But the last 2 or 3 weeks I have seen a Auth popup asking for my twitter api key/password. And as I didn’t know what this was I did not fill in any information.

Last week I had some spare time and started to look into this.

First I was a bit worried that I had gotten hacked, or that my Twitter Account was. But after googeling abit about this it seems that twitter has changed the way the it authenticate applications wanting to connect to twitter to post or read tweets. FOr the weeks leeding up to 1/9 2010 they did limit the access to the old way to 5 an hour i think it was and after 1/9 it was totally stop for the old way. You had to start using the new QOAuth way.

So what happened was that choqok was trying to connect more than 5 times and got the reject back and aked me for a new passord.

So I was NOT hacked – hurray. That’s good at-least.

But the bad part was that I now did not have an application that worked. Well I could still connect to the twitter website and post there but that was not what I wanted.

So I started to look for solutions.

First there is a new version of choqok out – not in the repositories so I have to build it from source, but that might work. Download it. Installed cmake as it requires that and then failed on that it did require QOAuth lib and QJson libs that I did not have. OK Skip this for now.

Checked other applications. They did not work, was not available for my ubuntu/kubuntu version or was not able to compile it. Some even crashed for some reason. No luck at all.

So now I have been without a microbloggning client for a while.

But today I came across one that I havent tried before. And that was qTwitter. Its kind of small and does what I want.

I did found it here, and at the bottom of the page you have a link to the ubuntu .deb packages, including the qoath library package.

So now I can do twitter again

Downloading the game

I decided to start the thing by trying the Demo game you can donwload and try.

The Demo allows you to create accounts, play through the tutorial mission and the “Stranded in Space” mission. This gives a nice test of the game and I wanted to see that I could run the game before committing and buying the game. I did found out that you cant go to the bridge, fly into any other star system (I wanted to visit Vulcan) so the demo is quite restricted, but works well for a test.

So lets get going. Go to http://startrekonline.com/demo and clikc on the download/Cryptic to get the download program down.

The downloader is some kind of P2P program and will NOT work under wine. It runs for a while but crashes fairly often. So I did end up running the downloader in Virtualbox and there it run nicely.

Make shure you download the game to a shared drive that you can access from both Virtualbox and from linux. This is because when you have downloaded the complete installer, you wil run that from Wine to do the install.

So when the download is complete, exit the installer and Virtualbox.

Installing the game

To install the game run the download/installer from wine when the download is completed.

You need to install some extra parts for Wine before the system and installer will run rightly. If you haven’t downloaded and installed winetricks there are some instructions here.

Then install ie6 from winetricks:

sh winetricks ie6

Then its time to do the install. I ran it with the command:

env WINEPREFIX="$HOME/.wine" ^Cne explorer /desktop=3,1600x1000 PathToInstaller/STO_Demo_Installer.exe

This will start the installer, Select the path where you downloaded the files from within Virtualbox and let it verify the files. Install and register the demo account with cryptic/startrek online.

When you run the game, website you will get some javascript errors, but just click yes to continue run javascript on this page and you should get there.

Fixing the Icon

The installer will create a desktop icon if you allow it, but I changed the command to run as I wanted the game to run in a virtual desktop and be able to run the game fullscreen in this. So I changed the Application:Command to be:

env WINEPREFIX="$HOME/.wine" wine explorer /desktop=3,1600x1000 "C:\Program Files\Cryptic Studios\Star Trek Online.exe"

There are a couple of things you need to fox to get the game to run well

Fixing the sound problem

Out of the box the sound will not work correctly. At least not for me. So to fox that you need to do the following:

• Startup winecfg
• Click on the “libraries” tab
• Under where it says “New override for library” enter mmdevapi then click add
• Select it on the list, click edit, select disable, and click ok.
• Click on the main OK to save and exit winecfg

This should give you sound in game.

Fixing the rolling problem

The keyboard repeat in X is a problem for the game, so when running the game, turn that off before you run the game with (in a terminal window):

xset r off

And turn in on after the game is run with

xset r on

That will take care of that

Game setting

The only thing left now is to change the graphics options in game to “low” and set the window to run in full screen mode (remember that you are running in a virtual desktop so the size is what you specify in the /desktop part on the command line)

Done!!!!!

You can now test play the game. If you later decides to try out the full game, you only need to get the Download key to convert the install to a full game. No need to reinstall the game.

So here are the instructions:

Start by going onto facebook, login as usuall and go to http://www.facebook.com/sitetour/chat.php

There you can click on Other Windows/Mac/Linux.

And you will get a popup shoing you the login, and other information you need.

Open up Kopete and Click “Settings” in the top menu and then click “Configure”.

Click on “Accounts”, and then click “Add Account…”

Select “Jaber (XMPP, Jabber, Google talk)” from the list of services. Then click on next.

Use the Jabber ID you got from the chat page, and type it into the Jabber ID field.

Click on Remember Password and type in your normal Facebook password. Then click on next.

Under Connection preferences, leave all three options unchecked.

ANd no clikc on OK and you should be able to connect.

Be warned that you will see all you facebook friends so it might give you a lot of new persons in kopete.

But today i could not edit the tags.

I have done this a number of times before so I was a bit perplexed over why it just now stopped working.

But I found the problem. When upgrading to 9.10 aparently i got the “normal” easytag packages installaed, and to be able to tag m4a you need the easytag-aac packages.

So install it with:

sudo apt-get install easytag-aac

and all works as it should again

So Im down at the summer house, but before I left Stockholm i needed to fix remote access to my home system.

Well I have had access via ssh to my home system regardless where I am but I would prefer so really control my desktop on my home server from anywhere.

So I started looking into how to do this. Preferable secure and safe and all that.

So my home system is a kubuntu 8,04 and yes I know its kind of old but i don’t want to upgrade it yet.

So i first started out trying the built in version krfb. Well regardless of what I tried I could not get it to work. I could connect to the system but as soon as I did that it crashed. So i investigated a bit and found a number of similar post out there so its not unheard of.

So what to do?

Well the solution is to remove that and get x11vnc working instead.

So I installed this with apt-get:

sudo apt-get install x11vnc

Then logout and restart the Xserver just to make sure all things are loaded correctly. Just to be paranoid I setup a password with vncpasswd to be used so that its harder to login.

Then you can login to the system and forward the defalt vnc port (5900) to the remote machine, start x11vnc on that system and it will work.

So on your local system start the ssh:

ssh -L 5900:localhost:5900 username@10.10.10.10

When youhave logged in to your remote system, fire up x11vnc:

x11vnc -usepw -display :0

And back on you local system in another window start up vncviewer:

vncviewer localhost::5900

and you should get a copy of your remote desktop.

So I fired up Virtualbox and started my itunes in there. Then i decided to check the route to my spekaers and clicked on the pulseaudio applet. But wait its not in the taskbar.

Not to much to worry about. Chose it from the menu and start it. And it won’t start up. What the f..ck.

So I fired up a terminal windows and tried to start it from there. And the error message I got was:

** (padevchooser:9169): WARNING **: pa_browser_new() failed.

OK – Not to obvious. Google to teh rescue. And after reading 2 or 3 pages or different things I found the problem.

Its the avahi-daemon that isn’t started. And what is avahi then. Well its a deamon that helps you discover printer, people tochat with and so on, on the local network. And for some reason it wasn’t started.

So i checked the status of ith with the comand:

sudo service avahi-daemon status

And got back:

avahi-daemon stop/waiting

I then started it with:

sudo service avahi-daemon start

(If you whant to check that its running after that you can do status again and will get avahi-daemon start/running, process 9228 as a result)

And now padevchooser starts as it should.

If you still has a problem starting up avahi-daemon, you could trie to remove the status/lock file with:

sudo rm /var/run/avahi-daemon/disabled-for-unicast-local

That might help. and the do the start again.

When I upgraded to a new veriosn 9.10. I actually got a warning from phpmyadmin.

To fix the logout problem i added the following line to /etc/phpmyadmin/config.inc.php file:

$cfg['LoginCookieValidity'] = 3600 * 4; // 4 hours

This will set the valid time for login to 4 hours – that shoud be ok for me atleast. But that exseeds the max time for a cookie in php so you have to make a change to php.ini also. Change /etc/php5/apache2/php.ini with the command:

sudo vi /etc/php5/apache2/php.ini

and add/change the line

session.gc_maxlifetime

to

session.gc_maxlifetime = 14400

This will set it to teh same as the phpmyadmin cookie validity and it will now work.

After googleing something about it i found the solution. First you have to make some changes to thunderbird.

Go into the config editor:

Edit -> Preferences -> Advanced -> Config Editor

Then adding two new config strings:

network.protocol-handler.app.http
network.protocol-handler.app.https

and setting them both to the value ‘x-www-browser’.

That will tell thunderbird to use the debian alternative system for what to start when following a link. So when this done it will now open konqurer as the broswer – not really what I wanted.

No we need to fix the alternative system. This is done with the command:

sudo update-alternatives --verbose --config x-www-browser

and select chrome as you default bowser. For me that was number 2. And now thunderbird will open the links in chrome.

So after some googeling I found what was the problem. And its that vim guesses what colour schema you have and it guesses wrong.

Fix this my adding this line to you .vimrc file in you home directory.

set background=light

This will make the syntax highlighting right for viewing on a light background.

In the old system I had pulseaudio to work. I really like that as it handles multiple input and outputs and you can even adjust volumes and individual applications. So its a really god system. In the old system (8.04) it worked very well, but tended to hang some times, and I had to kill the pulseaudion daemon and restart it from time to time.

In 9.10 pulseaudion is the default use and I like that. Its very stable and works well. Unfortunately some applications is not OK with pulseaudio and there is some problems getting them to work.

Things that works well is:

• System sounds
• Skype – latest version 2.1.0.81 has pulseaudio support
• Audacity audion editor works now well with pulse
• Flash via Chrome or Firefox works ok
• Virtualbox works very well

Not working:

• Realaudio player
• IP Telephone – ekiga can work by using the headset without pulseaudio
• VMWare – but im not using sound there
• Wine out of the box

So most is working but I need to get wine working.

I hate the fact that when you go out and surf for a problem with any application and pulseaudion the top 10 things you get is “Oh i solved the problem by uninstalling pulseaudio and now everything is working”. That is not a solution for me. I want pulseaudion to be able to handle multiple streams and so on and going back to an older system is a bit backwards in my opinion. Fix the bugs and get the new things to work instead.

In the old system I had wine working with padsp – the pulseaudion OSS wrapper. And it was working well. But in this new OS its not working well at all.

After googeling and checking I found the Winepulse project. Apparently someone in Canada did make a sound backend for wine that can talk to the pulseaudio server. Unfortunately it will not be included in the normal wine as the wine people are rewriting wine to use OpenAL and that has pulseaudio build in so it will be available when that is in. When that now will happen.

So in the mean time you have to use winepulse patch or download from their archive.

So did it work – YES – perfectly and without any problem.

So what did I do.

