John's Blog

This has bugged me for a very long time. Could never get this to work - had to watch ESPN live inside a Windows VM. I don't know how universal or even how long lived this solution will be but for now it's working.  Google-Chrome Version -  Version 23.0.1271.97 OS - Fedora 17 x86_64 Flash Plugins - see below - I had two versions installed. I disabled the  11.5.31.5 version and left the  11.   2r202 version enabled and things finally started working for me. Also enabled chrome popups for espn.go.com. Unfortunately not currently working for me on Fedora 18 :( Adobe Flash Player   (2 files)   -  Version:   11.2 r202 Shockwave Flash 11.2 r202 Name: Shockwave Flash Description: Shockwave Flash 11.5 r31 Version: 11.5.31.5 Location: /opt/google/chrome/PepperFlash/libpepflashplayer.so Type: PPAPI (out-of-process) Enable MIME types: MIME type Description File extensions application/x-shock...

Making sure KVM tools not pointing to the wrong (vbox) URI I decided to try the Fedora 17 command line kvm (libvirtd) tools to create a Fedora 18 virtual machine. This was met with failure, although for some reason the GUI tool, virt-manager, worked fine for this purpose. (I still don't know why the GUI "just worked", when the command line did not - I guess it assumes the correct URI while the command line tools do not.) I was following the Fedora VM "Getting Started" guide found here  and even the simplest commands resulted in failure. # virt-install --prompt ERROR    Error in network device parameters: Virtual network 'default' does not exist: An error occurred, but the cause is unknown # virsh dumpxml default error: failed to get domain 'default' error: An error occurred, but the cause is unknown # virsh net-list --all Name                 State      Autostart ---------------------------------------...

While using Deja Dup to do a "Restore Single File" operation from within Nautilus into /etc/cups/ppd, which is owned by root, I encountered an error.  Deja Dup finds the file and tries to do the restore operation, but fails due to "cannot create directory" error. Deja Dup needs to create a temporary directory into which the restored file is temporarily placed. If Deja Dup doesn't have write permission in this directory the restore operation will therefore fail with the "cannot create directory" message. I probably could have run Deja Dup as root but maybe that's not such a good idea for one thing, and also I would have had to reenter the info regarding the access key etc. This info is already in Deja Dup if it runs in normal user mode. There also doesn't seem to be a way when running in this mode to have Deja Dup place the restored file in some other directory  (/tmp/ or /home/me/tmp) other than the canonical one (i.e. the original locati...

Help, I'm composing a message in Thunderbird, I've pasted information in the form of a table, and I can't get out! Even information that does not appear to be a "table" may in fact have the HTML attributes of a table. Once you paste that info into a Thunderbird message no amount of mousing, tabbing, or other standard UI type manipulations will get you "out" of that table and allow you to add more information to the message you are composing. Have a look at this Mozilla Bug #535475 . The workaround therein provided is surprisingly easy. Select the entire text with Menu | Edit | "Select All".  Menu | Insert | HTML to get an HTML editing window. Add a "<BR>" HTML element where it is needed (typically before or after the table). Continue creating your message having freed yourself from the table that had previously constrained you.

While using Quicken 2010 I got repeated OL-293-A errors when uploading bill pay transactions to a bank (or to a Financial Institution or "FI" as Intuit like to say). Downloading transactions worked fine as did access via the bank's website. The printed message that accompanies the OL-293-A says something to the effect of the server at the bank has issued an unexpected response - try again later. In essence, an EAGAIN. Calling the bank for support (Are they experiencing other outages? Are they in a maintenance window?) was not helpful and it didn't seem that they were aware of any systemic problem. You can look at your "conn" and "OFX" logs from inside of Quicken (I think it's Help | customer support | logs) and this showed more detail - in fact the OFX is a complete XML record of every packet sent and received, while the conn is more concise and a bit more readable. You can see other error values being returned in these logs but it all ca...

