Last week i was confronted with a compromised custom Asterisk (Elastix) based PBX that suffered security issues, both in the configuration, such as easy passwords very lenient permissions, as well as the normal CentOS provided security updates since some software/drivers outside of the native repo’s was installed “yum update” was promissing the old RPM dependency hell.

In short: this PBX needed some love desperately.

I’m sure that someone more familiar with CentOS (red hat) could easily solve these issues and reinstate the existing install into a more stable and secure state, but since i limit myself mainly to Debian based and would like to get rid of the Elastix/FreePBX interface to Asterisk i decided that a fresh Debian install would be needed.

Pretty simple you would say, only there was one minor issue, the machine was about 7,814 kilometers down the road and remote hands were hard to find. Auch!!!!

After wasting about a week to find a solution to this I started looking for some more unconventional solution and bumped into some old HowTo’s how to do it remotely via SSH. It took me a number of runs in a Virtual testing machine but i finally got the required routine down.

The Basic idea is to do the following:
* Disable the SWAP partition.
* Debootstrap Debian into the old SWAP partition.
* Reboot into Debian on the old SWAP partition.
* Copy over the Debian installation to the normal / partition over the existing CentOS install..
* Reboot into Debian on / root.
* Re-enable the SWAP partition.

====Investigate the current situation====

First have a look at the current partition layout issuing the command.
fdisk -l

This gave me the following result
[root@elastix ~]# fdisk -l

Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 * 1 13 104391 83 Linux
/dev/sda2 14 19457 156183930 8e Linux LVM

fdisk shows a /boot partition on /dev/sda1 and a LVM setup on /dev/sda2.

Lets have look at the LVM setup by issuing the command:
lvdisplay

This shows the following LVM partitions.
[root@elastix ~]# lvdisplay
--- Logical volume ---
LV Name /dev/VolGroup00/LogVol00
VG Name VolGroup00
LV UUID tt92rH-Aoo9-TQUY-PlpS-4fot-MmzT-qZpkSa
LV Write Access read/write
LV Status available
# open 1
LV Size 147.00 GB
Current LE 4704
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:0

— Logical volume —
LV Name /dev/VolGroup00/LogVol01
VG Name VolGroup00
LV UUID qTAuIF-oIFN-95xV-C0jd-xmrd-Nf0v-dxe7eI
LV Write Access read/write
LV Status available
# open 1
LV Size 1.94 GB
Current LE 62
Segments 1
Allocation inherit
Read ahead sectors auto
- currently set to 256
Block device 253:1

Above you see the 147.00 GB / root partition on /dev/VolGroup00/LogVol00
and the 1.94 GB SWAP partition on /dev/VolGroup00/LogVol01.

The 1.94 GB SWAP partition is more then enough to hold the temporary debian minimum install so we will use that space.

First check your running services and the consumed RAM too see if we can safely do without the SWAP, disable unneeded services if needed.

====Remove the SWAP partition====
Disable the SWAP by issuing the following command:
swapoff -a
Too check the current status of the PV holding the LV’s:
lvm pvs
Below you see that there is 148.94G on the Physical Volume and no free space.
[root@elastix ~]# lvm pvs
PV VG Fmt Attr PSize PFree
/dev/sda2 VolGroup00 lvm2 a- 148.94G 0

Lets remove the LVM partition that holds the SWAP space.
lvm lvremove /dev/VolGroup00/LogVol01
Answer “y” as shown below to remove the active logical volume.
[root@elastix ~]# lvm lvremove /dev/VolGroup00/LogVol01
Do you really want to remove active logical volume LogVol01? [y/n]: y
Logical volume "LogVol01" successfully removed

After removing /dev/VolGroup00/LogVol01 investigate the Physical Volume again.
lvm pvs

Below you see that there is now 1.94G of free space within the Physical Volume.
[root@elastix ~]# lvm pvs
PV VG Fmt Attr PSize PFree
/dev/sda2 VolGroup00 lvm2 a- 148.94G 1.94G
====Resize the Physical Volume====
Now we can resize the Physical Volume to the size of the VolGroup00 partition and thus make the 1.94G available for use.
lvm pvresize /dev/sda2 --setphysicalvolumesize 148.94G
This will give the confirmation that the Physical Volume is resized
[root@elastix ~]# lvm pvresize /dev/sda2 --setphysicalvolumesize 148.94G
Physical volume "/dev/sda2" changed
1 physical volume(s) resized / 0 physical volume(s) not resized

Lets have another look at the Physical Volume using the cylinder size of the partition.
lvm pvs --units s

This will show as below:
[root@elastix ~]# lvm pvs --units s
PV VG Fmt Attr PSize PFree
/dev/sda2 VolGroup00 lvm2 a- 312344576S 4063232S

Lets have a look with parted to see the cylinder layout on /dev/sda.
parted /dev/sda unit s print

[root@elastix ~]# parted /dev/sda unit s print

Model: ATA Hitachi HTE72321 (scsi)
Disk /dev/sda: 312581807s
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number Start End Size Type File system Flags
1 63s 208844s 208782s primary ext3 boot
2 208845s 312576704s 312367860s primary lvm

Information: Don’t forget to update /etc/fstab, if necessary.