• First i removed the old wine installation.
• Then i disabled the winehw archine from the source list
• And added Neil Wilson’s PPA: https://launchpad.net/~neil-aldur/+archive/ppa
• And installed the wine1.2 – 1.1.31.oubuntu+winepulse0.33 package

And now I happily can play my eve. Almost!!!

There is a bug in the ubuntu kernel that apparently locks up eve when you start it up. There are some fixes but I changed the start-up icon to do this instead of  anything else. What is does is change the bitsCancelled=1 to bitsCanccelled=0 in the prefs.ini file before starting up eve. And this works 95% of all cases.

env WINEPREFIX="/home/xxx/.wine" sed -i -e '/bitsCancelled/ s/1/0/'
 "/home/xxx/.wine/drive_c/Program Files/CCP/EVE/cache/prefs.ini"
&& wine explorer /desktop=0,1600x1000 "C:\Program Files\CCP\EVE\eve.exe"  /end /LUA:OFF

It shuld be all on one line. And if you have multiple start-up for different clients you need to change them to. Don’t forget to change the xxx to you home directory.

There are a few times when the startup dont work for me. By doing a wineboot -k command and then start again usually works then.

Now I can play eve as I whant.

When upgrading i lost my accelerated nvidia driver, but that was kind of expected. So how to fix it?

Well I went in to my normal download directory and tried to rebuild the nvidia drover from there. I was running 190 and nvidia is up to 195 so i did not think it was to far out. It worked kind of and I copied the xorg config file back ffrom jy old copy and tried to reboot. No luck.

I got the message that no driver was found – dam. And had go to text mode.

So I started to see if I could find the problem. I did try a number of different version, installing the latest, reverting back to and old one. No luck.

And finally i found it. There where 3 files left in /etc/modprobe.d – probably left since the upgrade but they seems to be blocking modprobe to work and being able to install the new nvidea driver. So here s the solution:

Fix for installing latest nvideadriver

Remove alas-base, blacklist-modem and lrm-video from /etc/modprobe.d:

sudo rm /etc/modprobe.d/alas-base /etc/modprobe.d/blacklist-modem /etc/modprobe.d/lrm-video

Then add the new ubuntu repository from nvidea with the latest driver. Add

ppa:nvidia-vdpau/ppa

as a new software source (in System->Setting->System Setting->Settings->Software Source), and you should be able to select the new nvidea drive (195) and install that.

This worked for me.

OBS – Do not start the upgrade before you read this part. It might fail totally.

First of all – make a backup. I was going to change my hard drive, so i used the command dd_resuce to clone my harddrive to a new drive. I pluged in my new drive as a USB2 device – yes i have a nice cable for that and issue the command:

sudo dd_rescue -v /dev/sda /dev/sdc

OBS!! Make sure you check /var/log/messages so you ARE SURE that /dev/sdc is the right device for you. If you do this to the wrong disk it will overwrite the destination disk without any problem and destroy the content.

After this was done i removed my old boot disk and installed this new one and booted up. No problem. Started the upgrade and that was when things when wrong.

2 packages that I had installed on the old system did make the upgrade fail, or atleast one of them.

The 2 packages that was the problem was flashplugin-nonfree and libgl1-mesa-dev.

So i had to go back and start all over again. The new installatin was so screwed up that you could not remove the flashplugin-nonfree, and you could not fix it so I started over, copied the disk, removed the packages and started over and now it works.

So here are the tips what to do before you start.

Before Upgrading

1. First and formost, make a backup, or several.
2. if you have installed flashplugin-nonfree packages in the old system, please remove that.

sudo apt-get remove nspluginwrapper flashplugin-nonfree

3. If you have installed libgl1-mesa-dev in the old system, please remove that.

sudo apt-get remove libgl1-mesa-dev

Then you can go ahead and start the upgrade. And you should in an hour or so have a new kde4 system up and running.

Things to go over in the following parts (And the status of it):

• Get the bluetooth mouse to work (working).
• Get the sound to work (working).
• Get the nvidia advanced driver to work (working)
• Get virtualbox to work (working)
• Get vmware to work (working)
• Get palm sync to work (working)
• Get ipod sync to work (partly working)

I will now follow up this in more parts in the comming week. Most was easy to get working even of the nvidia driver took 1 days to fix due to some remaning stuff from old installation.

I have no problems with a program that requires a specific version of another package if you have a reason for it. For instance i am using opendnssec for signing and handling keys for dnssec and it requires sqlite version 3.3.9 or newer and the reason for this is that they have found bugs in the older versions. And then I don’t have any problems with that. Its a reason and its documented.

I also had an experience with a web designer for one other of my customers this week. This is the second flavor of version fascist. They had make a new website for my customer and on that page there fas a flash part for just changing images. You can do this in many ways but flash is one way. I think its overkill but its an OK option. But they demanded the latest version (or at least a newer version of flash that i had on my machine). And this for no reason.

Definitely bad. So i hope that people starts paying a bit attention to what they are demanding for websites and for packages.

parently its in beta test face acording ot ubuntu news.

A god project i think.

And I realy love it – except fo one small thing. And its not realy the mouse problem, but the mouse driver in linux.

I use the wheel to scroll quite frequently and I also is very used to use “the third” button as paste in X. And I use left and right click at the same time to emulate the 3:rd button. And that works perfectly except for those times when I managed to click the wheel button while scrolling. This happens some times and the the driver goes

- Oh – he has a third buttong – better stop using the emulation then

And there is no way to get back to the emulation. Except for rebooting, or atleast logout, restart X server and in again.

And today my patience ran out – I did start to read about the config options in xorg.conf to see what I could do.

And i found a quick and easy way:

Emulate a left click with the middle button instead.

So here it is – add the following line to the mouse config part in xorg.conf:

Option         "ButtonMapping" "1 1 3 4 5"

That makes the second button emulate left click and the driver never sees a button 2 and never disable the emulation mode.

Simple – and easy when you know how to do it!

I also had to do some small shopping, so I went to the shoppingmall and bought what I was missing.

On my way out I thought about getting a gps navigator for the car. I want one for 1 reason – warnings for speed cameras.

So i have a few requirements:

• Speed camera warning
• Updating of speed cameras
• Road hinderence and congestion update (I dont remember the name for it) by radio
• and should work with linux.

And most points where not a big issue except the last one. None of the ones they had in the store works with linux.

And tomtom is the most interesting of them. The machine itself is running a linux system, and they have even released the os or atleast the kernel as open source, but to run the updating proram on linux – no way.

So for now its no luck for me to get one. I have to look more.

Jag hade också tänkt mej lite grillad kyckling till kvällen men den jag köpte hade datumen gått ut på så den fick jag slänga. Jag är ganska liberal med att tolka sista förbrukningsdatum utom när det gäller kyckling. Den kan man ju bli så dålig av. Så det blev en liten tripp till affären. Handlade lite grillat och på vägen ut tänkte jag att jag går förbi och kolla vilka navigatorer dom har på OnOff.

Jag vill ha en navigator av 1 anledning och det är att få varningar om fartkameror. Så givetvis ska den uppdatera detta, så mina önskemål är:

• Varning för fartkameror
• Uppdatering av fartkameror
• Gärna uppdatering via radio om hinder
• Och skall fungera med linux

Och det flesta sakerna var inga problem utom detta lilla att jag vill köra linux. Ingen av navigatorerna fungerar med Linux.

Tomtom är extra interesant för maskinen i sig själv kör ett linuxsystem, dom har till och med publicera tos’et eller kerneln som open source, men att köra uppdaterings programmen på linux – inte.

Så för nu är det inte möjligt för mej att skaffa en verkar det som. Jag får väl leta mera.

I have no problems with typing in my password when I log in to a server, but now it was getting a bit annoying.

ssh-agent

Well one of the solutions is ssh-agent. This little program sits and holds the key and send them on towards a new system. I can also forwards it if you login with ssh to another machine (if you allows it). The best way to run it is as a main program and have all other programs to be a childes to it. Then all the settings and so it inherited. And guess what – ubuntu and kubuntu does that. When you login and startup X it seems that ssh-agent is started by default.

To make it even simpler its a good thing to add you main key to it. You can do that with the command:

ssh-add

and it will ask for the key for the private keys you have in your ssh directory. An even better version is to have ssh-add run as the X system starts up. To do this you need to install the package ssh-askpass. This contains a small X11 program that will grab the keyboard and allow you to enter the key. Install it with:

sudo apt-get install ssh-askpass

Autostart of ssh-add

When we now have that installed we can setup so that we get ssh-add to run when we tart up X.

create a file .kde/Autostart/ssh-add.sh

With this content:

#!/bin/sh
# set SSH_ASKPASS if not set elsewhere
export SSH_ASKPASS=/usr/bin/ssh-askpass
ssh-add </dev/null

This will start ssh-add and popup a windows asking for the password. Set the right permissions on it with:

chmod 755 .kde/Autostart/ssh-add.sh

Ths only thing now it to forward ssh-agent requests if you whant that. I did write about my way of creating small scripts fo conecting to other servers here, and I now added a -A flag to the ssh command line to get that to work.  Then when you log in to a new server you dont have to enter the password and if you run ssh (or svn over ssh) on that machone the password questions gets back to the original ssh-agent.

So off to google a nd there is a bug and the odule for virtualbox has not compiled and been added to the repositories.

Found some sites talking about that you can download and make one yourself. And when trying that the command fails.

What i did trie was

sudo apt-get install module-assistant

sudo apt-get install virtualbox-ose-source

sudo module-assistant auto-install virtualbox-ose-source

But the build in the end failed.  I got some weird asambler error in the end. So i did some thinking and thought of getting the source out and build it manually to see what the problem is. So i did:

sudo module-assistant unpack virtualbox-ose-source

That will unpack the sourve in /usr/src/modules/virtualbox-ose. So i cd there and tred a

sudo make

to se whats going wrong. And t comples – weird. Well lets test and install it:

sudo make install

and that works to – except that i make some complains about autoconf.

So i tested and loaded the module:

sudo /etc/init.d/vboxdrv start

And it loaded the module. And now i have virtualbox working again.

What was happening (I think) is that audible has timed up with itunes for the ipod delivery and as the system is not seeing my ipod it defaults to remove the ipod from my device list after a while (some years or so). So the drm will not let me play it on my ipod. I was kind of crossed as I just bought 2 new books that was a continuation of a book i already read, and now I could not read (well listen) to them.

One way of perhaps solving this would have been to reformat and connect the ipod to a vmware instance that I have for some windows programs that I do need to run, like the bank and so, but as i have had problems with that (linux sees the ipod first and grabs it) and that I do want amarok to handle my podcast that was not real way for me. And as vmware dont talk to pulseaudio I could not play it in “windows” and capture the audio.

Some days ago I started to play with virtualbox as an other way of doing other systems that I sometimes need to run uner linux. I have some customer sometimes that requires stuff to be done on a windows system with ASP and MsSQL server and so, so I am forced to run it some times. And as  Virtualbox is free I was looking into that.

