OpenSuSE 11.2 and Lotus Notes 8.5

Recently I have migrated from Ubuntu to OpenSuSE due to my work requirements.

Since OpenSuSE is an RPM-based distro, I thought that Lotus Notes 8.5 would be easier to install. I was mistaken.

Lotus Notes 8.5 depends on some libraries that were updated on OpenSuSE 11.2 causing some panels to appears blank. After a weekend debuging I was able to isolate the required libraries.

Lotus Notes 8.5 Blank Panel

The trick is that Lotus Notes requires the following libraries from OpenSuSE 11.1 (note that I am running a 64-bit version, so your required library may be the default one):

• gtk2-32bit
• pango-32bit

Go to the OpenSuSE 11.1 repository and grab the packages but do not install. If you install it will break the GTK applications on your desktop.

The trick here is to use these libraries only for Notes. To achieve these we must extract the RPM in a temporary dir, move the lib file to a isolate path and create a wrapper for the Notes to use LD_DIRECTORY_PATH.

So, as a root, do the following:

# mkdir /tmp/rpmsextracted
# cd /tmp/rpmsextracted
# rpm2cpio <PATH_WHERE_DOWNLOADED_RPMS>/gtk2-32bit*.rpm | cpio -idmv
# rpm2cpio <PATH_WHERE_DOWNLOADED_RPMS>/pango-32bit*.rpm | cpio -idvm
# mkdir /opt/ibm/ibm-lotus-notes-compat
# mv /tmp/rpmsextracted/usr/lib /opt/ibm/ibm-lotus-notes-compat
# chmod 755 /opt/ibm/ibm-lotus-notes-compat/lib
# chown -R root:root /opt/ibm/ibm-lotus-notes-compat
# mv /opt/ibm/lotus/notes/notes /opt/ibm/lotus/notes/notes.real

Now you have the libraries on /opt/ibm/ibm-lotus-notes-compat, so lets create a wrapper. Create a new file /opt/ibm/lotus/notes/notes with execute permissions (chmod 755 /opt/ibm/lotus/notes/notes) and the following content:

#!/bin/bash

NOTESCOMPAT=/opt/ibm/ibm-lotus-notes-compat/lib
export LD_LIBRARY_PATH=$NOTESCOMPAT:/opt/ibm/lotus/notes:$LD_LIBRARY_PATH
echo $LD_LIBRARY_PATH
/opt/ibm/lotus/notes/notes.real $@

That’s it. Now runs Notes and see if everything is working as expected.

Lotus Notes 8.5 Working as Expected

Final Note: You must have other gtk libraries installed as well. If you are running 64-bits, make sure that you have the 32-bits libs installed.

Ubuntu 8.10 and the Sony Ericsson MD300 modem

Since the 3G network is growing fast here in Brazil I decided to give it a try. My carrier offered me a Sony Ericsson MD300 modem as part of the service. Then I had the mission to make it work under Linux, in specific the Ubuntu 8.10 (my current distribution).

The MD300 modem is not officially supported but as I will show bellow, it is quite well compatbile with the Ubuntu 8.10. This USB device comes with 3 different operation modes. The defaul mode is the 1, which the device acts like a flash drive and of course we are not interested. Then we have to tell the Linux to switch to the mode 3 (serial modem) when the new device is plugged. To do so, we create an udev rule. It is quite easy.

Create a udev rule’s file (on the /etc/udev/rules.d) for the device

sudo gedit /etc/udev/rules.d/50-md300modem.rules

Then put the following content on the rule’s file:

ACTION!=”add”, GOTO=”3G_End”
BUS==”usb”, SYSFS{idProduct}==”d0cf”, SYSFS{idVendor}==”0fce”, NAME=”%k”,SYMLINK=”modem3G-%n”, PROGRAM=”/bin/sh -c ‘echo 3 > /sys/%p/device/bConfigurationValue’”
LABEL=”3G_End”

Be careful with the like breaks. The text above has only 3 lines.

After those steps, reboot your machine so the new rules can be applied and connect your modem at an USB port. Now it should be working and two devices must have appeader (/dev/ttyACM0 and /dev/ttyACM1)

YABP: MainStream

Yet another band project is starting. The MainStream (no idea where the name came from) is a band from some guys from my work. They recorded a demo with 2 musics, although it is not finishd I really enjoyed it. In my humble opinion, they are trying to make the type of music that is pop without forgetting the roots of metal and rock. I wish a lot of lucky on this project!