====Resize the /dev/sda2 partition====

Too resize the /dev/sda2 partition too fit exactly the Physical Volume we already made smaller and be able to use the free space, we first remove the entire LVM partition and recreate it later with a smaller size.

To remove the LVM partition issue the command:
parted /dev/sda rm 2

[root@elastix ~]# parted /dev/sda rm 2
Information: Don't forget to update /etc/fstab, if necessary.

And the recreate it, Note that the Start cylinder needs to be the same as the one removed earlier and the End cylinder is the same as the old one MINUS the free cylinders that were show earlier by “lvm pvs –units s”
parted /dev/sda mkpart primary 208845s 308513472s
Below shows no errors so it all seems to be fine.
/dev/sda2 VolGroup00 lvm2 a- 312344576S 4063232S
[root@elastix ~]# parted /dev/sda mkpart primary 208845s 308513472s
Information: Don't forget to update /etc/fstab, if necessary.

Set the LVM flag on the recreated LVM partition.
parted /dev/sda set 2 lvm on

To check the new state of the partition layout on /dev/sda with parted issue the command:
parted /dev/sda print
This will show the following result:
[root@elastix ~]# parted /dev/sda print

Model: ATA Hitachi HTE72321 (scsi)
Disk /dev/sda: 160GB
Sector size (logical/physical): 512B/512B
Partition Table: msdos

Number Start End Size Type File system Flags
1 32.3kB 107MB 107MB primary ext3 boot
2 107MB 158GB 158GB primary lvm

And to have a look with fdisk
fdisk -l /dev/sda
[root@elastix ~]# fdisk -l /dev/sda

Disk /dev/sda: 160.0 GB, 160041885696 bytes
255 heads, 63 sectors/track, 19457 cylinders
Units = cylinders of 16065 * 512 = 8225280 bytes

Device Boot Start End Blocks Id System
/dev/sda1 * 1 13 104391 83 Linux
/dev/sda2 14 19205 154152314 8e Linux LVM

We are not ready yet, to make the resizing of the Physical Volume final issue the command:
pvresize /dev/sda2
====Create a new partition in the free space====

We will create a new partition to hold the temporary debian install.
Start up fdisk on /dev/sda:
fdisk /dev/sda
Its never a bad thing to Print the current layout by issuing “p”
Command (m for help): p
Instigate a New partition with “n”
Command (m for help): n
Select “p” to make it a Primary partition
Command action
e extended
p primary partition (1-4)
We alreday have sda1 and sda2, so lets select “3″
Partition number (1-4): 3
Press “enter” to have it start on the first available cylinder
First cylinder (19205-19457, default 19205):
Using default value 19205
And press “enter” to have it end on the last available cylinder
Last cylinder or +size or +sizeM or +sizeK (19205-19457, default 19457):
Using default value 19457
Now we want to select what Type our new partition should become with “t”
Command (m for help): t
Select the partition number we want change the type for with “3″
Partition number (1-4): 3
And use “83″ to make it a linux type partition
Hex code (type L to list codes): 83
To Write the new partition layout to disk use “w”
Command (m for help): w
Since /dev/VolGroup00/LogVol01 is deleted from the disk and no longer available in the partition layout we remove it from fstab.
nano /etc/fstab
And make fstab to look like below:
/dev/VolGroup00/LogVol00 / ext3 defaults 1 1
LABEL=/boot /boot ext3 defaults 1 2
tmpfs /dev/shm tmpfs defaults 0 0
devpts /dev/pts devpts gid=5,mode=620 0 0
sysfs /sys sysfs defaults 0 0
proc /proc proc defaults 0 0
To finalize the new partition layout the disk has to be remounted, and since we can not unmount the / partition the system is running on we need to reboot.
reboot
And wait a while for it to reboot and log back in via SSH.
Yes, i know. Some sweat started to break out waiting for SSH to welcome you back in.

====Create a debootstrap RPM on another debian machine===

On a seperate Debian machine we first install alien
apt-get install alien

Get a debootstrap .deb for squeeze
wget http://ftp.nl.debian.org/debian/pool/main/d/debootstrap/debootstrap_1.0.26+squeeze1_all.deb
Use alien to convert the .deb to a .rpm
alien -rkv debootstrap*.deb
Copy over the .rpm to the system we are converting
scp debootstrap-1.0.26+squeeze1-1.noarch.rpm root@192.168.0.99:/usr/local/src/

====debootstrap Debian to the new partition===
Now that the new partition is ready to be used, we first create a new file system on it.

mke2fs -j /dev/sda3

Then we need to mount it but first we create an easy alias for the new mount.
export ASD=/mnt/asd
Create the location for the new mount
mkdir -p $ASD