And I found that the virtualbox works better than vmware does. It handles pulseaudio as it should. So i get the sound working in the windows instance that I run under virtalbox. Well this is interesting.

This guide also requires that you have pulseaudio working on your system but as I have covered that earlier I will not go into that here.

Install virtualbox

Installing virtualbox is fairly easy. You install the virtualbox-ose version (this is the open one thats in the normal repository). Im shure you can download and install the new versions from suns site directly but as this solved my problem i just did that. You also need to install the virtualbox-ose-modules-generic package also to gte the networks and so working.

sudo apt-get install virtualbox-ose
sudo apt-get install virtualbox-ose-modules-generic

Setup a virtualbox for playing

To set up a virtualbox its a straight forward job. You just click on new in the virtualbox control program, select a name and windows xp as os type. You create a new virtual disk of at least 5 gb (5g is enough for setting up the os and itunes, and some audiobooks).

You then select NAT as a network drive

Then setup the audio as pulseaudio

And after you have installed the OS and the guest additions you can add a shared drive to store the data on

This should give you an OS that should work well and play ipod music.

Then go ahead and install the audible download manager and the itunes program.

Configure pulseaudio

Pulseaudio has many config files and many options. I am sure that someone that is very knowable in configuring a pulseaudio setup can do this better, and if you already are using pulseaudio over the network you might need to work more on this,  but this worked for me.

As pulseaudio has a option of sending audio out over the network, and in creating this you have a sink to use to send the audio from our virtualbox os out on, and record from you can use this separatly from any other sound device so that normal system sounds and so will not disturb the recording.

You go into the “Pulseaudio Preferences”. and go to tab 2.

Click the “Enable Multicast/RTP sender” and then the “Create separate audio device for Multicast/RTP”. This is all that is needed to make it work.

Install pacpl

To convert the file later on we need the pacpl program. This is a perl audio converter and it can convert from mp3 to mp4′s for the ipod. It has quite some prerequisites so the easiest way is to follow the instrucrtions here. I think im using a slightly newer version, 4.0.5 insted of 4.0.4 but the procedure is the same.

In short:

• Install/Setup the mediabuntu repostitory
• Install Ubuntu Restricted Extras and build-essential
• Download and build pacpl package
• Install it and all prerequisites

The we can use the pacpl program later to convert the file.

Install gpac

To get the best files for the ipod we need to install gpac packages also. This will give us the MP4Box program that we can use later to fix the files for the ipod.

sudo apt-get install gpac

will install this.

Install lame

To convert the file to mp3 files we also need the lame encoder. Install this with:

sudo apt-get install lame

Running the audio capture

So now its time to run the audio capture. And yes – it will require you to play the audible file in your virtalbox windows, but as this is just running I don’t see the problem with that. It just take time. So startup itunes in there, and select the file you whant to play. But don’t start playing it yet. I have created one small playlist called ToConvert that I use and add one file to that the time. This is often the case because I like the unabridged versions of the books.

Then we need to move the audio output of the virtualbox to the rtp.multicast.sink. You do this by opening up the Pulseaudio Volum Control program, rightclick on the VBOX (with the right name if you are running multiple virtual boxes), select Move stream, and move it to the rtp.multicast.sink.

Then open up a terminal windows to start the recorder. The command to capture is (type in the command but don’t press enter yet):

parec --format=s16le --device="rtp.monitor" | lame -r -x --bitwidth 16 -m s -a - > /tmp/js.mp3

We use the parec to capture from the rpt.monitor and send ot out as 16 bits little endian raw pcm file.

Then we use lame to encode it to mp3. The flags for lame is -r for raw pcm file, -x for little endian, –bitwith 16 for 16 bits file, -m s for stereo and -a for downsample to mono.

When you are ready, have the virtual box ready and the parec ready, press enter on the command line to start the recorder and then click on play in itunes to start recordning.

This will give you the file /tmp/js.mp3 with the audio you have played in itunes. You have to have the windows open and press <CTRL> + C when the playing is done. If you miss and record silence in the end, you can use audacity to trim down the file.

So now we have the mp3 file and you can play that in your ipod. But there is some problems.

1. The file will not come in right without the ID3 tag
2. If you just copy over a mp3 you can play it, but if you navigate away from the playing and play something else you will loose where you where.
3. And the file will not be under Audiobooks.

Number 1 is sorted out by using a good ID3 tag editor – i prefer the EasyTAG one.

Number 2 and 3 is handled by converting it to a m4b file. The m4b file is just an mp4 file (m4a), but renamed to m4b.

Convert to m4a

To convert the file to mp4 we will use the pacpl program. The commadn line to do the conversion is:

pacpl --defopts 0 -v -to m4a --aacqual 50 --eopts="-b 32" /tmp/js.mp3

This will set no default options (–defoptions 0), be verbose (-v), convert to m4a (-to m4a), set the m4a quality to 50 (–aacqual 50) and set the encoding options to use 32 bits bitrate (–eopts=”-b 32). This will produce a m4a files that is just lightly bigger than the audible file we started with.

Setting ID3 tags

Before one renames the m4a files to m4b i run the easyTag to set the ID3 tags. Remember to add Part1.. and to to the title if you are spliting the file in multiple parts to keep the size down (i have found that files under 100mb works best). I usually sets the tags for both the mp3 file and the m4a file.

Fixing file for ipod

There is a problem with just converting the mp3 to m4a. Some times when you start up your ipod it will fail to read in the file in memory, or miss to skip to the right place, and then it will jump to next file. Enoying. But you can make it less frequent by just waiting for 5 seconds before pressing to play. But it seems that this last step will fix this to. Apparently the itunes creates an interleaved version of the file, and MP4Box program can do this to. So run the following command on the file:

MP4Box -ipod /tmp/js.m4a

And now you are ready to rename the file:

mv /tmp/js.m4a /tmp/js.m4b

And you can now tranfer the file to your ipod with amarol, and listen to it.

Hope you enjoy this as much as i do.

Men nu idag fick jag nya dosan och inloggningskortet från dom. Och om man vill använda USB och sladd så måste man installera programvara och drivrutiner för det, men jag brydde mej inte ens om att se om det kan fungera under linux utann kör den utan sladd. Det innebär att jag måste knappa lite mer men nu slipper jag att köra XP bara för den saken.

Hurra.

The old one we had at the summerhouse blow up last year – last weeks of the authum we where there and I did not have time then to fix it so I just took it home. After checking it now I did found out that the motherboard is probably fried and not the harddrive as I suspected, but I will have to do more investigation into this. I had more small boxes that can be used for this (the one we had in the house in uppsala could be reused) so i installed a new version of smoothwall on it. The old one was a Smoothwall Express 2 and there is a new version 3 out that I wanted to test.

So I installed the new version on the box and as I whanted to test it as much as I could I plugged the red network into my home network and that was good and well and it got a network adress over DHCP just as it would down at the summerhouse when I plug it into the telia ADSL connection.

Except for the fact that after the initial config you are suposed to login to the webinterface from the INSIDE (or the green network) to finnish the config. And my machines are ofcuase on my green home network, and that is now the red on the new smoothwall. So what to do. Well im shure there is a way of changing the config manually so I started too look around and checked if I could change it from the command line.

And there is a way but I could not find any one mention this so thats why I write this down. Hopefully someone else can use this aswell.

So you go into the machine (on the console) and login with root and the password you set.

At the prompt you do:

vi /var/smoothwall/xtaccess/config

And then add the line:

tcp,0.0.0.0/0,441,on

Save the file, back to the command line and reboot the system (reboot command).

This will open up the access to the 441 port and the web interface from the red network and after this you can login on the red network to the webinterface also and continue the config process.

What to do.

Well fist i did some googeling and found a god application for this. Its called PDFEdit and works perfectly.

So I installed it with:

sudo apt-get install pdfedit

And I had no problems opening the PDF – delete the 1 page and save it again. Wounderfull.

But he fixed that, wrote a small stub function that only returns OK, and that is now included in the 1.22 version so I would liked to have that going.

So i let the upgrade program do it works and had wine 1.22 running in no time.

Started up eve and “Oh No” – when I was inside the station i only had a black screen – no interior at all. So out in space and my show is permanently cloaked – even for me

So reading a bit in the forum it seems that this is a bug in the ATI driver and there is a patch for it, and apparently it should work on Nvidia cards. It might be the fact that i am running a bit old version of the Nvidia drivers. I am running the 169.12 version and the new version that is out is 180.51 so perhaps it is time to think of upgrading that, but as i am about to upgrade the whole system (I don’t like it and don’t have the time, but have to eventually) to the new version 9.04, I think I will do it then.

So now I need to downgrade to version 1.21 to have my game going as I whant it. I can live with the fact that I have to download the patches manually.

Downgrading wine

First i need to download the old version. You just go into the link on the wne HQ page and go to the Ubuntu pages and the old archives (here) and download the old version (1.21).

Then you install that with dpkg:

sudo dpkg --install wine_1.1.21~winehq0~ubuntu~8.04-0ubuntu1_amd64.deb

Next thing you need to do is to lock the version. if you dont do that the system will upgrade wine next time you run the upgrade program. And you do that with:

echo wine hold | sudo dpkg --set-selections

And if you as me have Synaptic package manager installed, you can go in there and lock the version also.

This now get me a god working wine that pays my eve, and I have to live with the not working autopatcher for now.

And yesterday was the date. I am a bit hopefull that this will solve some of my problems. I have a number of small anying things that I can get around but makes life a bit hard and I hope that the new version will solve some of it.

The problems I’m having is:

1. Amarok cant handle my audiblebooks any more
2. My palm don’t sync as it should
3. Some problems with realaudio and pulseaudio

Its not a major thing but its a bit annoying.

So I will do the update but not just yet on my laptop. Its to much hassle to do in the middle of everything but I need to plan it.

Torrent seeding

There has been quite some hassle in the press about Pirate bay and the fact that downloading has gone down quite a lot in sweden. So to help that I have now decided to download all the CD*s and DVD’s for the new kubuntu release and seed them over bit torrent. This is perfectly legal as its a non copyrighted materia and I think that it will be hard to see that torrent as right and other torrent as wrong.

Lets all start sharing real materials over torrent and se how the anti pirate bay maffia can track this.

Happy sharing for now.

So what to do. Well there are special panoramic cameras but they are very expensive. The second options is to use programs to take pictures with overlapping areas and mount them together as one picture. This is called stitching.

I was doing that quite a lot when I was running windows but I havent had time to do it under Linux. There was some god programs for it under windows but now its time to se what I can find under Linux.

After some searching on the net, i did come acreoss Hugin. This is a descendant of PanoTools, an excellent suite of panorama processing applications first developed by German university professor Helmut Dersch. The original PanoTools are no longer being developed, so Hugin is what I started to test.

Installing

Installing Hugin was easy – its in the kubuntu repositories so I only had to do an apt-get install to get it installed:

sudo apt-get install hugin

Testing out

So the program is installed and the default place is under graphics. Staring it up gives you the assistance page. First klick on the Load image button.

