Should Linux Standardize on a Single Distro?

When I demonstrate software for Linux Journal, I tend to use Ubuntu as my operating system. The reason is simply because Ubuntu is extremely popular, but it begs the question, should the Linux community standardize on a single distribution? Let's look at some of the pros and cons:

Advantages of a Single Linux Distro

  • Linux support would be simplified, as the quesion of "what distro" wouldn't be relevant.
  • Software vendors could release a single package that would install on all Linux desktops.
  • The apt/rpm/yum/up2date/synaptic wars would end.
  • Linux certification would be easier to define.
  • Tux would be everyone's logo. :)

I'm sure I could come up with many other advantages that a "One Distro to Rule Them All" idea would provide. The problem is that the disadvantages are so profound, I think it negates any validity to the first list. Just a few:

Disadvantages to a Standard Linux Distribution

  • A select few individuals at the top would control the present and future direction of Linux.
  • There would be no internal competition in the Linux community. How sad would it be if we only had Windows and OSX to compare ourselves to?
  • We lose the ability to choose, which is a fundamental part of everything Linux stands for.
  • We become a monolithic, bloated, close minded, inbred, operating system with no hope for innovation, and no motivation to think different.

Ok, I admit, the last point is starting to get preachy. It's important to realize, however, that our diversity is where we draw our strength. It's the community that empowers us. It's the freedom to choose that allows us find the solution that best fits our needs, instead of taking whatever vendor solution is provided.

So yes, I tend to use Ubuntu. For my purpose, it makes the most sense. Thankfully, you have the right to choose whatever you like. And that's the way it should be. Thanks Linux.

______________________

Shawn Powers is an Associate Editor for Linux Journal. You might find him chatting on the IRC channel, or Twitter

Comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

I agree with you for the

vagvaz's picture

I agree with you for the diversity is our strength but I believe that unified specifications for package management and file structure could also result in the same advantages without sacrificing so much...

standardization and preinstallation are the key

regina phalange's picture

the only two real reason people think about having a single distro are because of interoperability and gaining market share. standards take care of the first and pcs with preinstalled copies of linux are the the answer to the second. all the hurdles to bringing linux to the masses would be solved with these two solutions. if linux is preinstalled on pcs, noobs don't have to hassle with getting hardware and their mp3s to work.

slackware rules

I'm fed up with you newbie

Anonymous's picture

I'm fed up with you newbie Linux users trying to turn it into Windows: NO.

What kind of disadvantages are that?

Anonymous's picture

"A select few individuals at the top would control the present and future direction of Linux."

Is not that what the kernel mantainers are?, a select few individuals controlling the present and future direction of (the) Linux (kernel)?.

If this is what we have. It's not a disadvantage, it's a working reality.

"There would be no internal competition in the Linux community. How sad would it be if we only had Windows and OSX to compare ourselves to?"

How this pathetic sorrow is any argument?

And the thing about the lack of internal competition is just not true. There always gonna be moaners complaining about everything and screaming like hysterical offended nitpicky ladies about the most infimous bug or missing feature, they will pull the car. XDDDD

"We lose the ability to choose, which is a fundamental part of everything Linux stands for."

No.

The ability to choose is not a consequence of the distributions ecosystem but of the liberty provided by the GPL. As long as the software keeps the GPL you don't loose anything: you have the right to modify everything under GPL.

With a unified GNU/Linux OS you just move that decision in time, to choose after everything is installed. Like in: "it has KD'N'ME by default, well, throw that shit away and push in some G-NODE love".

"We become a monolithic, bloated, close minded, inbred, operating system with no hope for innovation, and no motivation to think different."

Yes... well, actually no.

Let's put it this way: With the adveniment of the standarization of Windows95 the OS industry became a monolithic, close minded, inbred, operating system with no hope for innovation, and no motivation to think different.

You can change w95 for any other OS, even mac system whatever.

As you see this sentence is not true for any case in history, and nothing makes this imagined "unified linux" OS any different (well, the GPL does make it different, but it pulls just in the opposite direction). Reality is just like a wall of stone in the middle of an open field: it's quite disgusting to find one if you're running blindfolded.

