More at DataMation...

What are your thoughts on the declining number of Linux distributions?

--jeremy

My thoughts. Ubuntu spinoffs. Like I am running on this touchscreen laptop.
If those disappeared. Then the stats would be divided in 1/2 and there would be less.

One man developers need help. Not vampires. But some folks don't think this way.
I help out on 2 distros that are not dead yet. One has had a resurgence of sorts teaming up with the Mepis community.
More hands means more help.

There is still lots of choice for the gear I own as my profile states. So I do not sweat the small stuff.

Tongue in cheek.

"They can take away our choices! But! They canna take away our Freedoms!"

http://linux.meuhobby.com/themes/gal...sword-4918.png

because funtoo exists... arch is rolling, but you're still at the mercy of the system devs because they dont have use options of each package. people know its better to concentrate efforts on distros instead of pointlessly fork creating redundant work. so yeah basically we've reinvented the wheel a few times.

First, I think the word "hobby" is vastly overused to describe "non-commercial" uses and interests. I think it is used intentionally by the larger commercial interests in a perjorative and dismissive way and should be avoided in all serious discussions of GNU, Linux, FreeBSD, etc., particularly at the distro level.

Now, why the decline in number of distros? I would suggest the competing forces of complexity and maturity.

The kernel alone is a whole different level of complexity than it was just 5-6 years ago. Same probably for a majority of useful applications that must be included to comprise a distro. It just takes a lot more effort.

That does not necessarily mean fewer people are involved. It just means that instead of wrapping your own great new idea into your own great new distro (because it was easy), it is now more likely to result in a great new idea packaged for existing distros.

Also, a few years ago everything was still new and relatively immature in some sense. A new immature distro competed mostly with other still new and similarly immature distros and more readily found acceptance. Now a new distro must immediately compete with much more mature distros, and among a user base that has more mature expectations.

So even a really good new distro must fight its way into a much more established distro-ecosystem, and it is significantly more difficult than in even recent years.

I think users just move upstream once they get comfortable with GNU/Linux -- I know I have.

Generally speaking, the further downstream a distro is, the more newbie-friendly it tends to be.

A corollary of this is the increase in abstraction in an attempt to add GUI-centric features that tend to cause more troubles than they solve (IMO).

So I would say this just reflects the increasing experience of the user base and a willingness to get their hands dirty with the "nuts & bolts" of their OS, rather than just clicking on the helper boxes provided by downstream spin-offs.

I think it is likely an artifact of distrowatch policy. Just because distrowatch lists less distros does NOT imply that the number of distros is declining.

If there really is a decline I imagine the economy has something to do with it. Less spare time.

"Some might say that the decline hardly matters."

Yep, I'd say that. As if 280-something distros aren't enough?

[Also, Distrowatch.com monitors *BSD systems as well, so I wonder about the accuracy of the count being actual Linux distros -- my guess is that the BSD count was not subtracted.]

I think that many persons who start distros do so without realizing how much commitment is required to maintain them. To build on what rokytnji said, there was a big increase in distros once Ubuntu gained traction, because of Ubuntu respins.

They may do it just to prove to themselves that they can (which is a noble motive) or to scratch an itch to have distros that suit their tastes, then their lives change, they get families or jobs or debts, and their distros wither. I will give a dollar to a doughnut that the ones that have disappeared in the last three years just sort of faded away . . . .

Just my two cents.

Commercialized use of Linux took it to this time where people do not have much time to maintain and concentrate on new distros. Everyone wants to make things quick and money-bonded.

Anxiety of the quarter end results made it possible to ignore such distros which are not qualifying the current much wanted scenario and people are more looking towards making that better which is currently fulfilling their agenda commercial or not.

Distros those had a long run since the start of linux era are so more reliable and qualitative that even if a new distro is introduced, it would need much more dedication, resources and hard work to gain the confidence of this result oriented commercial world, and even if they do so the current champions of the market would launch another update to the Kernel going furthermore ahead leaving the newer one behind with bugs notified as vulnerable and proving their reliability and durability.

For the most part i have been able to stay out of the " Cookie cutter" mess

in the RPM camp there are not all that many

the testing - fedora
the business - RHEL and SELS / SELD
the lab - SL and Fermi linux
The Home - CentOS ( or business or server)

Then a few cookie cutter forks

but Debian has had fork after fork and forks OF forks


getting back to a few MAJOR OS's with a few custom spins is a GREAT start to having Nix on A LOT of desktops and laptops

Since 90% of distros are Ubuntu spinoff, well I'm happy that number is decreasing.

Why didn't they just ask the distro devs why they quit? That might have been more entertaining than having the author make a guess about it. Oh well, I hope they enjoy the +1 to their page hit counter.

A lot of assumptions and guesswork, I guess they ran out of material.

I think the main problem is that the majority of the Linux distributions aren't really distributions - more like someone customising the desktop a bit, releasing a few of their own packages and slapping some paint and rebranding on and calling it a distribution.

To me a "fork" is where a developer takes a one time snapshot of the code base of some software and heads off in a new direction. They might take some ideas from the parent project and port over some software along the way - as well as from other projects - but continue development independently and no longer depend on it.

A fork is not where the new project relies on the old and bases it's releases on the releases of the parent.

This is why if you look at the "true" Linux distributions, there aren't really that many and a lot of them aren't that well known and have small user bases.

I think a large part of it is need. When distrowatch formed, a lot of distros were...clunky. So there were HUNDREDS of distro's just aimed at making them less clunky. But as time has passed, they've gotten less clunky, and quite frankly, we don't need as many spin-offs to fix things the mainline distro's haven't implemented, because at this time they've got just about everything working.