I always said I don't need Compiz and the like, I don't want useless eye-candy, and generally, I hate when efforts are directed nowhere, and a distro is praised or criticized by how it performs with regards to Compiz.

Brian Profitt tries to put things differently ("maybe it's useless, but who cares?"), in Eye Candy: So Bad, Yet So Good: «Many of the Compiz tools are simply there for fun; to expect otherwise is silly. If something productive can be done with them, then great. But sometimes we all need a little distraction, something to keep our minds fresh and our morale up. [...] So when someone says, that stuff is just a waste of time, tell them (like I have told myself) to lighten up and learn to play a little.»

Well, I would rather play with something I want to play, not with a creation I never asked for and they want me to play with!

I said I don't like Firefox 3.0 for various reasons, none of them being the "quality" of Firefox 2.0, of course. But the "awesome" bar was so buggy the last time I checked it, that I've found it too disruptive to worth the switch. Besides, Firefox 3.0 is not released yet, and I have no intention to use a non-released product from an organization not known for its high-quality products. No matter RHEL 5.2 comes with Firefox 3.0 Beta5, I follow "the other leaders" who are still on 2.0.0.14 (stable/released): Slackware, Debian...

A recent article not liking it, Firefox 3 'awesome bar' not awesome for all, has a striking comment at #2: «"Change is but inevitable..." - Don't know who said that, but well said. Some people really seem to have a problem with progress and they want to be stuck up in time. I guess if a survey is carried out, 95% of the people will vote in favour of the 'awesome bar'. The niggling 5% are the ones who think their knowledge base is good enough and there is nothing left in this world to learn or improve upon.»

Things are not that simple. If it ain't broken, don't fix it. You have "something" (the address bar) that is supposed to do "some task" (serve as a place to enter the URL to go to), and possibly to have a few "extra features" (such as keeping a history). If some people request this "something" to be "even smarter", this can be a legitimate request and a productivity enhancer, or just a way to more bloat. Generally, this is how bloat is generated. To rephrase Wirth's law, people are requesting for bloat and this bloat is delivered faster than hardware gets faster.

Thinking that "there is nothing left in this world" to discover or improve was a sufficient attitude in the scientific community around 1900. And they were definitely wrong. We shouldn't generalize though, as not all the requests for stability and consistency are retrograde views of people who "want to be stuck up in time". As I use to say, the wheel is still round and the Phillips screwdriver is still crossheaded: should we change them?

UPDATE: I decided yesterday that I will only switch to FF3 when all of the following conditions are met:

(1) FF3 is released;

(2) Support for FF2 is discontinued;

(3) FF2 has a discovered vulnerability, and the only fix is to move to FF3.

(4) Patrick replaces FF2 with FF3 in Slackware 12.1 (not -current).

This is what I call "aiming for stability".

I have constantly to deal in my RSS subscription with enthusiastic blog posts about how Plasmoids are a heavenly manna to the humankind. I usually manage to ignore them, but I am terribly frustrated to see how so many developers are focusing on things of so meager importance. (Here's the latest I've read: Returning to KDE programming.)

I wonder what I'll be using in Linux in a couple of years, when GNOME will be either fully Mono-ized, or totally disorganized (it's not coherent in development and it never was), when KDE3 will be dead and unmaintained (not even in LTS distros), and when KDE4 will be the only "decent" choice — but less decent in look & feel than any closed-source OS available at the time.

Another problem is that I should definitely get rid of GNOME somehow — I am still mad at those folks who kicked me out from their Bugzilla. This prevents me from having any "good feelings" for them, and leaves me with nothing else but the "consumer" attitude.

Or, a consumer simply makes choices based on what works for him or her, not on any "loyalty" or "moral principles". (Except that I can't use any Novell product for "moral objections".) If I don't feel GNOME is having the right karma, then maybe it really doesn't. After all, Slackware doesn't include it, right?

I tried to switch to XFCE, and I don't feel I like it that much. I am sensitive to some small design glitches. I am sensitive to them lacking leadership and being so sufficient not to admit it. XFCE's developers could have been developing some BSD flavor, and no one would have noticed the difference (except that some FreeBSD commiter called me names, whereas the XFCE devs have not... yet).

LXDE is unfortunately not good enough when comes to the desktop folder itself: the icons are simply displayed in positions you can't change, and you can't drag and drop anything to or from Desktop. The desktop has a "special treatment" by PCManFM, but this "treatment" is really annoying. I am not saying that LXDE and PCManFM are not marvelous pieces of work, I am saying I don't feel like adopting them. XFCE has also limitations with regards to the desktop as a folder, but not that restrictive though.

This leaves me with what exactly? No KDE4, no more GNOME, no XFCE, no LXDE... then KDE 3.5.9, which is on its road to death?!

Practically, if you notice that Slackware 8.1 and 9.0 are still maintained with regards to security patches (5-6 years later!), you can assume that Slackware 12.1 (most likely the last release to include KDE3) will be supported in 2013, which means it can be an option to use KDE 3.5.9 for years to come.

RHEL5 and its clones might also be an option to run KDE 3.5.4 for years — but 3.5.4...

I am still clueless, and I have even less valid choices than a typical Windows user, who should actually stay with XP if his business-critical applications don't run under Vista, or go with Vista for a light home usage. But with Linux, you simply can't plan for the future...

NOTE: ln -s /dev/null /dev/clue does not originate here, but actually here and here.

P.S.: But I still have a good deal of clues, as opposed to the many who must have believed the Distrowatch announcement for Debian GNU/Linux 5.0 Beta 2. Obviously, it was about a beta of the Debian Installer, but nobody in the Debian blogosphere noticed, except for Gunnar Wolf, yesterday.

Now I know what is the public understanding of the Linux world: ZERO. "Community" should be replaced by "Brownian motion" to be more accurate.