And besides with the GPL the energy of the innovation and the ability of people to do what they please will not stop until everybody: Every. Single. One. Stops moving. As someone else said before: a free software project is not dead until the last copy of the sourcecode has been erased from the last hard drive of the last user (or something like that). And that's thanks to the GPL (again: not to the distro ecosystem).

Regards.

PS: Damn you LJ. Because you allow em and i tags but only cite and not blockquote in the comments, your CMS Manager (does it exist such a position?) will burn in a hell of computers infested with Windows Millenium Edition and rebooting from BSOD's for the whole eternity (none the less). ;)

Standards

SkepticalMystic's picture

The ability to choose between distros is one of the most important things about Linux, and the competitive environment is conducive to innovation. By trying to put everything into one compiled distribution, you would lose the innovation that makes the Open Source world so fantastic. With that said, I would love to see some standardization of APIs for software developers to work with. The automated installers that exist in the MS world are one of the few things that they did right. Download the file, double click it, software is installed. Sure, more advanced package managers are making this point null, but what about a piece of software that doesn't have a package for your distribution, or the repositories don't have the newest versions? I'm perfectly capable of ./configure, make, make install, but that's a big hurdle for recent MS to Linux converts.

Skepticalmystic

Anonymous's picture

"The automated installers that exist in the MS world are one of the few things that they did right. Download the file, double click it, software is installed."

Keep in mind that is also what makes virus and malware infestations so easy on a windows system. Personally I think we have a perfect system as it is.

We can actually have both

ArtInvent's picture

Why is this question always framed as either/or? We can and probably will have both one overwhelmingly popular 'standardized' Linux OS - and at the same time there will continue to thrive myriad specialty distros.

For people who want the comfort and simplicity and 'know it will work and there will be packages available' there will be Ubuntu or whatever it shakes out to be.

For those who need something different, there will continue to be other distros just as there are today. Maybe fewer, who knows, but you couldn't kill them all if you tried.

Personally I think there is far too much developer time spent on the OS itself and fine-tuning hundreds of distros that all pretty much do the same thing; most of this re-inventing the wheel effort ought to be re-deployed to apps, in the places where FOSS is still severely lacking, such as video editing, CAD, etc etc.

Dont even think about it

Viktor's picture

For the record..i hate Ubuntu ..i cannot put a finger on it..i know so many guys are using it and it even came up as a google news item..but
i like the feel of Slackware and ..i am entirely at home with Fedora which is my work horse..I know everything is same inside..but i think u gotta choose which distro u need minimum tweaking to become ur pet..
Gentoo will ruin my whole day for compiling and installing everything..
Well, if u got lucky and hav ur way with Ubuntu as standard..u can expect assassination attempts on u..not just flame wars..so dont even tread that territory..oh fuck..u just did... ;-)

One Linux distro

W. Anderson's picture

The one and only question/comment I would make about the article premise is: What distro will be selected and who does the selection?

It would be horrible is Novell Suse is the choice, since Mono, OOXML and other Microsoft "Windows advantaged" technologies would become the standard, decided by the Linux Foundation sponsor corporations who show little if any regard or respect for the "core" developers of GNU/Linux or the community.

W. Anderson
wanderson@nac.net

Do it, do it, do it ---- I

someone's picture

Do it, do it, do it ----
I am all for a unified distro. One server edition and one desktop edition. Take the best of all the distros right now and do it.

Standard distro, diversified support

Phil Hughes's picture

Back in the GOD (Good Old Days) I remember writing about how the big difference between the distros was the support. It actually used to be that way. You "acquired" a distro because you liked the price (free), you liked the included support, you liked how it installed, or some other personal preference criteria.

What we didn't need was the packaging system wars (DEB did what was needed but RPM was marketed as better), the KDE vs. Gnome war (I prefer KDE but, for example, Red Hat shipped a few versions with KDE broken because they wanted to promote Gnome) and the general "X is more reliable/secure/... than Y".