And mount /dev/sda3
mount /dev/sda3 $ASD
Lets install the debootstrap rpm we cerated earlier and move to the location we copied it to, in our case:
cd /usr/local/src/
And install it
rpm -Uvh debootstrap*

Before we use debootstrap we need to install binutils, otherwise debootstrap can not unpack the .debs
yum install binutils
And finally debootstrap debian into our mounted /dev/sda3, always check which mirror to use because this could take a while if you have a slow far away mirror.
/usr/sbin/debootstrap --arch i386 squeeze $ASD http://mirrors.geekbone.org/debian

After debootstrap is finished we re-use some old configs, first copy over the resolv.conf file
cp /etc/resolv.conf $ASD/etc/
And then copy the hosts file
cp /etc/hosts $ASD/etc/

Create a hostname file
nano $ASD/etc/hostname
And use for example the hostname
voip

Now we change our root to the new debian system
chroot $ASD /usr/bin/env -i HOME=/root TERM=$TERM PS1='u:w$ ' PATH=/bin:/usr/bin:/sbin:/usr/sbin /bin/bash --login

Here we create the fstab file for the partitions we want mounted.
nano /etc/fstab

# filesystem mount fs-type options dump fsck-order

/dev/sda3 / auto defaults 0 1
proc /proc proc defaults 0 0

After that we mount the proc file system
mount -t proc proc /proc

hostname voip.example.com

Create our network details to be used when the new system boots
nano /etc/network/interfaces

# Used by ifup(8) and ifdown(8). See the interfaces(5) manpage or
# /usr/share/doc/ifupdown/examples for more information.

auto lo
iface lo inet loopback

auto eth0
iface eth0 inet static
address 192.168.0.99
netmask 255.255.255.0
network 192.168.0.0
broadcast 192.168.0.255
gateway 192.168.0.1
dns-nameservers 192.168.0.1

dpkg-reconfigure passwd

Set a password for root, without it we could not login remotely anymore
passwd

Install the locales package
apt-get install locales
And reconfigure the locales to be used
dpkg-reconfigure locales
Select en_US.UTF-8 UTF-8 to be installed and used as default.

After that we set up our time zone
dpkg-reconfigure tzdata

Set up the sources to be used with apt-get
nano /etc/apt/sources.list

deb http://mirrors.geekbone.org/debian/ squeeze main contrib non-free
deb-src http://mirrors.geekbone.org/debian/ squeeze main contrib non-free

deb http://security.debian.org/ squeeze/updates main contrib non-free
deb-src http://security.debian.org/ squeeze/updates main contrib non-free

update apt’s index with new repositories
apt-get update

And install ssh
apt-get install openssh-server

Check your system and find a matching kernel to be used
apt-cache search linux-image

In my case on this 32 bit machine, the often used 686 version
apt-get install linux-image-2.6-686
After the new kernel is installed we logout of the chroot
logout

And we copy over the newly installed kernel images to the /boot partition
cp $ASD/boot/vmlinuz-* /boot
cp $ASD/boot/initrd.img-* /boot
Now we need to update grub to use our debian kernel and partition on the next time it boots
nano /boot/grub/menu.lst

Mind that the boot stanza second from the top seems to be the default booted
# grub.conf generated by anaconda
#
# Note that you do not have to rerun grub after making changes to this file
# NOTICE: You have a /boot partition. This means that
# all kernel and initrd paths are relative to /boot/, eg.
# root (hd0,0)
# kernel /vmlinuz-version ro root=/dev/VolGroup00/LogVol00
# initrd /initrd-version.img
#boot=/dev/sda
default=1
timeout=5
splashimage=(hd0,0)/grub/splash.xpm.gz
#hiddenmenu
title Elastix (2.6.18-164.el5xen)
root (hd0,0)
kernel /xen.gz-2.6.18-164.el5
module /vmlinuz-2.6.18-164.el5xen ro root=/dev/VolGroup00/LogVol00
module /initrd-2.6.18-164.el5xen.img
title Debian!
root (hd0,2)
kernel /boot/vmlinuz-2.6.32-5-686 root=/dev/sda3 ro
initrd /boot/initrd.img-2.6.32-5-686
title Elastix-base (2.6.18-164.el5)
root (hd0,0)
kernel /vmlinuz-2.6.18-164.el5 ro root=/dev/VolGroup00/LogVol00
initrd /initrd-2.6.18-164.el5.img

After grub is updated, we unmount the mounted /dev/sda3
umount $ASD/proc
umount $ASD
Its time to hold your breath again and hope you did everything OK
reboot

====Prepare the final system for use====
If all went fine we have just booted in our new Debian environment on /dev/sda3.

