• Rising: Ubuntu, Fedora, PCLinuxOS, Linspire, Xandros.
• Staying the Same: MEPIS, SUSE.
• Falling: Debian, Mandriva.

• Ubuntu is not actually raising, it's only... being popular (as usual), even if it doesn't deserve this popularity. Jordan: «Canonical Ltd., the foundation set up by South African millionaire Mark Shuttleworth, is fanatically committed to its improvement.» C'mon. Grow up, man. Get a life. (There is a problem with some American people. They love to say "aggressively somehow", "fanatically somehow".)
• Fedora: what's that special in Fedora? Jordan: «Fedora 7, slated to be released in the spring, promises to be Red Hat's best ever.» Isn't ALWAYS the latest release supposed to be "best ever", no matter what product you're talking about? And what does it mean "ever"? Is Fedora 8 going to be worse than 7? Jordan: «Fedora has become the workhorse of the servers and desktops.» Nay, I wouldn't install Fedora on a server, and I won't use it in corporate environments. It is not really SUPPORTED. It has a short lifespan too. If there is something with a bright future in the enterprise, that's RHEL5 and CentOS5. Fedora is a test bed, and each release has important bugs, period.
• PCLinuxOS had a very bad last half of 2006. I almost thought 0.94 (aka 2007 now) will never release! Too many various "0.93 Whatever Edition". PCLOS is unpredictable when comes to releases. And, you know, PCLOS is not Debian. Not recommended (although the beta of 2007 features a fabulous rebranding).
• Linspire and Xandros: these are "pervert" distros, trying to be so very much like Windows. Well, Freespire is free, however this development model is not trustworthy. I don't see any future in them.
• MEPIS is not "staying the same", it is going down, after a considerable lost of identity (what makes it better than Kubuntu?).
• SUSE: nobody knows what will happen to SUSE.
• Debian: yes, it's about "fundamentalism and intransigence" lately. However, is it really falling?! Who cares that Etch was not released?
• Mandriva. This is the worst judgment of Jordan. Hey, Mandriva is not falling, by no means! Mandriva 2006 was the best release in years, and the One editions — both KDE and GNOME —, plus the focus on AIGLX, XGL and Matisse, make it very interesting to follow! Oh, Jordan, have you ever heard of Mandriva Corporate Server? (No, you have not.)

  [Français]  [Deutsch]  [Nederlands]  [Español]  [Português]  [Italiano]

     Patrick @ Feb 02, 2007 (07:16 AM)

With regards to Fedora, I think FC7 will vastly superiror to any release they have put out in the past. Most of the praise that FC7 is receiving (and will recieve) is really only merited when comparing Fedora to Fedora. The merge of Core and Extras, plus that huge list that you posted a month or so ago will mean a much more user friendly distribution. Also, they have increased the support cycle for Fedora a bit, so should make it more interesting for perspective users. I like the fact that Fedora is resonding to user input on what they want. While Fedora may be a test bed for Red Hat, there are hundreds of thousands of people who use this on a daily basis. Without them, RHEL would not be what it is today (or what it will be on 28 February).

     Béranger @ Feb 02, 2007 (07:26 AM)

The merge of Core and Extras is the most interesting part to me.

However, when you have a distro that features very "release-specific bugs" -- twice a year! --, you wouldn't use it in a production environment.

Of course, with a bad FC4 and an unimpressive FC6, but some good FC3 and FC5, it's time for FC7 to shine, following the "even is evil, odd is good" scheme :-)

Yet I still expect some serious bugs. It's not easy to release twice a year *and* to have such ambitious objectives like with Fedora 7!

I don't favor the Debian release cycle model, but I can't feel comfortable with Fedora.

     xabbott @ Feb 02, 2007 (07:42 AM)

Ubuntu will continue to rise, the community behind it is strong and people love having a leader figure.

PCLinuxOS is much better at doing what Mandriva and Unbuntu are trying to do but has poor marketing.

Linspire, I think it will be growing. They understand what needs to be done for widespread adoption.

Mepis, agree with you. Suse, agree with you.

Debian, This is one of the rocks of Linux that will continue for a long time.

Mandriva, bad. This last live disc with Metisse was horrible in my opinion. The fact that they would include Metisse and tout it as one of the main features borders on delusional.

     Béranger @ Feb 02, 2007 (08:18 AM)

> people love having a leader figure.

Stalin?

> PCLinuxOS is much better ... but has poor marketing.

Poor communication, agreed. And no clear *vision* for the future. It looks like it's for the gamers.

> Linspire ... They understand what needs to be done for widespread adoption.

Huh?!

> Mandriva, bad. This last live disc with Metisse was horrible...

Yes, it didn't work with me either. But it has a lot of potential.

     Anonymous Coward @ Feb 02, 2007 (09:16 AM)

Hmm.

Fedora 7 isn't out yet --> Vaporware

I find Distrowatch's statistical rankings for 2006 based on hits per day much more noteworthy:
!As we can see, Elive was the fastest climber in terms of number of places it rose during the year (+71), while openSUSE won by the number of visits - on average 576 more people per day viewed the openSUSE page in 2006 than in 2005. And the biggest losers? White Box Enterprise Linux (down by 55 places) and Mandriva Linux (down by 649 views per day).! (http://distrowatch.com/weekly.php?issue=20070101)

Rankings:

1 openSUSE +576
2 PCLinuxOS +348
3 SabayonLinux +285
4 Fedora Core +233
5 Zenwalk Linux +183
... ... ...
63 Yoper Linux -94
64 Xandros Desktop -94
65 MEPIS Linux -100
66 KNOPPIX -215
67 Mandriva Linux -649

     Béranger @ Feb 02, 2007 (09:28 AM)

I am afraid Distrowatch ratings are IRRELEVANT.