The behavior of Gnome 3 with more than one monitor is a bit counterintuitive, at least for me. Depending on the value of gsettings set org.gnome.shell.overrides workspaces-only-on-primary you get one of two behaviors. If this variable is set to " true " (default) then however many workspaces you may happen to have on your "main" monitor, you always get and see the same workspace on the external monitor. So you as you navigate up and down on the main monitor, what you see on the external monitor never changes.  So you get N + 1 workspaces, where N is the number on the main monitor and the 1 neverchanging workspace lives on the external monitor. If, on the other hand, this variable is set to " false" , than you get just as many workspaces on the external monitor as you do on your main monitor and they move in tandem or lockstep as you navigate through your dynamic workspaces. You can also obviously drag windows back and forth between workspaces. (I ...

This is really  https://bugzilla.redhat.com/show_bug.cgi?id=847479 and the bugzilla really says it all. Something broke for certain laptops around the time of Kernel 3.5.0-2.fc17.x86_64 in that laptop suspend works just fine but on resume the screen turns a lovely shade of gray and if you manage to get a TTY you are treated to a litany of messages like Aug 15 01:58:44 tp kernel: [13984.546007] [drm] nouveau 0000:01:00.0: PFIFO: channel 4 unload timeout Aug 15 01:58:55 tp kernel: [13995.124015] [drm] nouveau 0000:01:00.0: Failed to idle channel 2. Aug 15 01:58:55 tp kernel: [13995.125006] [drm] nouveau 0000:01:00.0: PFIFO: channel 2 unload timeout Aug 15 01:59:00 tp kernel: [14000.130020] [drm] nouveau 0000:01:00.0: Failed to idle channel 3. Again, as stated in the bugzilla, downgrading to an earlier kernel solves the problem. The bugzilla says that 3.4.6-2 (and presumably earlier) works fine, but when I used yum to look for earlier kernels, it was able to find 3.3.4-5.f...

Fedora 17 x86_64 - running the "Backup" utility aka Deja Dup. Set it up previously and ran at least one backup. Now a regularly scheduled backup starts and get an error message along the lines of "Computer Name Changed" "The existing backup is of a computer named foo but the current computer's name is bar. If this is unexpected you should backup to a different location". In my case the actual (incorrect) name of the "current computer" was something like 12-34-56-78.tvc-ip.com which made absolutely no sense at all.  (I made up the numbers which represent an ip address but the tvc-ip.com is what it really said). So what gives? "Backup" is really "Deja-Dup". "Deja-Dup is really a GUI front end for "Duplicity". "Duplicity" is apparently written in Python. When Python decides to figure out the current computer's name, it calls                           socket.getfqdn() So I ...

This is a complicated issue made even more complicated by the fact that although I got something "working" I don't yet fully understand all of the details. But assuming that having something working is better than not, then here goes with what I know, even if there are some holes in the understanding. Given a Fedora 17 x86_64 system, the challenge is to get Cisco Webex Meeting with Desktop Sharing working. The problem is actually quite easy if you ignore the Desktop Sharing requirement. All you really need to do is to get Java working along with the Java web plugin, which is sometimes referred to as libjavaplugin.so and sometimes as libnpjp2.so. I can recommend this page  at the "if-not-true-then-false" website. My takes on Java are I prefer the Sun/Oracle implementation to the open source OpenJDK. On the above referenced page I prefer the JDK option to the JRE option. On the above referenced page I prefer the "absolute" version to the ...

Case 1: Google Chrome Browser or Firefox Browser without the icedtea-web plugin installed. In this case the .jnlp file is downloaded but not automatically run by the browser. For some reason Firefox is icedtea-web plugin friendly but Google Chrome is not. I don't know why that is. Go to  http://docs.oracle.com/javase/tutorial/deployment/webstart/deploying.html . Click the "Launch Button" that is near the top of the page. This will cause the file "dynamictree-webstart.jnlp" to be downloaded to your machine. Type "javaws dynamictree-webstart". You should see this Case 2: Firefox Browser with the icedtea-web plugin installed.  > rpm -qa | grep icedtea icedtea-web-1.3-1.fc17.x86_64 Open Firefox and go to the same link as above:  http://docs.oracle.com/javase/tutorial/deployment/webstart/deploying.html . Click the "Launch Button" that is near the top of the page as before. This will again cause the file "...