Create a new file system on /dev/sda1, i like to use ext4 for this.

mkfs.ext4 /dev/sda1

Create a temporary /boot location
mkdir /boot2

And mount /dev/sda1 on there
mount /dev/sda1 /boot2
Copy over all from the old temporary boot location to the new temporary boot location
cp /boot/* /boot2/

Unmount /dev/sda1 again and remove the boot2 placeholder and the old temporary /boot
umount /boot2
rmdir /boot2
rm -rf /boot/*
nano /etc/fstab
Set up the new /boot location to be used on the next boot
# filesystem mount fs-type options dump fsck-order

/dev/sda1 /boot ext4 defaults 1 2
/dev/sda3 / auto defaults 0 1
proc /proc proc defaults 0 0

Mount the /boot partition
mount -a

Lets install grub2
apt-get install grub2

and make sure you choose to install it on the MBR of /dev/sda

Now lets prepare our new and final / root partition.
First we create a new filesystem on /dev/sda2
mkfs.ext4 /dev/sda2

And the use dd to copy over the running temporary debian system to its final location
dd if=/dev/sda3 of=/dev/sda2 bs=1024
Issue a filecheck on /dev/sda2 to be sure
e2fsck -f -y /dev/sda2
And resize it
resize2fs -p /dev/sda2
And then another file system check
e2fsck -f -y /dev/sda2

Besides using a new /boot location on the next boot we also want our / root to be on /dev/sda2
nano /etc/fstab

# filesystem mount fs-type options dump fsck-order

/dev/sda1 /boot ext4 defaults 1 2
/dev/sda2 / auto defaults 0 1
proc /proc proc defaults 0 0

Its time to update grub2 to reflect the new situation
update-grub

Het feit dat ze een “dongle”  “dongel” noemen en het KPN type nummer  overeenkomt met een release van Ubuntu voorspelde al niet veel goeds om te traceren wat voor type apparaat dit eigenlijk is en hoe deze aan de praat te krijgen onder Linux, in dit geval Ubuntu Karmic 9.10.

Wanneer je de deze USB dongle inplugt begint na een paar seconden een rood lampje te branden ten teken dat het apparaat herkend wordt.
Het commando
lsusb
geeft de volgende output
Bus 001 Device 004: ID 0af0:d057 Option
Zo makkelijk gaat het echter niet aangezien Ubuntu de dongle nog niet herkend als USB modem maar als een CD.

Dit kun je ook zien als je na het inpluggen het commando

dmesg
invoert, waarschijnlijk krijg je dan het volgende te zien

[ 466.993290] scsi 7:0:0:0: CD-ROM ZCOption Icon CD 1.00 PQ: 0 ANSI: 4
[ 467.007957] sr0: scsi-1 drive
[ 467.008438] sr 7:0:0:0: Attached scsi CD-ROM sr0
[ 467.008723] sr 7:0:0:0: Attached scsi generic sg1 type 5
[ 467.418375] ISO 9660 Extensions: Microsoft Joliet Level 1
[ 467.421443] ISOFS: changing to secondary root

De reden dat er een CD filesysteem op deze USB staat is niet zonder reden, onder windows wordt deze namelijk automatisch herkend en een autorun bstandje vanaf de “CD” opgestart waardoor alle drivers e.d. vanzelf geinstalleerd worden, een prachtige uitvinding natuurlijk waar we onder Ubuntu niks aan hebben.

Voor zover ik heb kunnen achterhalen gaat het hier om een GlobeTrotter GI1505 van de producent Option, al ben ik niet 100% zeker van het type.
En de vraag is hoe deze aan de praat te krijgen en als USB modem herkend te laten worden.
Om dit te doen heb ik het volgende bestandje geinstalleerd.

udev.tar

download deze naar bijvoorbeeld de Desktop en zoek de locatie waar het bestandje staat op in een terminal

cd Desktop
en pak het bestandje vervolgens uit
tar zxf udev.tar.gz
ga via de command line naar het mapje
cd udev
Om dit te kunnen installeren moeten we eerst nog het nodige installeren.
Eerst de build tools
sudo apt-get install build-essential
De kernel headers.
check eerst je preciese kernel versie met het commando
uname -r
En pas desnoods onderstaande aan aan jouw preciese kernel versie
sudo apt-get install linux-headers-2.6.31-19-generic
Vervolgens alleen nog het pakketje libusb-dev
sudo apt-get install libusb-dev

En dan zijn we klaar 0zerocdoff te compileren met het volgende commando
sudo make
En vervolgens.
sudo make install

Dat laatste commando zou ongeveer het volgende resultaat moeten laten zien.
install -d /usr/sbin
install -d /etc/udev/rules.d
install ozerocdoff /usr/sbin
cp hso.udev /etc/udev/rules.d/51-hso-udev.rules
install -d /usr/share/hal/fdi/preprobe/20thirdparty
cp 10-wwan-hso-preprobe.fdi /usr/share/hal/fdi/preprobe/20thirdparty
install -d /usr/share/hal/fdi/information/20thirdparty
cp 10-wwan-quirk.fdi /usr/share/hal/fdi/information/20thirdparty
install -d /usr/lib/hal/scripts/
install hal-serial-hsotype /usr/lib/hal/scripts/
install -d /etc
install osetsuspend /usr/sbin
cp hso-suspend.conf /etc

