Here’s a little trick to have you’re gcc compiler create enum entries of 1 byte instead of the usual 4 bytes,
simply add :  -fshort-enums

to you’re compiler options, this makes you’re enums smaller, but does render you’re code incompatible with code  that wasn’t created using this option

As using the preprocessor often leads to bigger code, and very ugly errors it’s often adviced to replace it’s functions using C/C++.
Here’s a tip on how to replace the following  preprocessor command

#define VARIABLE 5

now this could easily be replaces by using a const but a const doesn’t behave the exact same as this define, as you could still for example obtain the adress of the const wich is something you don’t want happining that’s why you use

enum{VARIABLE = 5};

instead, as it’s just about the same as the define, except for the fact that the compiler can trace any mistakes, and smaller code is produced using it.

this is currently a pretty popular habit on fedora planet so I just thought I might tell you how to do it and post mine

open a terminal and type

history | awk '{a[$2]++ } END{for(i in a){print a[i] " " i}}'|sort -rn|head

now the terminal will output you’re most used commands, here are mine

146 sudo
73 make
64 bzr
55 cd
18 echo
13 ls
9 rar
6 ssh-keygen
5 sensors
4 xcompmgr

its mostly sudo that’s being used here, but the amount of bzr usages is kind of suprising as I don’t use it that much

 Warning ! : the following tips can reasonnably improve speed on most systems but not all systems are the same so only apply these techniques if you think you’ve got the skills to fix you’re system incase it goes wrong.

1. improve you’re grub settings

start up you’re PC and press ESC to enter the grub conf mode move to the second line  and press e
now add a space and profile to the end of the line.

the first time you’ll boot after this tweak it’ll take a little bit longer but it should be going faster after that

2.  improve youre harddisk preformance

open a terminal window and type the following
sudo gedit /etc/fstab

you’ll now see the file containing you’re file system info, find  the following lines
# /dev/sda8
UUID=461bc0aa-acb0-4526-8845-1cc268bcac3c /home           ext3    defaults            0       2
and append ,noatime,nodiratime after defaults

you can do this for every partition you have mounted in Ubuntu (so no doing this for the swap partition), it should make accessing files a little bit faster

3. use concurrent booting

Concurrent booting allows to take advantage of that dual(or more) core that’s in you’re pc during booting, to enable edit were going to open the following file

sudo gedit /etc/init.d/rc

now look for the line including CONCURRENCY=none and change it to CONCURRENCY=shell

In some systems this alone may do the trick but it can also mess up you’re HAL stuff so be sure to do the following

sudo mv /etc/rc2.d/S12hal /etc/rc2.d/S13hal
sudo mv /etc/rc3.d/S12hal /etc/rc3.d/S13hal
sudo mv /etc/rc4.d/S12hal /etc/rc4.d/S13hal
sudo mv /etc/rc5.d/S12hal /etc/rc5.d/S13hal

4. prelink

prelinking is a technique that makes linking the right libary’s faster and more efficient wich causes you’re programs to generally boot up faster too.  The only downside to it is that you’ll need to prelink again every time you add additinal software but that’s no big deal

first we’ll download prelink

sudo apt-get install prelink

next we’ll simply run it , the first time this might take some time but after that this will be a one second operation

sudo /etc/cron.daily/prelink

 5. get swiftfox

swiftfox is identitical to firefox only that it is compiled especially for you’re processer making a lot faster (you can compare it to the speed difference between ubuntu and gentoo and that’s a pretty big one)
so simply visit http://getswiftfox.com/debian.htm and install it for you’re processor type, and you’ll be noticing the speed difference pretty soon.

Something I’d never heard of before but it’s actually pretty easy all you have to do is insert a ANSI escape sequence in front of the line you want to print and it will be printed in you’re chosen color

example :

std::cout<<”33[1;94m”<<”hallo adriaan”;

33[ is the ansi escape sequence initializer

1; - specifies bold text

94 - specifies to print in the bright version of blue

m - ends the ansi sequence

to learn more coloring options checkout wiikipedia

Saw a nice linux wallpaper made by a user on the ubuntu-nl forum today, just thought it would be nice to share

wallpaper

As a computer science student at the KUL here in belgium, I’ve been learned to write the documentation first and then the code.

all documentation here at KUL must be written using stuff like @param and @invar in java. But as much as I enjoy java I’m much more a C++ person so after investigating for a while I found this thing called doxygen and it is pretty much identical to javado, the only thing that differs is the way the documentation is generated (using doxygen is pretty much always done through the terminal while javadoc can simply be accessed through eclipse), and in most of the cases the doxygen documentation seems to be allot prettier to.

So now  I’ve only got to find a good replacement for JUnit, but I don’t hink that really exists

Here’s are 2 programs I find to be extremely handy now that I’m working on hacking the Statistica 7 file format

1. a hex Editor : this thing is crucial as most text editors don’t come with the ability to view binary files, so getting a hex editor is pretty much a must. On Linux I’ve been using KHexEditor which is a pretty solid hex editor coming with an amazing amount of binary conversion options. The only thing I’m missing is the ability to use tabs. On windows I haven’t really found a similar free experience but 010 Editor is said to be really good stuff.

2. Binary File Compare program : either written by yourself or downloaded somewhere on the web this thing is kind of crucial as it makes you’re life a lot easier.
edit : Kari nicely pointed me out that for ubuntu there’s this very handy program called vbindiff- thanks alot
Because cracking a file format is pure reverse engineering, the easiest way to do it is to make two files with one small difference and then searching were that difference is located in the file format, so when you have a program that can check what’s different in the binary part you’re going to be going a lot faster.

So I’m still determined to get those damn Statistica files running under linux but because of there severly outdated sdk ther seems to be no way to get it working. That’s why I’ve come up with the brilliant idea of trying to hack those things myself as it should be possible. I mean the openoffice guys can get all those ms formats running in there office so why can’t I ??

There’s only one problem, I have no idea how to do this thing so I’m still pretty stuck on it for the moment, anyone who has some ideas feel free to let me know.

A commonly heard point of criticism on Ubuntu is the fact that it doesn’t have any easy backup tools that can be used by every sort of user no matter what their PC skills are. An attempt to fix this problem was made during the development of Ubuntu 6.10 with a program called hubbackup, but sadly enough this program didn’t make it into 6.10 or any later versions for a still unknown reason.

But a new attempt to solve this problem is in the works. The program is called TimeVault (for a full installation guide check here), and it seems to be doing it’s thing really good. where hubbackup still had you manually backup stuff with little room for configuration, Time Vault does al those things for you (monitoring files + auto backing up) and even integrates into nautilus, something I personally really like.

While the integration with nautilus may seem amazing it also creates a big disadvantage, it’s only usable in a gnome environment so KDE people will have to look out for another program to do their auto backing up