Tomáš Pospíšek's Notizblock

Running Ubuntu

I'm a hibernating Debian developer but have been running Ubuntu on my personal machine since 2008.

Since then every upgrade to the next release has been a catastrophy. As a general rule the upgrade:

So an upgrade will take me around 1 1/2 days of work and will result in a system that has a lot of additional stuff running and installed that I don't understand, don't need and which is using up resources and finally the majoritiy of my polish, that made my machine work like I wanted it to, will be broken.

Why upgrade? Because old releases will start to be incompatible with current software and I won't be able to install software that I want or need for work.

Why not change distributions? Yes, I should. I think I will try to change back to Debian testing at some point. Last time I did my harddisk failed immediately after and I was lucky to be able to restore part of my world from the previous drive and some backups, which beamed me back into Ubuntu.

The assertion, that I should change back to Debian testing is interesting however. Why would that probably be quite an optimal choice?

Because for one Debian does not distinguish between "supported" and "unsupported" software - all software in the Debian repository is equally un/supported. It's up to the individual debian maintainer to keep up with the bug inflow and to maintain the package quality.

Ubuntu on the contrary has the "main" repository which is officially supported and universe and multiverse which are not. All of KDE is f.ex. not in "main". Both multiverse and universe are maintained mainly by the "Ubuntu Developers". There is no single person responsible.

The direction of Ubuntu is to some fuzzy part determined by a company, Canonical and a person, Mark Shuttleworth. As has been seen with Unity the goals are being determined by "visions", which are some big new ideas that are wanting to be realized.

There's not very much of such visions in Debian. Debian's realease goal is in major part to be as solid as possible and to keep upgrade problems as low as possible.

So I think, Ubuntu's goal is to lead by innovating. Stability is certainly not a higher priority.

Thus in the name of innovation, stuff will break. Software will disapear. It is possible that your settings, polish and configuration will self combust with the next upgrade.

I think this also means that Ubuntu's target audience are "endusers". People that will not look below the hood and change stuff. If you do that, that's fine, however it's not Ubuntu's goal to support you there.

On the contrary Debian rather targets the technically savvy person and will try to fix any and all possible configurations that are possible.

Let's see if these statements can be backed by anecdotical data from my last upgrade from 10.04 to 12.04.1.

Ubuntu's big goal was once to compete with Windows. Windows' target is not solely the "enduser". It's targeting developers, integrators, endusers, industry etc.

But I think that Window's application surface is much smaller and they have got those applications under much tighter control.

Windows' text editor is utter crap and has been utter crap for as long as it has existed. It doesn't change.

On the other hand Ubuntu's applications keep on changing a lot. That means breakage. And I don't think that's a strategy that will allow Ubuntu to compete with Windows on its traditional ground.

On the other hand Apple has extremely tight control over their platform. It seems to be quite reliable stability wise.

Tomáš Pospíšek, 2012-08-26

Articles