De specefieke dongle in deze post is te nieuw om al herkend te worden door datgene wat we net geinstalleerd hebben dus we moeten nog het een en ander aanpassen

sudo gedit /etc/udev/rules.d/51-hso-udev.rules
Onder het kopje “New Syntax”

zie een stuk of wat regels staan die lijken op deze
ATTRS{idVendor}=="0af0", ATTRS{idProduct}=="7601", RUN+="/usr/sbin/ozerocdoff -wi 0x%s{idProduct}"

Copier een van deze regels en plak deze onder de anderen voor de laatste regel “GOTO=”hso_end”" en pas vervolgens het “idproduct” aan.

ATTRS{idVendor}=="0af0", ATTRS{idProduct}=="d057", RUN+="/usr/sbin/ozerocdoff -wi 0x%s{idProduct}"
in ons geval dus “7601″ vervangen door “d057″, de d057 kennen we van de eerdere output van lsusb, namelijk 0af0:d057 dus eventueel aanpassen als je een iets ander type hebt.

Vervolgens in het zelfde bestandje onder de kop “Old Syntax” doen we hetzelfde en voegen we een nieuwe regel toe, copieer bijv. de lijn
SUBSYSTEM=="usb_device", SYSFS{idVendor}=="0af0", SYSFS{idProduct}=="7601", SYSFS{bDeviceClass}=="00", RUN+="/usr/sbin/ozerocdoff -wi 0x%s{idProduct}"
En vervang hier weer de “7601″ voor “d057″ zodat het er zo uit kom te zien
SUBSYSTEM=="usb_device", SYSFS{idVendor}=="0af0", SYSFS{idProduct}=="d057", SYSFS{bDeviceClass}=="00", RUN+="/usr/sbin/ozerocdoff -wi 0x%s{idProduct}"

Sluit vervolgens het bestandje af en sla deze op. En herstart de machine.
Als we na het herstarten de dongle weer in de USB poort stoppen dan geeft het commando
dmesg
Het volgende resultaat te zien.
[ 80.184102] usb 1-5: new high speed USB device using ehci_hcd and address 5
[ 80.318212] usb 1-5: configuration #1 chosen from 1 choice
[ 80.324180] scsi7 : SCSI emulation for USB Mass Storage devices
[ 80.325795] usb-storage: device found at 5
[ 80.325803] usb-storage: waiting for device to settle before scanning
[ 81.489855] usb 1-5: USB disconnect, address 5
[ 81.932162] usb 1-5: new high speed USB device using ehci_hcd and address 6
[ 82.066312] usb 1-5: configuration #1 chosen from 1 choice
[ 82.070176] hso0: Disabled Privacy Extensions
[ 82.072488] hso 1-5:1.5: Not our interface

Hier zijn we blj mee aangezien Ubuntu nu de KPN dongel herkent als een USB modem en met een linker muisklik op de “network-manager” zien we de optie om een “New Mobile Broadband GSM connection te configureren.

Klik deze aan en doorloop de stappen in de configuratie wizard, selecteer “the netherlands” en “KPN mobile” en hou voor de rest de default opties.

Bij de eerste keer verbinding maken zal er gevraagd worden om een pin en wachtwoord, voor beiden is het 0000 , dus 4xnul.

Het enige probleem is nog dat de versie van “modemmanager 0.2″ in Karmic geen verbinding kan maken, installeer hier voor het pakket modemmanager 0.3 van Lucid 10.04

After years of promises there finally has been a mayor overhaul of the Ubuntu theme after the “Brown Human Theme” that we have been accustomed to for so long, I know the usual complaints about the old Faeces Theme but i actually kind of liked it, or at least its what I’m used too and thought that Karmic was almost pretty damn chocolate perfect.

So, before continuing my tasteless rant here i would like to thank the people who worked on this new theme because its not something i can do or even can do better, so please understand where this is coming from before taking it personal and i’m glad the theming of the metacity theme, desktop background, GDM and splash are finally heading consistency.
Also that i have remarks does not mean its all bad, its just easier to be negative so take it as positive criticism ;)

The timing

My first head-shacking experience is about the timing of the introduction of the theme itself, actually I had this about various new introductions in LTS releases.

I understand that a new release, and especially a LTS, has to be pushed with as much media attention as possible and a new exciting shiny new theme is definitely something to advertise on the billboards, but this sounds like a commercial decision to me and not something I’m happy with as a user in the long term.

Software gets better with age and so do themes, why would you want to put a new untested theme in something that you are going to be stuck with a long time, especially with the “only bugfixes” approach.

If you start a new theme with the LTS then the next LTS -1 will have the perfect matured version of that theme, that does not make sense to me.