I tested with 2 of the images from my latest vacation:

You can select multiple images with CTRL. When this is down it starts the scanning and processing of the images. It takes some time and when its done it will present the panorma preview. If you are happy you can close that down.

I then went into the sticher tab and changed the output format to jpeg. Then you can generate the images with stitch now, or go back to the assistance tab and click on create panorama button. This will give you a nice panorama:

First you see tha the images is not blended. This is because the stitching engine can’t do that, I found out, when doing jpegs, so i did a new one and let it save it as tiff and then i converted outside of Hugin to jpeg and got this result:

Much nicer blending but the pictures are a bit off in balance/contract and so, so you need to do some processing first. And when that’s done i also cut down the image and removed all the black background. Here is the final result:

Conclusion

I think that this show that most image processing and specificaly creating panoramas can be done in Linux without any problem.

So i investigared the system a bit and there is a config flags for this.

Add the following line to your config.inc.php file:

$cfg['LoginCookieValidity'] = 3600 * 4; // 4 hours

This will set the valid time for login to 4 hours – that shoud be ok for me atleast.

So first of all i start installing the updates and the kernel. I’m a bit chicken and not going to install wine yet.

The upgrade works as it should. Still no built in support for the network cards so I have to compile and install that module again. Not a big deal.

And then i have to install/configure vmware modules also. Still not a big problem. And now all things is working as it should. Sort of.

Testing palm

Next step is to se if the new kernel has solved the problem with my palm – and guess what – it has not. Dam. But not a big deal as I can use the palm xfer programs to get my work done and its working – even if its a bit cumbersome.

So onto the big issue – Wine

Installing wine 1.1.12

When wine 1.1.10 came out I started to get some problem. The EVE online game had problems when starting or when warping into some new area. I could get it to work better is I remembers to not do anything, not move the mouse, pan or zoom or anything when warping in but its hard to remember this all time.

So I backed down to the 1.1.9 version that worked perfectly for me. But its kind of annoying to always have that reminder that there is new things to install all the time.

So this week the 1.1.12 version came out so I decided to test that.

Removed the lock on the install and let the system upgare wine. And the the big test – startup eve and run and test it.

And I can gladly now report that it works like a charm.

And had quite a bit of a hard time getting this going.

The fist try was to just use amarok program and play the CD. No problem. Then i went around the menus to see if one could import the files directly there but no sucess. There seems not to be any way to do it directly in amarok.

So the next try is to use the k3b and the item directly in the CD desktop shortcut. Starts up k3b and goes into the ripping. It fetches the album info from the net and all seems ok. So lets go.

Well its starts and pops up a window:

OK. Already in use – as an old unix guru it looked like l should quit the other application and get on with it. And that’s what i did – clicking on “Quit the other application” and after some confirmation it started up.

Run for 2 seconds – cd starts spinning, stops spinning again. Repeat. And it takes forever. This is not working. I had a similar kind of problem in the summer when i ripped another CD and I cant for my life remember how I solved it then.

What to do now. Well i started googeling to find some answers.

ripperx

After searching the net for some time I did found one program that looked promising. It was called ripperx and is using the lame encoder to get the files to mp3. So I downloaded it and installed it with apt-get:

sudo apt-get install ripperx

Test run and after some settings it seemed to do the work.

So i ripped the cd and got it into my amarok collection.

back to k3b

Well I think it was strange that k3b did not work so i thought that i should trie it one more time if it was just a coincident that it did not work.

So I tried again. Inserted the CD. Selected rip with k3b. Got the warning. Closed other applications and tried and the same result. Click, Click and no result.

Ok. This is weird. Just from a hint I did try again and disregarded the fairly strong wordings in the message  box and just let k3b go on with it and it did work. Very weird. But god.

So I think i will be using k3b now for the ripping. There are some settings that one needs to do to get it to work the way I whant it to work.

Configure k3b

So k3b can rip the cd’s for me, but I do need to configure it right.

First of all I have to tell it to store the mp3 files in the right format and the right names and so on.

You do this in the rip windows after you selected “Start ripping” from the main k3b window.

I have all my music stored in /CD/MyItunesMusic so I set that in the target. Then you have to get the filenames right. Click on File Names tab and you get this

And its the ripped file name you need to change. I set it to:

%{albumartist}/%{albumtitle}/%{number} %{title}

to correspond with what I have in my music album before.

The next thing to do is to configure the lame encoder to use 192 kb  bit rate instead of the default 128.

Go back to the fist tab and select the Mp3 (lame) file type and then press edit (the small button n the right if the dropdown).

I Changed the name to Mp3 (lame) 192 kbps and then inserted the -b 192 to the command line.

Click on ok and DONT FORGET now to save this. Its the litte save button on the bottom left of the windows. No you can go ahead and rip the CD’s and it will remember this setting.

Amarok

Next on the list is to get amarok to read in the new files names. You do this with the Tools->Rescan collection meny item and there you have your new mp3′s.

Syncing to home server

The last thing is to get the CD’s over to my main home server with all th CD’s for my Squeese Center to play on my stereo.

rsync does the job there and the command is:

rsync -avzuc -e ssh /CD/MyItunesMusic/ jan@192.168.0.10:/CD/MyItunesMusic

Thats all – i can now play my cd’s as i whant. Perfect

I’m using the Trip Deux program and have been using it for many years now. Some time ago i found out that the easiest way to get this to work under Linux is to export the month driving entries to a csv file in the memopad and get that into excel / well open office. But now I have a problem as my sync is not fully working.

So I did some more invenstigating in the pilot-link package and there is a program for manipulation the memopads exntries. So im running this command:

cd ~
pilot-memos -p usb: -s tmp/palm_memos

This will get all the entries out of the palmpilot and I can then get the memo i whant from the Unfiles categorei/directory.

The only small thing is to edit the datefield as it will inport wrong if not having a full year (2008) instead of 08.

I have been working over the internet for quite some time now and usually in a day i have over 8 different terminals open to different serves that I’m monitoring, updating, checking, maintaining, or all kind of other stuff. And I need to keep track about what windows is what. I usually uses the konsole application and open all different sessions in different tabs, and sometimes i have more than one windows so its a lot to keep track of.

I have previous used the old “Prompt” trick and changing the shell prompt for different machines, but that was not enogth. For some time i have been changing the tab name but doing this manuallay every time is a bit tideous. I have looked at different ways of chaning the tab name and there are eacape sequences that does it but I have not had a god result.

Some time ago I ceated a number of small scripts for connecting to different servers. One of the reason to do this is that I am using phpmyadmin to connecto to databases that I maintain and work on, so to avoid to keep typing all that I used the small script as i talked about here.

So why not change it to also set the Tab name.

And here is when DCOP comes to help. DCOP is KDE’s IPC/RPC mechanism. In other words is used to communicate between different DCOP aware applications on the KDE desktop. Its quite like the Microsoft Automation system, but it has no means of starting applications or services only runnning applications can be controlled.

The ‘dcop’ program provides command line access to the DCOP system. The syntax is:

dcop [application] [object] [function] [arguments ...]

To see which applications are available, enter ‘dcop’ with no arguments.

The interesting thing here is that konsole has some integration into this and there are 2 environment variables, KONSOLE_DCOP and KONSOLE_DCOP_SESSION.

So if you in a konsole window run the command:

dcop $KONSOLE_DCOP_SESSION sessionName

It will tell you the name of the sessson that is shown on the tab.

if you just run dcop $KONSOLE_DCOP_SESSION it will show all the tings you can get and you can set (and yes there is a nice graphocal tool also – kdcop – but thats a separate thing). And the interesting here is teh line

void renameSession(QString name)

This means that you can set the title to anyting you whant. So after realising this I cerated a scipt that i now use for all the different system i regularly connect to.

#!/bin/bash
#Start ssh to foobar

CMD="ssh jan@foobar.foo.bar"
NEW_TITLE="foobar"

DCOP=/usr/bin/dcop

echo $CMD

function restore_title() {
        $DCOP "$KONSOLE_DCOP_SESSION" renameSession "$OLD_TITLE"
}

if [ ! -z "$KONSOLE_DCOP_SESSION" ]
then
        # Get old title for restore
        OLD_TITLE=`dcop "$KONSOLE_DCOP_SESSION" sessionName`
        $DCOP "$KONSOLE_DCOP_SESSION" renameSession "$NEW_TITLE"

        # If SSH is interrupted (CTRL-C), restore the old title
        trap "restore_title" SIGINT
        $CMD
        restore_title

else
        $CMD
fi

This alows me to set a short name for the conneciton.

Today i was going to sync my palmpilot and I did as I usually did. Put the palm in the cradel. Press the sync and it all started. Wait – it did not. Noting happend for a long time and then i finally got the errro message on the palm that it could not connect to the computer.

I did some investigation and found this on the web. Apparently there is a bug in the 2.6.24 kernel that prevents the visor module to work. And as its not working the usb devices is not created and the kpilot will not be able to sync.

I did some test with different settings and could not get this to work in any way.

So i did unload the visor drive and uses the pilot-xfer program directly to get a backup atleast so i can do my timesheet.

cd ~
pilot-xfer -p usb: -b tmp/palm_backup

This will start pilot-xfer and connect it to the usb. When you then press the sync button it starts the sync and get all the apps and dbs over. I can then use teh tmp/palm_backup directory to get some of the db files into my timesheet progtam

So there are thing to fix and get going before its all ok.

Firefox and empty application list

So apparently after some googeling there is one package that’s not getting installed when installing firefox on a Kubuntu system. Its the firefox-gnome-support package.

So I install that with:

sudo apt-get install firefox-gnome-support

And that solves the problem with the application list and its now filled up with the applications to open for different file types.

Install Realplayer

So next in the list is to install RealPlayer.

There are some packages you can find for the RealPlayer out there but its not hard to download the player from Realplayer site directly.

I did the following.

• Create a temporary directory – if you dont have one already
• Go to teh directory you just created
• Download the bin file with wget
• Change permissions on the file
• and run it to install all the parts

So open a terminal window and do the following:

mkdir ~/tmp
cd ~/tmp
wget http://www.real.com/realcom/R?href=http://forms.real.com/real/player/download.html?f=unix/RealPlayer11GOLD.bin
chmod 755 RealPlayer11GOLD.bin
sudo ./RealPlayer11GOLD.bin

For all the question you can use the default. It will install itself and install a link in the multimedia sub menu.

Get RealPlayer to work with pulseaudio

Unfortunately realplayer is one of these other applications that don’t work with pulseaudio, and trying it to run to alsa will even crash it without any error messages on my system. But there are the god news that padsp program works with it. So running

padsp realplay

starts up realplayer with the emulates oss driver and you can the go in and select oss as the sound driver under tools->preferences->hardware and it will play perfectly under pulseaudio.

The only thing left now it to change the menu and file association for the realplayer files to play with padsp realplay.