Like I said in my predictions [ http://beranger.org/index.php?article=2189 ], what will matter in 2007, in no particular order, include:
-- Etch
-- FreeBSD 6.2
-- PC-BSD
-- RHEL5
-- CentOS5
-- FC7
-- Mandriva 2007.1
-- PCLOS 0.94/2007
-- Yoper
-- Sidux
-- SabayonLinux.
-- Pardus 2007

I don't care who clicks what on Distrowatch.

     Anonymous Coward @ Feb 02, 2007 (10:00 AM)

More aptly put, what you would LIKE to matter in 2007 are those...

Mattering...
-- Etch - only for a chosen few
-- FreeBSD - likewise
-- PC-BSD - likewise
-- RHEL5 - staying solid, but not moving
-- CentOS5 - niche software again
-- FC7 - maybe, maybe not
-- Mandriva 2007.1 - no
-- PCLOS 0.94/2007 - no
-- Yoper - no
-- Sidux - no
-- SabayonLinux - this one will
-- Pardus 2007 - only in Turkey...
-- SLED 10 - will make a huge, huge impact (see Peugeot as a harbinger)
-- openSUSE - 10.2 was a really great and increasingly popular release, and nothing indicates that 10.3 will be any less so; climbing quickly and steadily.

     Béranger @ Feb 02, 2007 (10:13 AM)

Oh, those Germans...

You shouldn't be proud of SUSE, as it's not a German product anymore. It's Yankee!

I guess Peugeot has bought the idea of an "easier migration from Windows to SUSE". Hovsepian has won. Or maybe they have IBM servers -- IBM was always "pro-SuSE".

     xabbott @ Feb 02, 2007 (10:36 AM)

I agree with Anonymous, what you want to matter and what will matter are far removed. That said Distrowatch isn't any better.

Some Distros like Sabayon only appear popular because it includes Beryl by default. So you have general geeks downloading it simply to see the desktop. They won't use it long and as more distros include similar features it will fizzle.

Also certain distros just have press releases every week to ensure a Distrowatch standing.

As for my "leader comment" earlier. Despite your personal feelings, I do think one of Ubuntu's strengths is Mark himself. It's very similar to Apple and Steve Jobs.

Let it be known I personally use Arch. :P

     Béranger @ Feb 02, 2007 (10:43 AM)

> I do think one of Ubuntu's strengths is Mark himself.
> It's very similar to Apple and Steve Jobs.

Right. And Microsoft and Bill Gates. (Including the wealth part.)

> Let it be known I personally use Arch. :P

Nice logo. Large repository. Important community. Awful pacman.

     Patrick @ Feb 02, 2007 (02:43 PM)

>Fedora 7 isn't out yet --> Vaporware

Well, Test 1 is out. Actually, I am a grad student at Boston University, and FUDCon 2007 Boston is currently going on in the building just behind my office. I just got out the seminar on Fedora 7. It is disingenious to call it vaporware because they are hard at work on it. In fact, there will be a lot of significant changes to the underlying system and build process that should increase quality assurance for this and future releases. I don't think they would admit this, but I would bet that Fedora 7 will be delayed. These are major changes, and there will be problems along the way. Although, a delay might not be that bad. The kernel maintainer said that the tickless kernel should be 2.6.21, and FC6 is currently at 2.6.19-1. This new kernel should be welcomed by all. Fedora 7 may not be final, but it is far from being vaporware, which implies that it will never be released.

     Béranger @ Feb 02, 2007 (06:23 PM)

> the tickless kernel should be 2.6.21

I have been reading abou the tickless kernel some time ago, yet I am not sure about what the advantages will be. Could you point us to some recent articles/papers/posts that would enlighten the audience on the subject? Thx.

     Patrick @ Feb 02, 2007 (09:00 PM)

As described by the people at FUDCon, the average kernel wakes up about 250 times per second (they said it could be more based upon the program in question). The kernel is asking if there is anything that it should be doing. For instance, is there a new blue tooth device connected? All of this constant searching by the kernel is just wasted energy when no new devices are being connected. The tickless kernel will not have all these constant "wake ups". The tickless kernel will respond only when a new device is connected. They did not go into details as to how this will be done, but it will do away with useless fonctions. Most importantly, this will reduce the power used by the computer. Batteries on laptops will now last longer. Servers will emit less heat. As an end result, less power will be wasted. I don't know of any papers, but will post anything once I can find it.

     Patrick @ Feb 02, 2007 (09:13 PM)

Here is a link for some info on a tickless kernel:
http://newkerneltrap.osuosl.org/node/6750

     Béranger @ Feb 03, 2007 (02:07 AM)

Thx. What I would like to know is whether this concept of "ticklessness" is something new to Linux or new to everyone. Are there tickless kernels in any of BSD, Solaris, AIX, anything else, or not?!

Then: isn't this improvement a change that would be too radical to keep it stable? Are they sure it won't break anything?

     Anonymous Coward @ Feb 03, 2007 (06:59 AM)

So the tickless patches produces by the OLPC-team have been accepted by the Kernel-devs already.

     Patrick @ Feb 03, 2007 (08:43 PM)

Linux is quickly moving into the mobile market, and battery life is of the utmost importance. I don't see AIX ever making it onto a laptop, and we shall see about Solaris. I think if Linux wants to gain market share, then it should be able to meet the needs of people who purchase more and more portable computers. When the lead kernel developer said that he expects the tickless kernel to make it into Fedora 7, the guy giving the lecture joked that something would break. Still, the whole point of putting it into Fedora 7 is to work out the bugs.

     Comments are closed for this article.