The colour

OK, this is as subjective as its going to get but I actually liked the brown, as did many others. At least it was distinctive to Ubuntu. But yes, any colour will have its fair share of lovers and haters, but purple?

A choice is a choice and it has to be made, fair enough, but it took me some time to find out that this theme was described as “purple” since my first google after seeing the new colour was “lucid new pink theme” Am I the only one who’s first idea was that it was pink?

Apparently plenty of people seemed to like purple because it was the same purple as MacOSX. WTH??? Ubuntu should not to try to mimic the branding of Mac or MS but should stand out as being Ubuntu, did Linspire become a success because it looked just like windows?

But there are so many people coming from windows so you should try to make them feel at home is the usual response. Yeah Right, I did not move to Linux 6 years ago because I was so incredibly happy with windows and I can hardly imagine a happy Mac user to switch to Ubuntu because it looks so much like what he is used too, a cheap rip off it looks like.

The window buttons

Another much heard of complaint is the placement of the window buttons to the left. (Again, just like MacOSX)

Why? well.. its different, Yes….its how some other software company (That I will not mention again) does it, Yes….is it a useful change, No…is it confusing, Yes….does it look misplaced, Yes.

Every blog stopped complaining about it and started bugfixing, so I will repeat the steps here to be complete.

window buttons to the right

press — Alt F2 — type “gconf-editor” —- go to “apps” — go to “metacity” — go to “general” — right click “button_layout” — edit key — and change the value to “:minimize,maximize,close”

nuff said!

window title in the centre

This still leaves the window title on the far left instead of centred and it still looks ridiculous.

Open a terminal and issue the following command

sudo gedit /usr/share/themes/Ambiance/metacity-1/metacity-theme-1.xml

and find that part that looks like this

x=”10″
y=”(((height – title_height) / 2) `max` 0)+1″/>
x=”10″
y=”(((height – title_height) / 2) `max` 0)-1″/>
x=”9″
y=”(((height – title_height) / 2) `max` 0)”/>
x=”11″
y=”(((height – title_height) / 2) `max` 0)”/>
x=”10″
y=”(((height – title_height) / 2) `max` 0)”/>

 
x=”10″
y=”(((height – title_height) / 2) `max` 0)+1″/>
x=”10″
y=”(((height – title_height) / 2) `max` 0)-1″/>
x=”9″
y=”(((height – title_height) / 2) `max` 0)”/>
x=”11″
y=”(((height – title_height) / 2) `max` 0)”/>
x=”10″
y=”(((height – title_height) / 2) `max` 0)”/>

and change it to look like this.

x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)+1″/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)-1″/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>

 
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)+1″/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)-1″/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>
x=”((3 `max` (width-title_width)) / 2)”
y=”(((height – title_height) / 2) `max` 0)”/>

It doesn’t work with every application

Its kind of hard to make a theme that will shine and gives the best result for every application thinkable, this again has something to do something with maturity and how fast and if the designers can fix this for the new release.

Firefox for example gives a very unreadable result when the history results come up when typing in the address bar.

Skype gives an unreadable result when right clicking on a contact to see the available options.

I’m sure there will be other non-gnome programs…can the devs fix this before the release or will we be stuck with an unlucky theming choice for many otherwise useful programs the next 3 years?

Size matters

The last part of my rant is that the window buttons are really tiny, don’t get me wrong I’m not blind and they don’t have to be huge, but they are half the size of any other button available. Besides being a visual distortion it also leaves room for error since you do have to aim pretty actually to hit the option you want, this is doable during most circumstances but I noticed when wanting to close some windows when coming /home late at night after a few beers that it was damn right impossible to hit the right one.

the below post fixed it, just set the default 2 to 0

http://support.mozilla.com/tiki-view_forum_thread.php?comments_parentId=403836&forumId=1

EDIT::::::

OK, have been working on the BETA now for almost a full workday, and there is much to like.

Gnome is still Gnome since i used it about 5 years, some apps come and some apps go as usual with any release but i must say the whole system runs smoother then i expected and was used to, even with Compiz for the nice effects, something under the hood must be going in the right direction…….but…Gnome is also still Gnome when it comes to evolution-mail, it stinks!!!

Maybe its me but it keeps crashing, i think it crashed 6 times on me today and just as many times i simply had to shut it down because it was starting to completely occupy my SWAP, not healthy behaviour.

Ok, i do have 3 accounts setup and about 10GB of mail store on an IMAP server, but that should not give it to much problems, i use my mail a lot during the day, i keep switching between my mail app and applications that have just opened my attachment or about to create my attachment so a buggy mail app is simply not workable for me.

Evolution simply ruins my complete Gnome experience which i do like for the rest so at about 4 in the afternoon i had it, once more, with evolution and will not use it again till the next 5 years.

Have just installed the trustworthy and admirable Thunderbird once again, i feel relief…

Nevertheless i do have the problem that my network shares do not show up in Nautilus (or is it file-picker?) when handling attachments in Thunderird, the package thunderbird-gnome-support should fix that i thought?