Go in and click on the kde menu, and the click on the multimedia sub menu and the right click on the realplayer menu item and select Edit Item from the menu. In the box that comes up change the command from realplay %U to padsp realplay %U. This will also change the file association for realplayer audio files so when clicking on a file it will startup padsp first and have realplayer working on that virtual device. Don’t forget to save the setting when you are done.

Realplayer and firefox

Last in the list of things to fix is fixing so that you can click and listen to a file directly from firefox. Unfortunately, no matter of how I trie to fix this it will not work. So the best solution for this is to get into prefrences->applications and set the filetype realaudio to save file. Then when you click on the link – save the file somewhere (in /tmp is what i usually do and then manually open it with padsp realplay). Im kind of sorry and anoyed that I could not get this to work but I guess its a combination from both the 64 bits version of firefox and that the realplayer is a 32 bits and the fact that realplayer also dont work with the pulseaudio. I hate being beeten but as I can workaround it its end of this for now.

So i downloaded and installed the asterisk package with:

sudo apt-get install asterisk asterisk-dev asterisk-doc

This was to get the development and the documentation also. After installing this i have asterisk working. I now connected my Grandsteram GXP2000 sip phone and  can hear the test and do echo and so.

So the next part is to get some of my special sounds and getting the conference services to work. On of the interesting things with the default ubuntu or kubuntu install of asterisk is that the sound is moved to /usr/share/asterisk instead of /var/lib/asterisk (well that is there to) but all the sound files are in /usr/share/asterisk/sounds/.

So i installed the sound files and changed the extensions.conf to include my meetme conferences. And its not working. Hum. When accessing the meetme application if get the error message that asterisk cant access the timing device. Yes – I forgot to install zaptel and the ztdummy device. So checking and ofcourse there is a package with the zaptel devices so I installed it with:

sudo apt-get install zaptel

And got errors – the installer cant find any zaptel devices and that absolutely true as I dont have any in this machine. And it will not install the ztdummy device. Dam. So i need to build it probably. I can download the source with apt-get source zaptel but i hope for someting else if i could.

The I read about the module-asistance comand in ubuntu. This is actually to start with a debian comand. Remember that ubuntu and kybntu comes from the debian system and a lot of the tings in the system id debian based. Module assistant is a script that helps download and build kernel modules from source. It has a lot of different commands but if you whant to make it all you can use the auto-install command that will just do all the job. There is also a short name for module assistant – m-a that you can use if you dont what to type that much, and even a short cut for auto-install – use a-i instead.

If you dont have module assistant installed you do that with:

sudo apt-get install module-assistant

In my case it was already installed, but if you dont have it installed, install it and ofcourse install all the dependencies also. So I went ahead and installed the zaptel vith this:

sudo m-a a-i zaptel

This did the job and installed the zaptel driver for me – and now i have ztdummy running and I can run the meetme application.

If you do need to install the package again you find it under /usr/src, and you can use m-a install to install it also.

So I downloaded it and have a nice mp3 set of the books – 70 files of it. Not ideal. I could ofcourse transfer them and make a playlist, but thats not ideal. One can use audacity to merge all the files to one big mp3. That would work to. But there are a problem with that to.

All my normal audiobooks is under the audiobook menu in the ipod. When transferring a audible book (.aa) file it pops up there. And transferring a .mp3 file it will go to the normal list of music. A mp3 will also start to play from the beginning when you go back to it. For podcast it remembers the place where you left of from playing and the same goes for an audiobook.

Checking a bit I found out that to have the ipod recognize the file as an audiobook as has to be an AAC encoded mp4 file (.m4a) file and that the filename has to be .m4b.

AAC

AAC or Advanced Audio Coding, is a MPEG (Motion Pictures Experts Group) audio standard first adopted as part of the MPEG-2 family of standards. The best known example of this is Apple’s iTunes online store and software. iTunes AAC files are stored in the standard MP4 container, but generally with an extention of .M4A to denote that they only contain audio.

The audiobooks is a normal m4a file but just renamed to m4b. So we need to convert the file. This can be done with mpg123 and faac. So we download and install them.

sudo apt-get install mpg123 faac

Now we can use mpg123 to play the files and pipe it to faac to convert it. This files I have for the ringworld book is normal mp3 files – MPEG 1.0 layer III, 128 kbits/s, 44100 Hz joint-stereo files. So we can convert them with the standard commands:

mpg123 -s * | faac -b 80 -P -X -w -o output.m4a -

This works nicely and create a big file (460mb), bu in all workable on the ipod. So i started to do the same for the Ringwold Engeneer – book 2 in the Ringworld series and it is playing like on 4 times the speed. Annoying. It seems that the files are 22khx mono files and that faac or mpg123 is not making the conversation right.

So i found another tool  – MP3 to iPod Audio Book Converter – and its free but runs under windows – well actually if you look under the hood and gt the source you can se that it seems to be a java program, but I downloaded the installer from here, and tested it under wine and it works perfectly. It creates m4b files. It handles my 22 khz mono files right. It is graphical – kind of OK – but not really needed, and it has a built in ID tag editor. So thats what i’m using now.

When you run the instaler it installs it and created a meny item in Wine -> Programs -> MP3 to iPod Audio Book Converter ->  MP3 to iPod Audio Book Converter and you can run it from there.

Ipod and homemade m4b

There are problems with homemade m4b files, and it seems that when the ipod goes into sleep mode and you press play again to wake it up and press play again to start the load of the 460 mb file takes to long time and it fails and goes back to the main menu. It seems like its timing out.

I have noticed that all the audible files is max 100 mb large and if the files gets bigger they divide it into multiple parts. Can that be a trick to use. I reencoded my ringworld file into 4 parts and now its seems to play ok.

You can ofcouse use the command line for a lot of things and many times I do use that. But some times a graphical tool is god to use, especially when you are converting old applications and need to jump between tables and getting an overview of a system.

When I was using my old machine under windows there was a program called EMS SQL Manager (comercial) so you had to by it but it was very good. So i was looking for a simmelar program under linux and came across phpmyadmin ofcouse.

This is a webapplications written in PHP and you can install it on a server or on your own system. I would prefer to have it running on my local server / laptop and access databases remote as I don’t want to install web and software in all serves im working on. So I started with installing phpmyadmin on my laptop. There is an package that is called phpmyadmin os its just to install that with:

sudo apt-get install phpmyadmin

I can then go to locahost/phpmyadmin and login and work on the local databases.

Access other servers

Thats all god and fine but now i want to access other servers. There is 2 different ways of doing this. First you can enable the “Any Server” field so that anyone can enter in an address and that phpmyadmin will connect to that database. Its something I kind off need as sometimes you dont want to add the server to the phpmyadmin config file, but just access it a few time.

It is ofcouse a security problem as anyone with web access to your machine can use it to connect to any other mysql sever, but as most of the time my laptop is behind a firewall with very restricted access to it, I can cope.

What you do to enable this is to add one line to the config file (/etc/phpmyadmin/config.inc.php):

$cfg['AllowArbitraryServer'] = TRUE;

This will pop up one more input field on the login page called Server and that will allow you to enter the server you whnat to connect to. You can actually enter both the IP or hostname of the server and the port name (after a : ) if you are running mysql on a special port.

And this leads into ssh tunnelling.

Mysql ssh tunneling

One other way to get to a mysql database even through firewalls and other hindrance is to use ssh tunnelling.

You do this with the -L flag to ssh.

So as en example:

ssh -L 7778:localhost:3306 jan@www.foo.net

will take port 7778 on my local machine and forward that to port 3306 on the other end (localhost on that end). And that is ofcuase the mysql port. So if you can log in to a server ytou can forward a port that way.

This is ofcuase a great way of working with severs and mysql servers, and you can then type in 127.0.0.1:7778 in the server field on the login page of phpmyadmin and work with the remote database.

If you have one or more servers that you regularly are working on the is a way of makeing this even easier.

I have created on small shellscript in the bin directory in my home directory. Its has the same name as the server – lets say foo and consists of:

#!/bin/sh
#Start ssh to foo with forwarding of mysql port
echo "ssh -L 7778:localhost:3306 jan@www.foo.net"
ssh -L 7778:localhost:3306 jan@www.foo.net

This way i only have to type foo to access the server and the right ports get forwarded. You could do the same with the config file in ~ssh also.

The you can add the folloing lines to the config file (/etc/phpmyadmin/config.inc.php):

//-------------------------------------------------
$i++;
$cfg['Servers'][$i]['host']          = '127.0.0.1';
$cfg['Servers'][$i]['port']          = '7779';
$cfg['Servers'][$i]['socket']        = '';
$cfg['Servers'][$i]['connect_type']  = 'tcp';
$cfg['Servers'][$i]['extension']     = 'mysql';
$cfg['Servers'][$i]['compress']      = FALSE;
$cfg['Servers'][$i]['controluser']   = '';
$cfg['Servers'][$i]['controlpass']   = '';
$cfg['Servers'][$i]['auth_type']     = 'cookie';
$cfg['Servers'][$i]['user']          = '';
$cfg['Servers'][$i]['password']      = '';
$cfg['Servers'][$i]['only_db']       = '';
$cfg['Servers'][$i]['verbose']       = 'foo.net';

This will give you one more server named foo.net that can be accessed in the drop down Server Choice box.

This is how I solved my need. Hope it will help someone else.

Summer time and reading is a god combination. As some of you know that knows me I do a lot of reading but I do most of it as Audio Books. I have for many years done quite some travelling and sitting and driving for a long time requires something to do. And I listen to Audio books. And preferable Science Fiction. I also listen to quite a lot of Pod Cast, both in general interest like the Weekly Science News and some for more specific interest, like the EVE pod cast for the EVE-Online game.

And I have already talked erlier about my set-up of Amarok so that it would handle my ipod and download my music and pod cast to that. No Problems.

To get my Audio books I have mainly used the Audible Site. This is quite a god site for audio books. Its ofcuase commercial and you have to pay for the books, but its kind of OK. I have been a premium member for 4 years now or so and it works very well. Except for one small think. It only Windows or Mac oriented.

The Audiobooks are encrypted and locked down to you devices – don’t ask me how but the files is a .aa file and you have to have special player for that. And they exists in the ipod, and on mac or windows. So you cant play the files on a Linux machine. That is not a big deal for me as I’m only transferring them to my ipod and listen to them there. On my old machine I used the VMware installation I had, and had a Itune setup on it so that it and the audible download helper downloaded it. And then I transferred it to my ipod using the files that itunes created and had Amarok to download it to the ipod. Worked kind of ok even it was an extra step to starup vmware as soon as you whanted to download you books. Well it did not happen that often, so it was ok but would have been nice to no have to do that all the time.

But now when i have moved over to the new machine itunes will not import the file.It complains about write permissions to my music directory and as far as I can see it has all the rights its should have. I was fiddling a bit with the download and found out that:

1. I could move the download directory of the Virtual harddrive and to my shared device – the download helper had no problems with writing on that disk.
2. I could also get that file into Amarok and transfer it to the ipod, so it works again.

