The LMDE3 installation I had done 18 months ago was in need of a refresh. This was based on Debian 9, so I thought Debian 10 would be a good place to start. Unfortunately, the newest release breaks my laser printer, so I've looked at continuing with Debian 9 (aka
oldstable) for the forseeable.
In the course of my efforts recently, I've tried:
- Debian 10 (breaks printing)
- Arch (couldn't install network drivers)
- Manjaro (failed to load OS on boot)
- CB++ (could not see network devices)
- Debian 9 + LXDE (working)
- Bunsenlabs Hydrogen (Debian 9) (initial install ok, just too bloated and complex)
- Bullseye (Debian 11) Alpha 1 with LXDE (working nicely)
Bullseye has the advantage of being
testing which is a combination of stability and new development. It is not clear to me the process of going from
stable in terms of the so-called freeze. The idea is to remain in
testing indefinitely, however that is possible.
Once one realizes the way the distributions are made, it seems to me to come down to getting close to a major distribution, so that one has a lot of options on software to install. The main choices are: Debian, Fedora, Arch. Fedora has a bit fewer options, in terms of applications available, than Debian and Arch. Ubuntu is derrivative to Debian, but its distribution cycle does have some advantages. However for someone taking responsibility for which repositories and how to use them, then it makes the most sense to stay close to the original source. Debian warns about mixing repositories and creating a franken-distribution, but if one sticks closely to
stable-backports and extends based on limited application needs (especially those with few dependences), then it should be possible to have a good experience.
There is a bit of work, because one does have to upgrade from one release to another, but it is not so bad because of
backports and taking place every 2 years or so. In any case Debian seems to be the primary distribution source of import.
When it comes to desktops, a minimal approach can be quite freeing. It seems that the newer, or ongoing distributions have issues regarding maturity/stability and resource requirements. So, projects such as Gnome, Cinnamon, KDE are bloated and require a bit of resources, and projects such as XCFE and Mate are a bit frozen in their development (perhaps just a personal preference). The LXDE desktop is a bit of fresh air, since development has largely stopped. Openbox, the window manager for LXDE is basically a finished product. This is the kind of platform one really needs to stop somewhat needless development. (Though there is some development needed which is neglected, such as newer monitor connection support.)
Openbox and LXDE
Openbox and LXDE are a great learning tool when used together. In order to configure or make changes, one learns where various settings are, and what applications control what functions. For example, changing the default filemanager from
nemo one also learns (besides various gnome settings and a symlink) how to install and configure a wallpaper-setting application (
hsetroot), as well as startup applications to run the wallpaper setup script on login.
Essentially, openbox and lxde were set up with simplicity and modularity in mind, unlike monolithic desktop environments. Even though these tools are a bit old and few resources are going into them, swapping out the various components is both possible and relatively straightforward. Along the way, one learns how the the linux desktop works in a hands-on approach.
Crunchbang no more
Just as when the crunchbang project ended, the same can be said of crunchbang successors (cb++ and bunsenlabs) that these projects are not better than vanilla Debian (plus one's choices of desktop components). For this, openbox and lxde (and perhaps lxqt) live on as useful tools.