I have not used Gnome in a while or Karmic with another desktop so i don’t know who to blame yet,but i must say my desktop feels more responsive opening individual programs then ever before, is it Karmic, is it Gnome or is it the really smooth default Compiz that even works on my modest onboard graphics card?

First thing to do is to get the restricted multi media going.

Installing the “ubuntu-restricted-package” will get us a long way.

This package depends on some commonly used packages in the Ubuntu
multiverse repository.

Installing this package will pull in support for MP3 playback and decoding,
support for various other audio formats (GStreamer plugins), Microsoft fonts,
Java runtime environment, Flash plugin, LAME (to create compressed audio
files), and DVD playback.

Please note that this does not install libdvdcss2, and will not let you play
encrypted DVDs. For more information, see

https://help.ubuntu.com/community/RestrictedFormats/PlayingDVDs

Please also note that packages from multiverse are restricted by copyright
or legal issues in some countries. See

http://www.ubuntu.com/ubuntu/licensing

for more information.

After restarting Firefox FLASH worked instantly, no problems checking out youtube.

MickeySoft fonts, no problem

MP3 and LAME i did not check, i hardly ever use that.

But, the java JRE did not work out of the box, i did hear something about dropped support or replacement, nevertheless you can still find it in the repos by installing “sun-java6-plugin” after the old and trusted buggy and slow java worked. I wonder how far the free java replacement is and if its somewhat usable by now.

Almost there with the codecs, just the medibuntu stuff left to go

sudo wget http://www.medibuntu.org/sources.list.d/`lsb_release -cs`.list –output-document=/etc/apt/sources.list.d/medibuntu.list && sudo apt-get -q update && sudo apt-get –yes -q –allow-unauthenticated install medibuntu-keyring && sudo apt-get -q update

And install w32codecs and libdvdcss2

The upstart graphics don’t seem to be completely ready yet if you want a smooth transition from grub to desktop but its a good start, still some console text flying around here and there.

Also i receive the message that the “computer will shut down in 60 seconds” when shutting down from the menu, this is either a bug or something that has to to with the delayed allocation of the ext4 file system bugs? i’m not sure.

Still plenty to to about fine tuning the personal preference and the bug fixing but the conclusion so far…not the best ubuntu beta i’ve ever seen, but it could become one of the better stable ones

The Beta version of Karmic Koala shows the same cd boot screen that we have been used for a while, so lets start with the option ¨Try Ubuntu without any change to your computer”

A new loading screen for the desktop CD appears and it looks quit minimal, not too bad,

although some text based messages about crypto flash quickly before before going to the silvery black loading bar.

The Gnome desktop looks about the same on the live cd apart from a new desktop image and a slightly new color brown.

Lets not waste too much time and start with the installation right away with the “Install Ubuntu 9.10″ icon on the desktop.

The welcome screen appears and we will use “English” for the installation process.

Next are the “time settings” and have already seen the new lay-out since Jaunty, looks good to me.

For the “Keyboard Layout” i will use the suggested  USA.

I usually keep my / and /home on a separate partition since i can’t be bothered making a backup in case of a reinstall, this means i choose the “Advanced Option”.

Yeah!!!! always makes me feel good to be an “advanced” user ;) .

I already had everything manually formatted as ext4 from my previous installation but it seems to be the default choice now. The /root will be used as ext4 and formatted and obviously /home as ext4 and not formatted.

The setup for the first user user on the system (the sudo account) looks familiar but there’s the GDM option to log in without a password and to require the password for the ecrypted home folder.

My previous installation did not have an encrypted /home folder, if i choose to encryption now on the existing partition containing my personal data should i have made a backup first? Well…lets not take too many changes and i still don’t feel to wait for a 30 minute backup to finish.

The next screen will show the options selected in the previous steps, so lets confirm this.

OOOHHHH!!! some flashy introduction appears during the installation, i must say Ubuntu is slowly shaping up to look like the, looking as slick as OSX, promise.

No GRUB confirmation? i think i noticed that the confirmation to install grub to the MBR already disappeared in the Jaunty version.

Restart the system or continue to use the live cd appears as usual, so lets give that a try.

The cd comes out as usual with the continue to restart after removing the cd

I did not time the time it took for the default installation since i have been writing this ans making the screen shots during installation but it should roughly be about the usual 20  minutes for the complete procedure.

As suspected after previous releases, even the BETA runs smoothly without any hick-ups and i must say it looks smooth

But with the new 9.10 in beta i’m willing to put it up to the test to see if i can switch to the standard Ubuntu for all my installed desktops (+- 12 of them) by the time of 10.04 .