What we all want is:

  • A distribution that is easy to install on our hardware
  • A distribution that is easy to upgrade
  • A distribution that has the applications we want available
  • A distribution that is reliable and secure
  • A choice of KDE, Gnome, ...
  • Finally, something that fits our price and support criteria

What we have is the equivalent of a Ford, a Chevy, a Dodge, a Toyota and more. While the Ford/Chevy/Dodge war went on, Toyota continued to gobble up market share much like Ubuntu is doing today. Why? Because it is works well and at least it used to cost less.

What if we think of Debian as the people that make the frame of the vehicle, the drivetrain, ... Canonical (Ubuntu/Kubuntu/Edubuntu/...) makes the various models of the complete vehicle. Then, you can be a do-it-yourself mechanic or seek professional help. Everyone gets all the choices they need and we get the added benefit that there is only one "drivetrain" so it can be perfected rather than having to identify a problem and they see how to implement the fix in all the drivetrains out there.

Now, if you believe Red Hat, SUSE or some other distribution is a better starting point than Debian (I don't and it is looking like most people don't) then, fine, start there but then build the diversity at the "various models" level rather than back in the drivetrain.

Phil Hughes

You've just suggested what is already happening.

Bunkai's picture

Seems to me like your suggestion is exactly what happens now, in my opinion.

Just take a trip over to www.distrowatch.com and browse through the selection of available distros. You'll soon realize that the vast majority of them have all been created to fulfill some niche, but are developed out of only a handful of base distros (such as Red Hat, Debian, Knoppix, etc.). And when you get right down to the bottom of it... the real "drivetrain" that you speak of is actually the Linux kernel... everything on top of that is what makes up the customizations that you speak of.

In addition to that, there ARE already sufficient standards for GNU/Linux operating systems. Just have a look at The Linux Foundation. From their site: "The Linux Foundation is a nonprofit consortium dedicated to fostering the growth of Linux. Founded in 2007, the LF sponsors the work of Linux creator Linus Torvalds and is supported by leading Linux and open source companies and developers from around the world. The Linux Foundation promotes, protects and standardizes Linux by providing unified resources and services needed for open source to successfully compete with closed platforms."

So, how is what you suggested not already happening?

More importantly, taking all of this into consideration, I firmly believe that any more merger of the various "linux" operating systems would be a detriment to the community and the freedom that it was built on.

Ubuntu works well

T. J. Brumfield's picture

They've had several buggy releases. In all fairness, every distro has a few problems, but I would never suggest that Ubuntu is more stable than another desktop distro. I'd contend Debian stable and Fedora would take those crowns. Gentoo is also pretty fickle about their stable packages.

Can we switch Phil's comment with this story?

Rob Enderle's picture

Congrats, this comment made a hell of lot more sense and explains itlsef bettern than anything the original article had.

One Linux...One Target.

Anonymous's picture

Having a completely unified Linux would also make it a larger target for hackers/crackers/script kiddies. At some point the ubiquity of the single distribution would warrant enough attention and weight to become a viable target.

It is good though that there is a distribution that helps people realize that Linux can be used by non-geeks (The Eee helps in this area too). However, diversity is one of the best parts of Linux, so even though I've really only used Ubuntu, I'm glad that there are other distros to check out once I get confident to move on, or Ubuntu doesn't fit my needs (may also be read I get sick of it). For a starter though, it's been great.

I believe that there should

Debian Linux Geek's picture

I believe that there should be one main distro and the distro is debian. It is the most stable. All of the good distros are based off of it anyway. It has the best package management and would definatly make it easier for software makers to make software for linux. Debian for life baby!

All the good distros are not

Anonymous's picture

All the good distros are not based on Debian. Debian is good, but so is Slackware, Redhat, and Suse. Out of the lot, I prefer Slackware. Anyone that says that one distribution is the best, has probably only tried one.

Single Distro

alphakamp's picture

FLAME BAIT!
In all seriousness, hell no. The kernel is standardized, but the beauty of linux is its ability to be customized.

