: in your dhcpd.conf file for the
network segment
Jan 2 11:59:37 new-gateway dhcpd: to which interface wlan0 is
attached. **
Jan 2 11:59:37 new-gateway dhcpd:
Jan 2 11:59:37 new-gateway dhcpd: Listening on
LPF/external/00:1d:72:05:4a:68/24.0.0.0/8
Jan 2 11:59:37 new-gateway dhcpd: Sending on
LPF/external/00:1d:72:05:4a:68/24.0.0.0/8
Jan 2 11:59:37 new-gateway dhcpd: Listening on
LPF/internal/00:e0:4c:03:07:ba/10.10.11.0/24
Jan 2 11:59:37 new-gateway dhcpd: Sending on
LPF/internal/00:e0:4c:03:07:ba/10.10.11.0/24
Jan 2 11:59:37 new-gateway dhcpd: Sending on Socket/fallback/fallback-net
Jan 2 11:59:38 new-gateway dhcpd: DHCPDISCOVER from 00:15:99:6a:9c:94
via internal
Jan 2 11:59:38 new-gateway dhcpd: DHCPOFFER on 10.10.11.152 to
00:15:99:6a:9c:94 via internal
Jan 2 11:59:38 new-gateway dhcpd: DHCPREQUEST for 10.10.11.152
(10.10.11.251) from 00:15:99:6a:9c:94 via internal
Jan 2 11:59:38 new-gateway dhcpd: DHCPACK on 10.10.11.152 to
00:15:99:6a:9c:94 via internal
Jan 2 11:59:40 new-gateway dhcpd: DHCPREQUEST for 10.10.11.152
(10.10.11.251) from 00:15:99:6a:9c:94 via internal
Jan 2 11:59:40 new-gateway dhcpd: DHCPACK on 10.10.11.152 to
00:15:99:6a:9c:94 via internal
Jan 2 11:59:51 new-gateway dhcpd: DHCPDISCOVER from 00:0b:82:40:0e:93
via internal
Jan 2 11:59:51 new-gateway dhcpd: DHCPOFFER on 10.10.11.182 to
00:0b:82:40:0e:93 via internal
Jan 2 11:59:51 new-gateway dhcpd: DHCPREQUEST for 10.10.11.182
(10.10.11.251) from 00:0b:82:40:0e:93 via internal
Jan 2 11:59:51 new-gateway dhcpd: DHCPACK on 10.10.11.182 to
00:0b:82:40:0e:93 via internal
but later it will stop recognizing the DHCPDISCOVER packets. No log
message about failure or error.
Any help appreciated.
sean
10 years, 3 months
en.mo.gz not found/invalid magic number
by Don Levey
In my research I've not yet seen anything for Fedora about this, and
certainly not the combination of two errors. I'll try to be as complete
as possible:
After a power loss, I was unable to reboot my desktop machine (Dell
Dimension E521 running F19). Upon first hitting grub, I received large
numbers of syntax/bad command errors. I was eventually dumped into a
grub shell, but did not know enough to do anything inside it. Using the
F19 install disk I was able to get to an Anaconda shell. I noted that
while my root directory was mounted to /mnt/sysimage, the /boot
partition was not mounted and I couldn't figure out how to get to it
manually.
I found that adding GRUB_DISABLE_SUBMENU=y to /etc/default/grub allowed
me to proceed, at least somewhat. Now, at least, I am able to get
further in grub. Just before the grub menu, I see the first error:
error: file `/brub2/locale/en.mo.gz' not found
This flashes for a brief moment, and then I am presented with the kernel
choice menu. I see three different options there: two kernels and the
F19 rescue option. I can boot from neither kernel - I get the following:
Loading <kernel>
error: invalid magic number
loading initial ramdisk...
error: you need to load the kernel first
Press any key to continue...
And if I wait long enough it boots into the WinXP partition.
If I use the rescue option I gen what looks like a nasty dump, including
what appears to be errors in LVM loading lv_root. I do not know how to
capture that output.
The solutions I've found so far (at least for the first) seem to revolve
around accessing /boot (for example, ensuring that en.mo is in
/boot/grub2) but I've not yet found a way to get there. The solutions
to the second seem to all revolve around booting under another, working,
kernel - which I can't seem to do.
It had been operating properly before the power cut, but it had been
some time (and a number of updates) since it had been rebooted. Does
anyone have suggestions?
-Don
10 years, 3 months
Fedora 20 Live stick install fails
by Patricia Wilson
I am trying to run a Fedora 20 Live stick on my system and the
graphics is failing, probably due to my NVidia graphics card. I can
get it to boot into single user mode but then I don't know what
file(s) to modify to get into a "safe" graphics mode. I am hoping for
a pointer to more recent docs than I have been able to find. Thank you
for your assistance.
NVidia GE Force 240
ASUS 64 bit system.
--
Patricia Wilson
Apache Junction, AZ
10 years, 3 months