If you wanna listen the recorded music, just go to the Andre’s Blog and download them.

e1000e driver is back on Intrepid

Since my last kernel update the driver e1000e was not blacklist anymore. I am not sure if it is the final solution. This is the bug fix solution (from kernel.org):

Set the hardware to ignore all write/erase cycles to the GbE region in the ICHx NVM. This feature can be disabled by the WriteProtectNVM module parameter (enabled by default) only after a hardware reset, but the machine must be power cycled before trying to enable writes.

Although the e1000e was unblacklisted I had to modprobe it again to work on my Ubuntu Intrepid. On a terminal:

sudo modprobe e1000e

Shockwave Theory applied on Traffic Jams

Japanese Scientists executed an experiment to demonstrate that some Traffic Jams can appear from nowhere due the shockwave principle. It’s amazing how the variation of speed may induce a complete stop on the traffic.

XBMC, the solution for a media center

Last week Murillo told me about the XBMC project and since my media center was lacking of a good software to browse my collection I decided to install it. My set up is composed by a Mac Mini running Mac OS X. Front Row is terrible, it does not support a lot of formats and some media codecs are very slow.

My first impression on XBMC was very good. The interface is totally eye-candy with support for skins and other customizations. The support for file formats and codecs are so extensive that if I had to describe it here I would spend a lot of time . But, for my purpose it only have to play mkv (matroska) in 720p smoothly, which it did flawless.

Also, the support for plugins and scripts enables the XBMC to play YouTube videos, download lyrics and subtitles, and much more. And the best thing is that the plugins are written in python!

Of course the XMBC has some issues since it is on a beta version. I witnessed some freezes on playback and on some scripts. Also, the navigation using mouse is not fully operational (at least on the skins available). Some subtitles don’t work very well and are not displayed.

Check out the project at http://xbmc.org/ .

Video: Airtraffic density over 24h, just crazy!

This video on YouTube shows the air crafts on the world over a period of 24 hours.

It is amazing and terrifying. The amount of information required to manage all those air crafts is so huge, that even the smallest mistakes may lead to huge accidents.

Flash High CPU Usage on Intrepid: a better solution

There is a better alternative to solve the Flash High CPU usage on the Ubuntu Intrepid 8.10, using a possible official patch. This patch will generate a new .deb that can be installed and removed using the Debian’s Package Manager. To quickly install just follow the commands below on a Terminal (Applications -> Accessories -> Terminal).

sudo apt-get install build-essential fakeroot devscripts
cd /tmp
wget http://launchpadlibrarian.net/17178225/flashplugin-nonfree_10.0.1.218%2B10.0.0.569ubuntu1.debdiff
apt-get source flashplugin-nonfree
sudo apt-get build-dep flashplugin-nonfree
cd flashplugin-nonfree-10.0.1.218+10.0.0.525ubuntu1
patch -p1 < ../flashplugin-nonfree_10.0.1.218+10.0.0.569ubuntu1.debdiff
debuild -uc -us
sudo dpkg -i ../flashplugin-nonfree_10.0.1.218+10.0.0.569ubuntu1_i386.deb

This will generate a .deb and install it.

Update: patch was misspelled. Thanks Doug!

e1000e banned on 2.6.27-4

As a desperated measure, the maintainers of Ubuntu Kernel decided to put the e1000e driver on the blacklist. It is a temporary measure to avoid more problems, but also disabled ethernet on a huge number o machines (including mine ).

I hope this issue will be solved as soon as possible.

Intrepid ships hardware-blower driver

In the Ubuntu 8.10 Intrepid Alpha, the kernel used is the 2.6.27 which contains a very dangerous version of the e1000e driver. This specific version of the driver maps the firmware of the hardware on memory and gives write access to it, so a misbehaved kernel (which is common in a development branch) can overwrite the firmware.

The attempts to recover the firmware using the Intel’s tools just make the things worse and brick the entire network card (the card is not recognized as PCI device anymore).

In other words, the driver allows a sequence of events to destroy the network card, which in many cases is onboard or at a laptop (there is already some occurrences of this fact).

You can track the bug on LaunchPad.