Background

Many years ago I had just barely learned about Linux. Some friends at Cistron, then one of the first Dutch companies to offer internet access through UUCP, had convinced me to install Linux on one of my own computers and I immediately felt ‘welcome’ in the then much smaller ‘community’. Coming from MSDOS where I used some tools rather frequently, I really missed my Norton Commander. So when I was told that there was something similar on Linux, called ‘Midnight Commander’, I immediately had to try it out. Sadly, at the time Midnight Commander or ‘mc’ only operated on files, not on whole directories.

Midnight Commander

This is where I ‘joined’ the development, using some of my own C coding skills to add directory support to mc and sending a patch to Miguel d’Icaza. My patch was used, praised at first, and is probably long been overwritten by far more efficient code than my first code, but the fact remains that this made me one of the ‘official license holders’ for mc. What was much more important to me however was that I learned a great deal of technical knowledge about C coding on Linux. Currently being somewhat settled in my ways after having worked for large corporations for years as a freelancer, I am looking to revitalize some of those skills that are at the basis of my current position because I sometimes feel I’ve drifted away from that a bit too much, getting complacent about my own ‘expertise’ at times.

Finding the Right project

So I decided to search for a project to join and found out that it’s no longer very easy to find such a project. I’m very interested in virtualization, both from a personal as a professional perspective. So I decided to check out the KVM project. While this is a very interesting project, I also found that it’s incredibly complex and there seems to be very little community spirit. I get the impression that KVM is largely an IBM incentive and one needs to have very deep technical skills or you’re simply not welcome. I did not feel encouraged to help with the project and believe that at this point in time it’s too complex for me. In the same context of virtualization, I also checked out the libvirt project, a management and visualisation tool for virtualization. While I felt much more welcome while talking to some coders on IRC and got good feedback on questions on the mailinglist, also in the libvirt project I found a decidedly corporate push in this project, this time by Redhat. It’s difficult to join an open source project when the developpers seem to not ‘be there’ in the weekend on IRC and emails get answered ‘from 9 to 5 on weekdays’. I may contribute some more to libvirt at some point, I’ve done some testing, reported a bug that got analyzed and discussed on IRC and fixed by a small patch I made, but I don’t think it’s a project I’d really put a lot of effort into.

MPLAYER

Again, for personal and professional interests, I also tried to join the mplayer project. I had a movie I wanted to view on my ‘Home Server’ and it did not want to play, so naturally I started asking questions on IRC. I noticed that at least the mplayer community on IRC is far more of a hobby effort by the people involved. One gets this ‘bazaar’ kind of feeling of the past. I helped fix a bug, sent a patch to the mailinglist, but started wondering to myself if playing movies is, well, something that really interests me from a technical point of view.

EFL

One of the projects from ‘way back in the mc days’ is an interesting one; Enlightenment. It is interesting in my opinion because of it’s history. Enlightenment ‘was’ a window manager very much ahead of it’s time in the early open source days. Some called it ‘just eye candy that burns a lot of CPU’, others thought it was brilliant. Either way, the corporatized open source world (sorry that’s how I sometimes feel these days), did not do a lot with Enlightenment but to use some good parts for the GNOME desktop. Interestingly the GNOME desktop was being pushed very well by the main coder of mc. The GNOME desktop pretty much overtook the Linux desktop with the only real serious competition being KDE. However, Enlightenment is still very much ‘alive’ and has been in hiding for nearly 9 years at the time of writing this. All this time, a very dedicated group of people have been working on completely redoing the whole basis of Enlightenment, turning it into a two-fold effort; The creation of a big set of graphical libraries, the EFL or Enlightenment Foundation Libraries, and the creation of Enlightenment on top of that. I’ve checked into the source code, the current state, and I believe Enlightenment and the EFL will emerge very strongly and I’m very much interested in participating, nay ‘helping’ in this project. I believe the project has a strong open source ‘spirit’ to it.

VUURMUUR

While working on my ‘Home Server’ concept, I decided it was time to update my good old firewall shell script and started looking around for a ‘good firewalling application’. I found many half-assed GTK+ attempts (apologies to those that feel offended) with very few extra things to it than ‘having a GUI’. I found a few incredibly elaborate command-line tools, hats off to the people that made those. But then I bumped into a tool that had a nice ncurses interface and it zoomed me back in time to the ‘mc days’. With internet security and forensics being one of my longest held interests, I immediately decided to find out more about ‘vuurmuur’.

I’ve so far found vuurmuur to be the ‘perfect project’ for me.

• It’s pretty stable and useful in it’s current (0.8b2) version
• There are many opportunities to improve it by new features
• There are chunks in the code that could be done different/better
• It has a very small group of relatively active coders some of whom also happen to be Dutch

So at this point I’ve taken on trying to implement something that I had never even heard about before, called ‘inotify‘. I will document my efforts on my Wiki server, mostly supporting my own memory but also hoping it will be useful to somebody eventually.