Alternatives utility gets stuck doing install. Fedora 17 x86_64. Removing the /var/lib/alternatives/javaws file resolved the issue (for me). # alternatives --install /usr/bin/javaws javaws /usr/java/jdk1.7.0_07/jre/bin/javaws  20000 failed to read link /usr/bin/javaws: No such file or directory # ls /usr/bin/javaws ls: cannot access /usr/bin/javaws: No such file or directory # ls -al /var/lib/alternatives/javaws -rw-r--r--. 1 root root 105 Sep 22 11:39 /var/lib/alternatives/javaws # rm /var/lib/alternatives/javaws rm /var/lib/alternatives/javaws rm: remove regular file `/var/lib/alternatives/javaws'? y # alternatives --install /usr/bin/javaws javaws /usr/java/jdk1.7.0_07/jre/bin/javaws  20000 # which javaws /bin/javaws # ls /usr/bin/javaws /usr/bin/javaws # ls -al /var/lib/alternatives/javaws -rw-r--r--. 1 root root 65 Sep 22 11:50 /var/lib/alternatives/javaws # cat /var/lib/alternatives/javaws auto /usr/bin/javaws /usr/java/jdk1.7.0_0...

Fedora 17 x86_64 After a recent update got a bunch of these in a terminal session. kbuildsycoca4 running... "KConfigIni: In file /home/me/.local/share/applications/wine/Programs/Spotify.desktop, line 4: " "Invalid escape sequence "\ "."  Most of these warnings referred to .desktop files whose "Categories" line did not terminate with a ";". OK fine. Those were easily fixed. But the above message related to this line in my Spotify.desktop file, Exec=env WINEPREFIX="/home/me/.wine" wine C:\\\\windows\\\\command\\\\start.exe /Unix /home/me/.wine/dosdevices/c:/users/me/Start\ Menu/Programs/Spotify.lnk and it took some experimentation to resolve the issue.  Adding an extra back slash solved the problem. (Added back slash in red ). Exec=env WINEPREFIX="/home/me/.wine" wine C:\\\\windows\\\\command\\\\start.exe /Unix /home/me/.wine/dosdevices/c:/users/me/Start\ \ Menu/Programs/Spotify.lnk

All of a sudden the sound on my FC17 system sounded terrible. Very distorted. Long story short - I *think* that a recent update which updated kde-settings-pulseaudio for some reason caused my output volume for my volume control to go all the way to the right (I know I didn't put it there). Moving the slider back to it's normal spot of 100% reverted the sound quality back to what it was before.  Even though I don't even use KDE, I think this is the culprit because the timing was exactly right. Updated kde-settings-pulseaudio-4.8-16.fc17.noarch @anaconda-0 Update                          4.8-18.fc17.noarch @updates

I would like to type special characters - for example À la recherche du temps perdu or Du côté de chez Swann.  There are a LOT of methods to do this on Linux up to and including memorizing the unicode hex codes for each accented character and escaping these with ctrl-shift-U.  Sorry, not going to go there - that's too much like work. But many of the other methods involve using the ALT or Windows or CTRL keys as modifiers and again - sorry - I need those for all sorts of other uses. What seems like a decent solution - at least for me - is to use the Caps Lock key which up until now I had disabled on Linux. One can turn this into a "compose key" and you simply hit Caps Lock, then for example hit the back tick key, and then the "e" and you have composed è. Seems easy enough to remember, easy enough to type, and doesn't muck with my Xemacs, my Gnome Shell, my VNC etc. etc. To do this all that's needed is System Settings | Region and Language | Layo...