Wine to the rescue

After thinking a bit and wondering a bit about it, I did some testing with Firefox 3 and Linux. When audible send a download, it will send a small .pl file. On windows the small AudibleDownloadHelper.exe program will then start and download the file and I guess decrypt it or something. So I looked into my vmware installation and saw that there is a directory where the audile things are installed. C:\Program\Audible and it has a Bin directory with 2 .exe files, 2 .dll files and .AUD file. SO i copied that over to my home diretory.

I then run the AudibleDownloadHelper.exe manually (with wine AudibleDownloadHelper.exe) and it popped up the config screen for the program and I could set where I wanted the downloaded files to be stored.

I then run it again with the .pl file as the argument and it actualy downloads the thing and stores it where i whanted. Horay.

So the last thing to test is to set the AudibleDownloadHelper.exe as the program to handle .pl files and it will start the program automatically. This works perfectly and I can now download my Audible books without even having to start vmware.

When reading about pulseaudio it seems to promise to solve all the problems, multiple applications and even individual volume controls for each application, so it seems promising. The only problem is that kde is not handling pulseaudio direct so there might be some work necessary to get the kde notifications going.

First thing ofcuase was to get the sound to work at all. Without ALSA working i dont think this would have succeeded at all. I did found 2 notes/links to information about how to do this so i followed that and added some things of my own.

Install pulseaudio

Started of with backing up important configuration files (don’t worry if you see “no such file or directory” warnings):

mkdir ~/pulse-backup; sudo mv ~/.asoundrc* /etc/asound.conf ~/pulse-backup/
cp /etc/pulse/daemon.conf /etc/pulse/default.pa ~/.pulse/

Then we install some packages to ensure that all the pulsaudio will work right. I also added the pulseaudio package to this (all on on line):

sudo apt-get install libasound2 libasound2-plugins libasound2-dev padevchooser swh-plugins
libao-pulse libsdl1.2debian-pulseaudio ladspa-sdk pulseaudio

Create a new “~/.libao” file and set the default driver to “pulse” for libao applications:

echo "default_driver=pulse" >~/.libao

Create a new “~/.asoundrc” file:

vi ~/.asoundrc

And add the following to the file:

pcm.!default {
  type pulse
}
ctl.!default {
  type pulse
}
pcm.pulse {
  type pulse
}
ctl.pulse {
  type pulse
}

This should have set everything up and so. The only thing left is to go into sound and hardware and set the sound system up to ALSA.

And still no sound.

So reading some more and found another guide, so I followed that to. This will probably be some overlaps but lets try it also.

Install additional packages:

sudo apt-get install libao-pulse libasound2-plugins

Edit /etc/asound.conf and set the following content:

pcm.pulse {
    type pulse
}
ctl.pulse {
    type pulse
}
pcm.!default {
    type pulse
}
ctl.!default {
    type pulse
}

Edit  /etc/libao.conf and have the content sat to:

default_driver=pulse

If the files dont exists, create them..

Install more packages and packaged for the default libSDL:

sudo apt-get install libflashsupport padevchooser pulseaudio-module-hal pulseaudio-module-x11
sudo apt-get install libsdl1.2debian libsdl1.2debian-alsa

The last install will warn about conflicting installs but that as it should be. Allow it to install.

Now add the following lines to /etc/apt/sources.list (this will add a new repository, from zman0900 so that we can install the patched SDL driver):

## zman0900's PPA
deb http://ppa.launchpad.net/zman0900/ubuntu hardy main
deb-src http://ppa.launchpad.net/zman0900/ubuntu hardy main

Then run the following to get the patched DSL driver installed:

sudo apt-get update
sudo apt-get upgrade
sudo apt-get install libsdl1.2debian-pulseaudio

You will get a warning about installing from an unknown source but I went ahead and did it anyway.

Last thing to do now is setup one environment variable for the patched SDL driver and you do that by editing your .profile and add the following 2 lines to the end of the file:

# Make SDL audio work properly with Pulse
export SDL_AUDIODRIVER=pulse

This should be it and we can now reboot the machine and see if it works. And guess what, It did not. Hum more googeling and a bit of searching and I found out that for some reason the pulseaudion daemon did not start up as it should. First I tried to start it up manually with:

pulseaudio -D

And now the pulseaudio things works. Well kind of. I still don’t have any notifications from KDE but all other applications i start works so this is kind of on its way at least.

Configure kde

So KDE is not playing along so far. More sufring and checking for answers. Apparently KDE (and kde3 as I’m using) is using the artsd to play sounds. This daemon is not ready to work with pulseaudion, but it can be configured to work to esd (Enlightened Sound Daemon) and that one is OK with pulseaudio. So i reconfigured sound hardware and wolla I got it to work.

Sound from kopete when someone pings me, at the same time as i play music via amarok. Nice.

If you whant to do this you can edit the .kde/share/config/kcmartsrc

[Arts]
AddOptions=
Arguments=\s-F 10 -S 4096 -a esd -n -s 1 -m artsmessage -c drkonqi -l 3 -f
AudioIO=esd
AutoSuspend=true
Bits=0
DeviceName=
FullDuplex=false
Latency=250
NetworkTransparent=true
SamplingRate=0
StartRealtime=true
StartServer=true
SuspendTime=1

No its the only thing lft to test and thats to reboot the system and se that it works after that. So I reboot again and no sound – dam.

More checks and the sound daemon is not running. One suggestion is to make a small Autostart file in .kde/Autostart and that did start-up the pulseaudio daemon but there is still problem with the kde notifications and I think its due to the fact that Autostart starts up the pulseaudio deamon long after the time when the artsd has started and tried to access the sound already. Have to find a way to start the pulseaudio daemon earlier. After reading some manuals i found the way and its to configure Xsession (the program that normally start up Xsession in the X11 world). It has a number of config files and you can add new things in /etc/X11Xsession.d. So I created a new file /etc/X11/Xsession.d/99-pulseaudio withthe content:

#!/bin/bash
pulseaudio -D

And tested and rebooted the system (after removing the Autostart file first) and now the sound works perfectly as it should.

I did have a bit of a problem getting skype to work so you need to configure the daemon to get away with some of the stuttering. Se next part under skype for that config also.

Applications

So now the basics things in the system will work as it should and I can even adjust individual application sounds levels with the pulsaudio volume control – very nice. But not all applications are that well-behaved.

amarok

Amarok is using xine as a backend to play sounds and its fully integrated into the pulseaudio system so thats just works out of the box. Very nice.

wine

When reding the notes about wine it seems to be a problem but it also said that the latest version should work with padsp that emulates the oss layer onto pulseaudio.

So I started up winecfg and selected oss as the sound device. I then started up my EVE client with padsp:

env WINEPREFIX="/home/jan/.wine" padsp wine "C:\Program Files\CCP\EVE\eve.exe" /end /LUA:OFF

And it works prefectly. I changed my desktop icons to do the same and this is now working.

skype

Skype has some problems with pulseaudio it seems. There are some cases of stuttering an you can get that out of the way by changing the priority of the pulseaudio daemon.

Add your user to the groups “pulse-access” and “pulse-rt”:

sudo adduser $USER pulse-access
sudo adduser $USER pulse-rt

Edit ~/.pulse/daemon.conf:

vi ~/.pulse/daemon.conf

Add the following lines to the end of the file, and save:

high-priority = yes
nice-level = -11
realtime-scheduling = yes
realtime-priority = 5
default-fragments = 8
default-fragment-size-msec = 5
resample-method = speex-float-0

As I’m using my USB headset for sound for my phone, I could set the output and input of the skype client to the USB plughw device, and set the ring device to pulse. This made my skype work.

ekiga

ekiga and my other softphone (sjphone) has a bit of a problem with pulseaudio, but the same thing as I did with skype, can be done with these to. As they do release the audio hardware after a call, and you really only can talk to one person at the same time, its not a big deal. Ekiga uses arts for its ring signal and that goes over the pulseaudio without any problem. So I can use it as it is now.

audacity

Audacity is the very nice sound tool for recording and manipulation sound files. It can unfortunately not use the pulseaudio system and will lock up and not work nicely if you try.

As this might happen for some applications there is a small utility called pasuspender that will stop the pulseaudio system, run one application and start up the pa system again after its done. Very handy. So to run audacity you do:

pasuspender audacity

That will make audacity having full access to the hardware while its running and when it stops you get the sound back over pulsaudio again. I changed my menu item for audacity to do te same.

Links

HOWTO: PulseAudio Fixes & System-Wide Equalizer Support (Hardy Heron)

How To: The (almost) Perfect Pulse Audio Setup

PulseAudio – The Perfect setup talks a lot of different applications

When moving over to the new machine i installed Wine version 0.9.58 as this version worked and had almost no bugs. I have been trying new versions on and off when they come out but so far I haven’t found a newer version that was OK and worked well. So I stayed on 0.9.58. I have the Premium Graphics working and I can run 2 clients at the same time without the machine being to tired.

I did lock the version with both the synaptic package manager (package >> lock version) to lock it for synaptic but I also had to do a dpkg command to lock the adept package manager:

echo wine hold | sudo dpkg --set-selections

This was necessary for me so that the automatic updater would NOT update wine, but it was complaining all the time that there was a new release out there.

When I checked the updated today i saw that a new version of wine (1.1.2) was out there and after some checking on the forums it looked promising. So I decided to try it out. I just removed the hold of the application and let adept updated it:

echo wine install | sudo dpkg --set-selections

and then run the updater.

And it works – evryting except one small thing is working as it should. The thing that isn’t working is that when i start up the wine and eve client i get s 10 pixels or so alignment problem of the screen. So this requires a bit more googeling. But one way to solve it it to press <ALT> + <ENTER> to that wine switched to full screen more and then so the same again so it goes back into window mode – that works.

So after some googeling and finding out that the card i have in the machine is a new Intel HDA card, a Realtek ALC889. Apparently Intel has done this new standard of soundcards for on the motherboard mounting and as usually with a new standard its implemented slightly different on different machines.

So I did not find any notion of how to solve the problem, but there was this post that said that you could solve the problem by installing OSS4 instead of ALSA. There was ofcause a lot of discussions for and against OSS and if ALSA is better or of OSS of free enough, but frankly I don’t care. I want my stuff to work. I need the audible notification from kopete when someone send me a message, and i need my VOIP softphones to work, prefably with the USB Headset. What does it and how it does it I could not care less about.

This link >> will show you how to install OSS4 on a ubuntu 8.04 machines.

So I did this and now the sound worked. Kind of. I got the sound partly to wok. Speakers works and I can here when someone send a message, and I can use wine to play my eve and get sound and use amarok to play music.

One small thing that they forgot to mention in the post was that OSS don’t fully come up when you reboot the machines, and you have to run ldconfig to get audacity to find the sound drivers right.

So i did edit /etc/rc.local with the command

sudo vi /etc/rc.local

and added the line

ldconfig

to the end before the exit 0 line.

Then I started to test all my softphones and getting them over and working. And I found that no matter what I did and how I changed devices and setting I could not get the dam mic to work. Not the one on the machine itself or the USB headset one. So I put the project one ice for a while as sound kind of worked and it vacation time here in Sweden.

