I have now decided to go away from Foresight Linux. Why that? Well first of all after my FreeBSD experiences I got back to Foresight Linux Alpha 2. I knew that it worked mostly and also I had to have something that worked for a presentation. And that it did ok. To more detail please read the link above.
I dont really have statistics about OpenBSD – I would be interested in those – but from my very subjective impression if I do “man xy” OpenBSD mostly gives me doc that I understand – and I like that. Foresight for example is worst when it comes to manuals – which is also due to the fact that it only installs what is needed. Which is or can be a plus – But like with “man mplayer” I like my system to just have that.
I also think maybe its because I am getting older that I am seeking different things. My Linux history started with SuSE Linux, which was really ugly with all that Yast stuff – but also due to Linux being much more in an early stage. back then 1998 – you had to build your own kernel to include the sound drivers. Sound was a real issue. I then had to execute isapnp. I cant say I loved to build kernels but I did it often.
Today on Linux mainstream kernels are mostly accepted. I still see people hacking on kernels on new releases activating this and that. I can see that it can be fun and also we need people testing stuff. But I havent built a new kernel since about 10 years – and never had the need. I am also never too excited about new kernel releases.
I have also followed GNOME for a long while, because I wanted to have the one desktop the felt the best – and as an old Mac user GNOME felt very much GNOME like. The nautilus file manager later was also developed by some old Mac folks at Ximian.
After SuSE I used Debian and that was much nicer than SuSE. But there was this Debian dilemma: Stable was often very outdated – so people told me: use testing – but this wasnt so stable – and I even followed the advice of some people to use unstable – but this was really bad. I also used Debian on a web server. It had Apache 1.3 but I wanted Apache 2 – so there was something called backports. I installed that and erased all my site configs. Also the dpkg database got trashed often. I could count endless issues about “stable” Debian.
After Debian I took a wuick look at Red Hat 8 – but updates where a pain (you only got updates if you subscribed to channels) – so I gave Gentoo a try. What was wrong with that? Well, all in all it was too complicated to look at all the USE flags and things. So I switched to Fedora 1.0 as this came out. And Fedora really was great as it resolved all that made using Red Hat a pain. So I was a loyal Fedora user for some years.
But then they made some evil hacks on the community – and also it crashed on my notebook all the time without anybody seeming to care. The rest of the story can be read in “It’s final: Good bye Fedora!”
And now yet another switch. Foresight had MANY good ideas – but I think they made too much wrong choices. On too many cross-ways they walked in the wrong direction. Which is kind of sad, because it had very good starting conditions and also had Conary as the only desktop distro on its side. But conditions where not always best and for the past months it wasnt so much about being creative with new ideas but struggling with all kinds of dependencies. Developers really invested a lot. But as I always said – a distribution is for me never more than the best compromise. I am open if its more to me – but I really dont like all that common distro marketing thing. There is this blog about passionate users that tries to bring in some fresh ideas. But I think the whole idea to “create” passionate users is false. You cant do that unless you lie. Otherwise – if they suggest that software should suck less I would agree – and also I would agree on the message that software only needs to be good enough. The thing about blogs like the one I just mentioned is that it all sounds good and looks good – but I can not connect to it from my user perspective.
Currently I look more forward in a nice base system which does its job good enough and which doesnt do toomuch flip flopping and breaking. I like to be able to boot my machine every day wihtout the fear that it wont boot. This means less updates is better. On Linux updates have become some kind of mysticism – even when it comes to security one can assume that for 99% of the users the bugs and issue are irrelevant. What is more relevant is that an update works and doesnt break anything. What also matters tome is that fixes are available but I am not forced to apply them . I also like distros to stick to some basic promises like deliver a new stable version each so and so month. I also like new features – not to the cost of stability – bu generally I think its nice if my desktop can do new things – especially if they are useful. I also like it when things get fixed without me noticing it – so I do like fixes under the hood.
What about statistics? Well some people like to play with them and maybe could show me that openBSD is not the fastest or whatever. I dont really care – first I cant really check that data without a lot of investigations and secondly often the conclusions are false. So I think for many cases it doesnt matter if you use Postgres or Mysql – or what file system you use, etc. – But many techies tend to look at the latest trends and propagate one technology over the other. There may be issues in some technologies but you really need somebody with a lot of experience to really tell the difference, because statistics alone seem to suggest soemthing to be true although its irrelevant in practice .