This apparently can happen in a number of situations when using Pidgin under Linux but for me it happens while using Pidgin integrated with "pidgin-conversation-integration@kageenshi.com" Gnome 3 shell extension. The exact sequence for me is that I get an IM Notification via the Shell (as expected) and via a series of mouse clicks in the notification area I am able to bring up the Pidgin Chat Window. However, I am unable to type into the chat window as it won't accept keyboard input. The solution is to right-click the chat window then click in "Input Methods" "System (Simple)" even though this setting already appears to be checked. This immediately fixes the issue and it stays fixed (until possibly the next  chat message is notified). Apparently other users have reported the same symptom with Pidgin but under different circumstances but for me it only occurs in the setting of the aforementioned Gnome 3 shell extension. Addendum: This is being...

Addendum: This has been "working" for a long time but ... often the printer would get "stuck". I noticed that in the CUPS interface the default paper size was set to A4. I changed that to US Letter and the reliability improved considerably. Most info I got from this Ubuntu Board - I don't have too much here that adds to that but anyways .... Get the Canon Print Driver from here:   http://software.canon-europe.com/products/0010757.asp Specify Linux, English, and choose the RPM version. This turns out to be a 32-bit driver which seems to work fine on my 64-bit system. What gets downloaded is                   MP640_rpm_driver_pack.tar So first we need to               tar xvf MP640_rpm_driver_pack.tar The tarball contains another tarball (weird but that's how they did it). So next need to               tar zxvf cnijfilter-mp640series-3.20-1-i386...

Most of this stuff is easily available on the web and easy to do. But just for fun here it is anyway. Fedora Core 16 (x86_64), Gnome 3 Shell, Cisco IPSEC VPN. Get the required software.                   sudo yum install vpnc NetworkManager-vpnc Get the .pcf file from your entity that is providing the VPN service. Get the pcf2vpnc perl script from here for example. Run the pcf2vpnc perl script against your .pcf file and p ut the result into /etc/vpnc/default.conf. Invoke vpnc - for me it's /usr/sbin/vpnc. And vpnc-disconnect does the expected. To integrate this into the Gnome Shell Go to Network Manager from top panel. Very Important - on a new setup until I added an entry via nm-connection-editor there was NO VPN section on top panel Network Manager! Run nm-connection-editor if no VPN entries are showing up in your NM panel. In the VPN section create a menu entry and make sure it's of type "Cisco Compatible VPN (vpnc)" fr...

The Problem: Allow one linux workstation to be simultaneously used for both work and non-work purposes so that work and non-work traffic are segregated. The Requirements: I would like to be able to use the following applications in the corporate environment VNC Web (typically Google Chrome but potentially other browsers as well) Email - IMAP, SMTP, LDAP using Thunderbird IM - pidgin As already mentioned here the corporate network has an SSH server that you authenticate and connect to and from there you can reach the other corporate network resources. I would like non-work related traffic to bypass the corporate network - this is mostly web traffic but might also include things like Spotify, yum updates, and other assorted tasks. Finally it would be nice (but not essential) to wrap all this up in a handy GUI operation or series of operations. The Environment:  Fedora Core 16 (x86_64) and Gnome 3. The Solution:  Use SSH, SOCK...

The title pretty much says it all. Fedora Core 16, and Gnome shell 3.2.2.1. metadata.json from weather@gnome-shell-extensions: { "uuid": "weather@gnome-shell-extensions.gnome.org", "name": "Weather indicator", "description": "Adds weather information menu", "shell-version": [ "3.1.90", "3.1.91", "3.1.92", "3.2", "3.3" ], "localedir": "/usr/share/locale", "url": "https://github.com/simon04/gnome-shell-extension-weather" } If you set via Network Manager the proxy settings to something other than "none" your weather shell extension will not be able to query the server and the extension will display that is continually reloading.  Turning off the system proxy (and restarting the gnome shell) fixes the problem. Some discussion of the issue which seems to be rooted in the gnome shell infrastructure h...