Back to ALSA

After some time on vacation its time to start working on the sound problem again.

For those who are interested just in the solving this is the solultion:

Install alsa 1.0.17 (not totally shure if thats nessesary)
Set the model to acer-aspire in 7etc/mopdprobe.d/alsa-base
Download hda-verb from here >>
Compile and copy the hda-verb to /usr/local/bin
Add one line to /etc/rc.local:
/usr/local/bin/hda-verb /dev/snd/hwC0D0 0x15 SET_EAPD_BTLENABLE 2

So the full explanation and all the links and so on.

I started by googeling and reading the forums. No full explenations what given enywhere, but some pointer here and there was given. The first one was to upgrade to ALSA 1.0.17. So i folowed the instructions here>> and installed it. Make sure that you read all of the posting as the alsa_2.sh script is not working but the last post on page 3 has the right code to install the new ALSA on the right place. The code is:

cd /lib/modules/2.6.24-19-generic
sudo cp -a kernel/sound/* ubuntu/sound/alsa-driver/

But you will have to change the version of the kernel you are running.

Here is the output of my aplay -l command:

**** List of PLAYBACK Hardware Devices ****
card 0: Intel [HDA Intel], device 0: ALC889 Analog [ALC889 Analog]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 0: Intel [HDA Intel], device 1: ALC889 Digital [ALC889 Digital]
Subdevices: 1/1
Subdevice #0: subdevice #0
card 1: Headset [Logitech USB Headset], device 0: USB Audio [USB Audio]
Subdevices: 1/1
Subdevice #0: subdevice #0

And the output of the grep for Codecs is:

Codec: Realtek ALC889
Codec: LSI ID 1040

This installed the new version of ALSA, but the default config did not work. If you read the second replie in the previous mentioned post, you will find a list of all models and you can also find it in /usr/src/alsa/alsa-driver-1.0.17/alsa-kernel/Documentation/ALSA-Configuration.txt file after installing ALSA 1.0.17.

Unfortunately the alc889 card is not in the list, so I started out and tried all the models on the alc888 list. And after changing and rebooting like 40 times it felt like – none of them work. So i when back into googleing again. And found a link to a post talking about the same problem and that you needed to send a code to the sound system to get it to work. I set the model to acer-aspire in /etc/modprobe.d/alsa-base as that seems to be the most logcal one.

So i downloaded the sourecode for the hda-verb program here>>. Unpacked it and compiled it. And then tested it by just running the command:

./hda-verb /dev/snd/hwC0D0 0x15 SET_EAPD_BTLENABLE 2

And it worked. So I copied the hda-verb program to /usr/local/bin and edited /etc/rc.local and inserted the line:

/usr/local/bin/hda-verb /dev/snd/hwC0D0 0x15 SET_EAPD_BTLENABLE 2

To the end of /etc/rc.local.before the exit 0 line. And now my sound works as it should.

Testing with all the applications and all of my sipphones and it works.

I did offcause have to reinstall the alsa version of the skype program, but as i started with the one it ok.

Update

After some testing i found that I have a small problem with the sound and that is that the sound caard is locked up when using it from multiple applications. So i probably need to test with pulseaudio to se if that solves the problem. But ths a minor thing.

This is one of the new #G modems that behaves like a USB stick at first when you enter it as it has the drivers and stuff for windows on that. The driver the sends a command to the device to switch to Modem moden and can then use it for the connection.

This is ofcuase a small problem for us LInux users as linux will se the device as a USB stick and the drives there is ofcuase not any use to us linux users.

Fortunately a program for handling this as been developed by Josua Dietze at http://www.draisberghof.de/usb_modeswitch/

This little program has a config file for different cards and will be used tgether with some config of the udev system to get all of this to work. I am using version 0.9.3 of the software.

So you start with downloading the software, and unpack it into a directory. Then compile it for your machine (a simple shellscript – compile.sh) is included. You need to have libusb installed as its using that so if your minssing that you need to do a:

sudo apt-get install libusb

I then copied the config file to /etc/usb_modemswitch.conf and the binary to /sbin/usb_modemswitch.

In the config file you make shure that (and this if ofcuase for my card) the following lines are uncommented and all other lines are commted out:

#######################################################
# Option GlobeSurfer Icon 7.2
#
# Contributor: The Pharscape Forum

DefaultVendor=          0x05c6
DefaultProduct= 0x1000

TargetVendor=           0x0af0
TargetProduct=          0x6901

MessageEndpoint=0x05
MessageContent="55534243123456780000000000000601000000000000000000000000000000"

This setup will switch my card over.

Udev

So the next part is to configure udev so that it will run usb_modemswitch when the device is plugged in.

I created a small udev rule file called 99-opteon.rules in /etc/udev/rules.d with the content:

##
#
# udev rule for Opteon ICON 7.2.
# ------------------------------
# Jan Saell <jan@irial.com>
#
# This rule switched over a Opteion ZeroCD (TM) usb modem card
# from storage mode to modem mode
#
##

BUS=="usb", ATTRS{idVendor}=="05c6", ATTRS{idProduct}=="1000",  RUN+="/sbin/usb_modeswitch"

The idVendor and idProduct values has ofcuase to be changed if you have another card.

So when i now plug in the device, udev sees it, and runs usb_modemswitch and switched it over to modem mode.

kppp

Next step is to get kpp working.

I created one modem called “Card” and one account called “Telenor”.

For most of the time you don’t need any user name and password for the things so only fill it in with some dummy values.

The Card is configures as follows:

And you have to go into the Modem tab and change some settings:

All the important settings is under the Modem Commands one:

The important settings here is:

AT+CGDCONT=1,"IP","services.vodafone.net"

The services.vodafone.net part is the APN and that is different depending on what service and ISP you are using.

A list of a lot of APN can be found here: http://franson.com/forum/topic.asp?TOPIC_ID=6669, http://www.vysoo.com/apn.php#416 or here: http://www.modmyifone.com/wiki/index.php/Carrier_APN_Settings.

The last thing is the Account setting:

This took care of everyting for me and I can now use the telnor services to connect over #G and get internet even if I dont have any other connectivity.

As a lot of you know I have had some change in my personal life, so I did this work the 1:th but haven’t had time to publish it. So thats why the date is perhapes a bit missleading.

I think now its 3 major things I need to get over:

• amarok and my music and connection to my ipod.
• my palm and syncing the palm
• the mobile network setting is have as a backup if networks fails.

And today we will do the amarok.

Amarok is the preferred music program that plays my mp3′s and more important it handles all my pod casts and downloads them and transfers them to my ipod. So the first thing is to get Amarok going. But to get that going I have to get the mysql databases over from the old machine.

I have already transferred over all my music collection. But amarok keeps the podcast in its own directory. And it also uses mysql and a database for ll the music collection data. Well amarok can be configured in a number fo ways but as I am using the mysql-database for other things I can as well use it for amarok.

The fist thing is to dump the databases from the old machine. I have copied over all the data but not the databases, and the best thing to use is to dump the databases with mysqldump. Over to the old machine and do the dump. I then check what databases I had on the machine, and use mysqldump for each database to a file:

mysqldump -p --extended-insert=FALSE amarok > /tmp/amarok.sql

And the repeat it for all the databases.

Ti start on the new machine I first have to get the right user and persmission into mysql as i plan to copy the amarok settings over. I dumped the mysql database also on the old machine just to be shure. And the user data is in the user table and the permissions is in the db table, so I edit the files and just take out the permissions and user data for the amarok db and amarok database user. Save it to a separate file. Then i ceated the amarok database and loaded the users with:

mysql -u xxx -p mysql < mysql_userload.sql

and the loaded the amarok database with:

mysql -u xxx -p amarok < amarok.sql

That should take care ok the database. Now it up to the settings of amarok itself.

Amarok

The config files for the amaok application is in .kde/share/config/amarokrc so i copied that from my old directory. The data for the application is in .kde/share/apps/amarok so that one to is copied over from the old directory. And now its time to start up nd test.

Starting amaork works fine – no error messages or anything and all my podcasts shows up. But nothing in the collection. None of my old album is in there,. Strange.

So after some more digging and googeling I found out that not only does the amarok stores the songs and albums and info in the database – it also stores the disk id’s in there. That to be able to handle removable disks and so its kind of god – but in this case its a bad thing.

The database table it is storing the information in is devices, so I cleaned out the amarok database and just started it up blank so it would fill in the information. I then dumped the database out and edited the old database dump so it reflected the new devices. I then reloaded the database but it still did not work.

Well I gave up the idea of reloading the database and just did a rescan of the music and that also worked.

ipod

So now when the amarok is working and downloading the podcasts I am listening to I need to get the ipod syncing going also.

So fist thing is just to plug it in and see if the system recognize it. And ofcause it does, and ask what to do with it. For now I select noting as im just testing. I had amarok turned of as i did not what that to interfere yet and it did not pop up and start automatically.

So eject the ipod and …. wait – that did not work. Ipod still attached. Eject again – noting works – now its not even mounted on the system but still connected according to the ipod – I hate hardware. Unpluging it did not make anything so i have to reset the ipod. I found the link to how to reset it here>>.

So reset the ipod and that did not damage the ipod and more googeling.

Apparently its a change in the udev system that makes the problem. I could plug in the ipod and do a sudo eject, and that worked but its kind of inconvenient and also – amarok will not make that working.

ATTRS{type}=="0",                     GROUP="disk"

sudo vi /etc/udev/rules.d/40-permissions.rules

sudo udevadm control --reload_rules

One of the main things still to fully solve is the sound issue. I have tried to switch over to oss4 and that works and I do get sound from the KDE system, but it seems that a lot of the IP phones I’m using will not work. But lets go over the things one step at the time. The things I will cover today is the

• Blurtooth mouse
• Flash for firefox
• krdc
• skype

Bluetooth mouse

I do think that the touchpad is ok for some minor work, but when using the keyboard fully during the day and writing documents and other things, I think its in the way. I do have the tendency of moving the pointer while im typing and that usually means that I insert text wher I dont whant it. So I normally use a external Bluetooth mouse (from Microsoft ). But that has to be configured into the system.

Its realy not hard. I uses the command hcitool scan to se what different bluetooth devices is around and the do:

sudo hidd --connect 00:xx:xx:xx:xx:xx

where 00:xx:xx:xx:xx:xx if the adress of my mouse. That connects the mouse (and you have to remember to turn on bluetooth every time you boot the machine, but thats only one button). But I whant the connect to be done everytime the system boots, so we need to edit the /etc/default/blurtooth file.

Change the lines HIDD_ENABLED and HIDD_OPTIONS to the following

HIDD_ENABLED=1
HIDD_OPTIONS="--connect 00:xx:xx:xx:xx:xx --master --server"

This means that the system will connect to the mouse at startup and – as long as you remember to turn bluetooth on – you will have it running even during the login screen.