Each distro serves its purpose, with one version, there would end up being different flavors anyway. i.e. Ubuntu/Kubuntu/Xubuntu/Edubuntu/Mythubutnu

+1

Kartik Mistry's picture

We can all use Debian instead of forking it more :)

I agree the forking of

alphakamp's picture

I agree the forking of Debian is a little bit crazy-- Debian-Ubuntu-Mint for example

LSB could be a good start

Gianmaria 's picture

In my opinion the most relevant advantage of Linux is the freedom for developers to modify him (yes, him :-) ) in order to customize efficiency and usage accordingly to their needs; it would be a great loss to limit the birth of new Linux branches.
Indeed I think it would be great if LSB (http://www.linux-foundation.org/en/LSB) became a real deal in penguin computing. Being able to use one distribution would mean to feel comfortable with all others since libraries and commands would be found anyway. This wouldn't prevent from adding new features to a distribution but only stick with a fixed set of standards. Nowadays only a few distros stick to this and I think that is the main reason why Linux ports are so rare.

Totally agree. What I see as

Anonymous's picture

Totally agree.

What I see as the missing key is the "total" lack of standards for applications and distributions.
Most Linux distributions today consists of a whole lot of "hacks" to make things work.
Yes I mean this seriously.

Some examples:
* Configuration files.
A major part of OSS applications implement config solutions that fit the mind of the developer.
Yes there is /etc and $HOME/.kde but that doesnt cut it. I want an API to access and change all these settings. In Windows you have the registry with its API. The problem about Windows registry is not its interface, but lack of application settings documentation.

Which configuration file do I need to change in order to make my computer dim the LCD after 5min idle?
Which configuration file do I need to change in order to change the KDM theme in Kubuntu?
No, not /etc/kde3/kdm/kdmrc, because they made their own "system" with some dirty hacks in the init files..
Which configuration file do I need to change x,y,

and PLEASE tell which API i can use to make all these changes !

* How do I make a service restart automatically if it fails?
* How do I make a service start at boot?
For some reason Debian/Ubunut decided you define service startup in custom /etc/default/*servicename* , and not with the regular /etc/rc*.d/ linking. But mind you that this is not the truth for all services..
* How do I change service startup options?

Get my point?
It is so frustrating that a platform with all these possibilities has such a major lack of standard APIs.
But I guess lack of standards is the reason why we have such a rapid growth in developement(lack of management overhead perhaps).

Standard solutions

Anonymous's picture

Totally agree.

What I see as the missing key is the "total" lack of standards for applications and distributions.
Most Linux distributions today consists of a whole lot of "hacks" to make things work.
Yes I mean this seriously.

Some examples:
* Configuration files.
A major part of OSS applications implement config solutions that fit the mind of the developer.
Yes there is /etc and $HOME/.kde but that doesnt cut it. I want an API to access and change all these settings. In Windows you have the registry with its API. The problem about Windows registry is not its interface, but lack of application settings documentation.

Which configuration file do I need to change in order to make my computer dim the LCD after 5min idle?
Which configuration file do I need to change in order to change the KDM theme in Kubuntu?
No, not /etc/kde3/kdm/kdmrc, because they made their own "system" with some dirty hacks in the init files..
Which configuration file do I need to change x,y,

and PLEASE tell which API i can use to make all these changes !

* How do I make a service restart automatically if it fails?
* How do I make a service start at boot?
For some reason Debian/Ubunut decided you define service startup in custom /etc/default/*servicename* , and not with the regular /etc/rc*.d/ linking. But mind you that this is not the truth for all services..
* How do I change service startup options?

Get my point?
It is so frustrating that a platform with all these possibilities has such a major lack of standard APIs.
But I guess lack of standards is the reason why we have such a rapid growth in developement(lack of management overhead perhaps).

Choice is everything

Galactic-ac's picture

The increasing pervasiveness of Ubuntu has been troubling to me lately. Of course it's great that desktop Linux adoption has begun to spread much more quickly because of that particular distribution's success, but I'm afraid too much Linux-related information on the Internet is being diluted by Ubuntu. Nearly every How-To article I read in numerous publications these days (including this one) includes a paragraph to the likes of "to do this in Ubuntu, click here, then here, then here..." without a counterbalance of other distributions or more universal procedures which are what highly technical users are often after.

Most importantly, choice is everything to me and I'll be very frustrated when a single distributions popularity begins to leach development time from others, or when applications not typically included in or used on that distribution lose community support because they're unpopular.