While safely posting this from my Xubuntu Jaunty 9.04 eeepc netbook i will be keeping track of the preparation for my personal use case on a modest Pentium IV with 512MB RAM as my test installation

Roundcube downloaden

De laatste versie van Roundcube is te vinden op http://roundcube.net/downloads

Op het moment van schrijven is dat versie “roundcubemail-0.2.2.tar.gz”

Voor we Roundcube downloaden gaan we eerst naar een locatie waar we m willen hebben, en in mijn geval kies ik daar altijd de locatie /usr/src voor

# cd /usr/src/

vervolgens gebruiken we wget om roundcube te downloaden

# wget http://garr.dl.sourceforge.net/sourceforge/roundcubemail/roundcubemail-0.2.2.tar.gz

Roundcube zit verpakt in een tar.gz bestand en om deze uit te pakken gebruiken we tar -xvf

# tar -xvf roundcubemail-0.2.2.tar.gz

Het resultaat is een mapje met de naam roundcubemail-0.2.2 en om dit mapje op een mooie plek te krijgen met een duidelijke naam te geven doen we het volgende.

# mv roundcubemail-0.2.2 /var/www/webmail

MySQL voorbereiden

Er van uitgaande dat MySQL al geinstalleerd is gaan we eerst een database aanmaken voor Roundcube

(LET OP, MySQL gebruikers en de MySQL root account staan volledig los van de systeem gebruikers en root account)

Om in te loggen op de MySQL comando prompt als root doen we het volgende.

# mysql -u root -p

vervolgens zal MySQL het wachtwoord vragen van de MySQL root gebruiker.

Om de database “roundcubemail” aan te maken doen we het volgende.

mysql> CREATE DATABASE roundcubemail /*!40101 CHARACTER SET utf8 COLLATE utf8_general_ci */;

Vervolgens maken we een aparte gebruiker met de naam “roundcube” aan die gebruik kan maken van die database, LET OP!! verander het woord password hieronder in iets veiligers!!

mysql> GRANT ALL PRIVILEGES ON roundcubemail.* TO ’roundcube’@'localhost’ IDENTIFIED BY ‘password’;

Nu de database en gebruiker zijn toegevoegd gebruiken we het comando “quit” om weer uit de MySQL te komen.

mysql> quit

Vervolgens moeten we nog de tabellen e.d. importeren in de roundcubemail database en om dit te doen gaan we eerst naar het mapje webmail.

# cd /var/www/webmail

En vervolgens importeren we de tabellen

# mysql -u roundcube -p roundcubemail < SQL/mysql.initial.sql

Vervolgens zal MySQL vragen om het paswoord van de roundcube gebruiker, als het goed is is dat het “password” wat we eerder verandert hebben.

Roundcube configuratie

Roundcube is verder te configuren vanuit de browser mar voor we dat doen zorgen we dat de locatie temp en logs schrijfbaar zijn voor apache.

# chown -R www-data:www-data temp

# chown -R www-data:www-data logs

In je webbrowser ga je vervolgens naar de locatie

http://url-to-roundcube/installer/

Als alles OK is ga je door naar het volgende scherm met de knop ” NEXT”

Op het volgende scherm zijn een hoop opties te vinden, !!!!!vervang in ieder geval het goede MySQL password!!!!

Zodra alles OK is gaan we verder met “CREATE CONFIG”

en krijgen we het volgende scherm, dit is de configuratie met de waarden die we in het vorige scherm hebben ingevuld.

Zoals aangegeven moeten we die configuratie handmatig in het goede bestandje zetten.

de eerste is main.inc.php

selecteer alles uit dit gedeelte en doe copy met de muis, vervolgens gaan we we naar de comando prompt en maken we het bewuste bestandje aan.

# nano /var/www/webmail/config/main.inc.php

en pasten

CTR + X om te saven

Vervolgens de tweede db.inc.php copieren

# nano /var/www/webmail/config/db.inc.php

en pasten

CTR + X om te saven

Zodra dit gedaan is is roundcube klaar om te gebruiken maar eerst halen we de “installer” folder weg.

# rm -r /var/www/webmail/installer

Maak een link voor elke nieuwe gebruiker

In het geval van virtualmin maakt wordt er voor elke nieuwe klant een systeem gebruiker aangemaakt met ieder zijn eigen “public_html” folder, om te zorgen dat elk domeintje via http://mijndomein.nl/webmail toegang heeft tot webmail maken we een symlink aan in de skeleton files.

# mkdir /etc/skel/public_html

en vervolgens de symlink

# ln -s /var/www/webmail /etc/skel/public_html/

Apparently the man pages for Ubuntu were not to be found on the web to be read in your browser and the man pages website for debian manpages.debian.net cgi generates the pages meaning its not browsable by the search engines, so the chance of running in to them if you don’t know the exact address are next to none.

So.. i downloaded the work done for Ubuntu and adjusted the scripts and appearance to work with Debian, the result can be found here .

Not sure what i’m gonna do with this site but if there seems to be appeal for it then i will keep the site up and maybe extend it a little.

In case anybody has any idea’s on that i’d be glad the hear about it.