Flash for firefox

As im running hthe 64 bits version of the kubuntu OS and therefore the 64 bits version of firefox there is some notes out there about getting the flash plugin to work. After some googeling there seems to be quite easy to get it to work now. I needed only to install nspluginwrapper – this wrapper alows you to run the u386 versions of the plugins on 64 bits version of the OS and firefox. SO just run:

sudo apt-get install nspluginwrapper flashplugin-nonfree lib32nss-mdns

This installs the wrapper and the flash plugin and all works very well. Ther eis a small issue with th oss sound system but I will get back to that when I talk about the OSS install.

krdc

krdc is the KDE program for running remote desktop in the Linux world. Its a wonderful tool for using when you are a consultant like me and have customers with a lot of windows system. It allows you – among other things -  to connect to a windows server.

There is ofcuase a number of setting you need for that and I preferred to have my old setting copied over to the new system.

The config file for krdc is only in the file .kde/share/config/krdcrc so i copied that from the old system and now that works.

Skype

I have stuggled and keept out of the skype world for some time as I do work with free IP phones but more and more of my clients have skype and some is using that as a chat system also, so I did start and installed a client on the old system. I then ofcuase have to copy that over to the new system.

So the prerequsites for skype is ia32libs, and lib32asound2 so i installed this with:

sudo apt-get install ia32-libs lib32asound2

and the download the skype client with:

wget -O skype-install.deb http://www.skype.com/go/getskype-linux-ubuntu

and then install it all with:

sudo dpkg -i –force-all skype-install.deb

The –force-all is required as the .dep package is for the i376 platform specific and you whant to install it on this 64 bits version. This gives you the skype klient, and besides the oss / sound problem it works ok.

1. Sound system
2. Camera
3. Media card reader

But none if this is very urgent – even if the sound part is a bit annoying and has to be solved fairly soon. The sound bit seems to be a bit OK as I managed to get the sound working with my USB headset so that should be OK, and means that I can use the system for the normal uses I think, like all my different softphones. I do work quite a lot with IP telephony and not being able to test on my main production system would be hard.

knetworkmanager

As I’m now in the summerhouse on the west coast of Sweden I connect to the network over wlan. This means to get the knetworkmanager settings over. This will help you get all the configs and networks you have connected to over to the new installation.

knetworkmanager has its setting in the file .kde/share/config/knetworkmanagerrc so i checked it and the only thing in the new one was that it had a new version in the [General] sections so i copied over all the rest of the lines.

kwallet

knetworkmanager also has all the passwords and so stored in a kwallet so i have to copy that settings over. The settings is in the .kde/share/config/knetworkmanagerrc and in .kde/share/apps/kwallet so i copied both tf hem over.

And then relogged in again and it just connected to the network. Nice.

MSN/Kopete

Well one of the main things to get over is the kopete or MSN messenger program. Kopete gives you the possibility to connect to msn and other messenger server and I do uses that as everyone else. The god thing with this program its that it can connect to more than one system at the same time.

It uses kwallet to store its passwords but thats already take care of, so the only files here is .kde/share/apps/kopete and .kde/share/config/kopeterc and .kde/share/config/kopete.eventsrc so i copied all of that over and now kopete starts up nice and easy and connected to all messenger services as i had.

1. Move the old directory into place and trie to get KDE4 to work with that
2. Or to copy data over and get one application after another to work

And I think that the second one was the best way. The old system i was running and moving from was an Kbuntu 7.04 so there was 2 versions to upgrade over and I was not shure that the upgrade system and all would cope with that, and I realy did not have much that I already had running on the new system so I decided to get application after application over.

There is not that many ones that is absolutly crutial, and there are some I would like. Im not shure I can rangorder them so im just going to talk about each one as I go along.

xplanet

As im kind of happy with how my old system worked I whant to have xplanet as the background. This program lets you have a background of, well any planet in the solar system, but I uses it to display a earthglobe centered over stockholm. This way I can see what the time is on different places on the world by just takeing a glance on the desktop. It can also download a cloudmap and ovelay that obver the normal earth picture so you also gets a notion of how the weather will be.

For some reason the xplanet was not installed so I had to install that:

sudo apt-get install xplanet xplanet-images

did the trick. Then i had to copy over the config files and the doanloader. The config files and the coordinates is under /usr/share/xplanet. The config file I uses is default.conf and the marker file i uses for the locations I whant is markers/earth. Copy them over and its ok.

The downloader is called doenload_clouds.pl and is also in /usr/share/xplanet. I  copied that one over and then installed it with sudo crontab -e. The line to add is:

0 * * * *       /usr/share/xplanet/download_clouds.pl 2>&1 > /dev/null

The script will only do it of the file is older that 3 hours so thats ok.

Firfox

One other thing t get going is the bookmakrs and evryting with firefox. All the sertings fr firefox is in the .mozilla directory, and I was i bit curious if it would handle the old files right. The new system uses Firefox 3 and the old one had 2.

I moved teh new directory (.mozilla) over to .mozilla-new and copied in my old .mozilla directory. And the started up firefox. And that worked directly – no problems – and I it just worked. All my bookarks, password and even cookies is moved over – very nice and easy.

So while the eve system was down I spent some time trying to sync over all data from my old laptop to the new one.

First thing was to get all the installed applications installed the same way on the new machine. I found some references out on the net on how to dump the installed applications on the old machine to a file and then use that to install the same applications on the new one. So i tried that and it almost worked. Not quite.

To get all the installed applications out from a ubuntu/kubuntu or any debian based system you use the dpkg command and some grep and awk to make a nice list:

sudo dpkg --get-selections | grep '[[:space:]]install' | awk '{print $1}'

the command dpkg –get-selections gets all applications on the system. The grep then filters out only the installed ones and awk just prints out the package name.

Then do some editing of the list as there are some things you don’t want to install on the new system. Like i don’t want to install the fgl xorg drivers as i’m not using a ATI card on the new system, but already have installed the nvidia driver. So remove the following package in that case:

fglrx-amdcccle
fglrx-kernel-2.6.20-16-generic
fglrx-kernel-source
xorg-driver-fglrx
xorg-driver-fglrx-dev

Then I removed all the linux version packages like linux-headers-2.6.20-15 and others like that as i dont want old version on the system.

I then transfered the list over to the new system and use xargs to get apt-get to install it. The documentation out there just did a cat on the file and piped it into xargs – if you do this and you get some questions from apt-get you get a break as the input is connected to /dev/null. So i used the command:

xargs -i /tmp/apt-get-list.txt -n 10 sudo apt-get install

The -i flag tells xargs to use the file instead of stdin. and -n tell xargs to only do 10 args at a time.

Now it ran fine and I get a number of installs and I also had to say Y a number of times to install extra packages and so, but all in all I got all my applications installed now.

Empyrian Age

When I had installed everything I just – to make sure that the system still worked as it should – started up into the EVE client. I did not expect that the server would be back up, but it was. It asked if it should download the new patch and I said yes. It downloaded it and installed it – No problems. And finally I was in and all worked. A very nice patch indeed.

So i tested the official client also and that upgraded and installed a new version also without any problems.

So i can now run the new version of EVE. I’m happy.

We have been moving from the house to the apartment for the last month and the only real thing left now is to finish off the garden so its in hand over state. And thats what we have been doing during the weekend.

But there has been small times when we haven’t moved or carried things. And I decided to test the wine and premium graphics to see if I could get it to work. From what I have read on the forum the wine version of EVE runs without any big problems.

So i copied all the .wine directory over form my old laptop and used that as a start – that way I also would retain my old setting in the EVE client and thats very useful.

First thing i tested was installing wine from the normal repository and that one did not work.

I got some strange errors and it would just not start. I then added the wine-hq repository with the command:

sudo wget http://wine.budgetdedicated.com/apt/sources.list.d/hardy.list -O /etc/apt/sources.list.d/winehq.list

And then do the install with:

sudo apt-get install wine

The version I got installed this time was the 1-0-rc3 and I still got the error. Problems – what to do. The error i got was:

err:ole:CoGetClassObject class {9a5ea990-3034-4d6f-9128-01f3c61022bc} not registered
err:ole:CoGetClassObject no class object {9a5ea990-3034-4d6f-9128-01f3c61022bc}

After checking and doing some reading is seems that the bug is due to some old initiation error and a patch for wine is out there. But its strange. I did run the old version and had no problem.

I decided to install the client from scratch. First i installed the msttcorefonts and copy them to ~/.wine/drive_c/windows/windows/fonts.

cp /usr/share/fonts/trutype/msttcorefonts/* ~/.wine/drive_c/winedows/fonts/

Then I went to the EVE-Online site and downloaded the windows premium client from the EVE-Online site. and then started the installer. After installing the client, dont start it just yet. Go into the eve directory and run the RedistD3DXOnly.exe with

wine .wine/drive_c/Program\ Files/CC/EVE/bin/RedistD3DXOnly.exe

That worked but i still got some weird error

preloader: Warning: failed to reserve range 00000000-00010000

This is because the new version of kubuntu has a harder memory protection and that conflicts with what wine does, so you need to relax that setting. This can be done with the command:

sudo sysctl -w vm.mmap_min_addr=0

But that only changes it temporary. To set it fully you need to edit /etc/sysctl.conf. Go in and change the line:

vm.mmap_min_addr = 65536

To

vm.mmap_min_addr = 0

This will change it for the futrure and keep the setting when you reboot the system.

Thats was it. I have wine working. BUt the latest version has some weird graphics errors, so I did some more googeling. Aparently the best version for running the eve premium grpahics is the 0.58 version.

So i downloaded it from the wine debian archive, removed the normal version and installed it. And perfect. You need to turn of shading in the client and change the wine registry setup for OffscreenRenderingMode.

Edit the ~/.wine/user.reg file and add the following lines:

[Software\\Wine\\Direct3D]
"DirectDrawRenderer"="opengl"
"OffscreenRenderingMode"="fbo"
"PixelShaderMode"="enabled"
"VertexShaderMode"="hardware"
"VideoMemorySize"="256"

Then startup the client and change the shadow quality to of in the setting. This makes the premium graphics work very well,

I downloaded the official client for EVE-Online and saved the .deb file and run the installation as I should. No problem. Started up the configure program to configure it and I got an error that it missed some 32 bits libraries and I had not ofcourse installed that yet so that was kind of expected.

A bit more googleing gave me the info that I needed ia32-libs, util-linix, lib32nss.-mdns libasound2-plugins, so i installed that with the command:

sudo ap-get install ia32-libs util-linux lib32nss-mdns libasound2-plugins

Then it was time test again. Into the config and run all the test – no problems, all passed ok and no problem.

So fire up the game – and wow it starts but in full screen mode.

Bakc into the config program and change the “Run in Window” from No to 1280 x 1024. Back into the game and now it start as it should, but i can move the mouse outside the windows. Back into the config and change that to – thats the DCGrab check box and now the game works just right.

And the framerate i get outside our pos in 0.0 is almost up to 30 fps and on the old one I had around 8-10 fps – very nice and smooth.

When I get the wine version running its suposed to be able to run even the premium graphics moce so that will be nice.