Ubuntu's pervasiveness

QuickBrownFox's picture

Ubuntu's pervasiveness should not trouble you. It draws in many users who may or may not stick with that distro. For most people they need to be sold an operating system as a product whether they pay for it or not. This means there needs to be one name and logo and they should both appear on startup. This is less confusing and more welcoming. Ubuntu is not Linux. Ubuntu is Ubuntu, and within its own area of specialisation it's bigger than Linux. Why should Linux need to be mentioned every time Ubuntu is? Perhaps the person making the tutorial has never used any other distro? How and why should they know that the steps apply to other distros? There's nothing wrong with this and it's a reality you will have to accept assuming you wish for Linux to become mainstream. The only way to fully acknowledge the heritage of a distro is to mention it's name along with KDE/Gnome, Firefox, Xorg, and all of the other bundled applications, which on the desktop are just as important as the kernel itself but that would be ridiculous. I'm all for keeping it simple and promoting a single distro to the majority. But there will still exist a variety of others that many people will use and the more computer minded users may then learn more and experiment. But those who have no interest could happily use the most popular one, which looks like Ubuntu, and set their minds to much more important matters.

What kind of user you say?

Anonymous's picture

Don't be offended, but if you don't know how to translate the "click here, there and there" of the ubuntu how-to into the apropriate config files on your distro of choice... I'm sad to tell you that you're not a Highly Technical User (maybe you aim to be one, but you are not there yet).

Re: What kind of user you say?

Galactic-ac's picture

Well, I'm certified by Red Hat, for what it's worth to you. But if I need to troubleshoot a failed join on one of my GUI-less servers to an Active Directory running Win2xxx, I'm sorry but GUI-driven, distribution or desktop environment-specific instructions are not helpful. There's quite a lot of functionality available in most configuration files which isn't exposed or operable by GUI, and my point is that much of the information out there has begun to get very seriously watered down with Ubuntu-specific "click here, there and there" directions. It has become difficult to weed through all of that to locate the information that is really valuable to technical users.

Re: Choice is everything

nix2ways's picture

I couldn't agree more. The 'buntuisation of Linux is a double edged sword. Ubuntu is a good distro, but we need to keep it perspective. It is not the best distro, there's no such thing. It does not 'just work' as people keep saying ad nauseum. It needs as much post install configuration as any other distro. We seem to be in a situation where Ubuntu is just accepted as the standard by which other distros are judged and that is wrong.
The variety of distros available is one of the great strengths of Linux.
Let's never move away from the diversity and equality that is one of the cornerstones Linux and keep the playing field level.

Single point of failure

GaryM's picture

The reason different distributions exist is the fact that no one distribution is viable for everyone. If there were such a distro why do we have 400+ to choose from?

Having one single distro would also be a single point of failure. Putting all of the eggs in one basket. No. Linux' strength lies in her diversity. There is no reason to single up on any one distro.

This appears to be the dreams of those who already have the distro picked out.

As long as there are unmet needs, developers to code, and a user base, there will be a plethora of distributions to choose from.

And that's the way it should be.

Webinar
One Click, Universal Protection: Implementing Centralized Security Policies on Linux Systems

As Linux continues to play an ever increasing role in corporate data centers and institutions, ensuring the integrity and protection of these systems must be a priority. With 60% of the world's websites and an increasing share of organization's mission-critical workloads running on Linux, failing to stop malware and other advanced threats on Linux can increasingly impact an organization's reputation and bottom line.

Learn More

Sponsored by Bit9

Webinar
Linux Backup and Recovery Webinar

Most companies incorporate backup procedures for critical data, which can be restored quickly if a loss occurs. However, fewer companies are prepared for catastrophic system failures, in which they lose all data, the entire operating system, applications, settings, patches and more, reducing their system(s) to “bare metal.” After all, before data can be restored to a system, there must be a system to restore it to.

In this one hour webinar, learn how to enhance your existing backup strategies for better disaster recovery preparedness using Storix System Backup Administrator (SBAdmin), a highly flexible bare-metal recovery solution for UNIX and Linux systems.

Learn More

Sponsored by Storix