Did we really need yet another Linux distribution?
I saw over on Linux Planet that Sidux, another Debian derivative, has been born. It never ceases to amaze me at how needless some of the open-source community's development efforts are. Did we really need yet another Linux distribution?
What is Sidux's claim to fame, besides an unfortunate name?
...Sidux is a great choice for Debian fans who want Debian Sid in a nice polished package, and who want it as plain vanilla as possible, rather than heavily-modified as so many Debian derivatives (such as Ubuntu) like to do.
So there you have it. What? You're still here? You haven't rushed out to download it?
Of course you haven't. There's no need.
Yes, I could be eating my words if Sidux becomes the next OpenSUSE, Fedora, or Ubuntu, but let's face it, it almost certainly will not. There are hundreds (thousands?) of Linux distributions, yet only one handful that get serious traction (SUSE, Red Hat/Fedora, Ubuntu, etc.). The rest are there to serve someone as a scratch to a very narrow itch.
Fine. I'm not suggesting it's somehow wrong. I just don't see why anyone bothers. It's one of the core promises of open source that anyone, anywhere can tweak the software to fit their needs. In the case of Linux, apparently many people feel the need to tweak.
My question is, why? Why not just use those efforts to build up the core Linux distributions, instead? Why not give more time to Ubuntu, in this case, rather than building Sidux? Does it really break significant new ground?
Matt Asay brings a decade of in-the-trenches open-source business and legal experience to The Open Road, with an emphasis on emerging open-source business strategies and opportunities. Matt is vice president of business development at Alfresco, a company that develops open-source software for content management. He is a member of the CNET Blog Network and is not an employee of CNET. Disclosure. You can follow Matt on Twitter @mjasay. 
As all the existing Linux distributors have failed to break through the Geek Ceiling, there is still room for one more, particular one who promises "as plain vanilla as possible".
Take a chill pill; go with the flow; give them a chance to be *different*. Get over it - and get over yourself.
If they fail, by all means give them your usual kicking. But not yet.
Well, this is *how* we have OpenSUSE, Fedora, and Ubuntu. The licensing permits fragmentation, so we end up with the thousand-grains-of-sand treatment. It's what you might call a different paradigm.
This is a huge strength of open source. If we ever get to the point where we have "consolidation" within a few popular distros, that's when you need to be worried. It means something has gone wrong.
>Why not just use those efforts to build up the core Linux distributions, instead?
Because their goals are different? Mac OS and Windows will be emulating this someday. We already see this happening in industry. Car manufacturers will be allowing you to choose a chassis and build on top of it. You'll be able to personalize like never before.
Another reason: How many flavors of ice cream are there? Why not combine them all into one mega-flavor?
I have a specialized multimedia distro (Dyne:bolic) with a realtime kernel for doing audio and video production. Then there's another distro (Puppy) that allows me to recycle an old machine with new software; it runs in 128MB RAM without needing a hard disk. Then there are more desktop-intensive distros like those you mentioned. To combine all of these into one distro would be a joke.
Ubuntu tries to do this to a degree; you can install ubuntu-studio easily, or xubuntu. But that can only go so far before these sub-brands start diluting the overall brand strength.
So we look at the lower level: One beauty of this is that I'm able to customize these distros for my customers for use in things like kiosks and appliances.
This all goes without mentioning how difficult it is to come in and make major changes to the way an existing distribution works. There is red tape to be dealt with, and half the fun is the DIY part anyway.
This distro seems to be one of the less is more distros. In that it isn't as bloated as Debian has become or other Linux distros have become. The more bloat an OS has, the slower it runs and the more hardware it needs to run at a decent speed. You'll see that even Mac OSX and Vista have added on so much bloated features that customers want a plain vanila version of those OSes that aren't so bloated, but none exist.
One of the few OSes that really got it right was AmigaOS. It always had a small footprint in memory and loaded fast enough and works really great. But Commodore and others nickled and dimed it to death by not marketing it properly and not getting enough third party support for it. I can recall the Amiga BCPL and C compilers where less than 2000 bytes and the original AmigaDOS/Workbench disk fit neatly on a 880K floppy disk that still had room to spare, and still did true pre-emptive multitasking and at that time Windows 3.X, OS/2 2.0, MacOS, and others needed a hard drive and had much higher footprints in memory.
Ubuntu = Google...
SuSE = Novell
etc. etc. The huge fragmentation across Linux is also a strength because it means a rogue programme or virus would never be able to hit all Linux machines because the variations from one distro to the next are soo different.
It is not enough to put up a so-so Office-like, or a wannabe Visual Studio-like. What most business users want is something at least as good, preferably better.
THAT would be a trully liberating effort!
Meanwhile lets get back to complaining of Vista :(
a) Developing better drivers to support a wider variety of hardware.
b) Develop more feature rich applications for various hardwares.
c) Improve documentation for novice users. Currently, distros or new versions of existing distros are released with scant documentation.
d) Try to standardize the file locations across various distros.
These enhancements will result in a wider adaption of Linux as a desktop OS. Till then, a new distro is just one more among the crowd.
To the original posting: In regards to Matt's original question the answer is "Why not?" - 'free' (as in people are free to do what they wish) open-source software is much closer to an evolutionary process then closed source or commercialized open-source. So it produces many dead-ends, but makes leaps also that may never have taken place in a commercial environment. All these have their value and I don't dismiss closed/open/whatever - but likewise nor should we dismiss the choices of others. I hear endless complaints in the J2EE community about the number of web frameworks for example - I'm sure Matt you here people complain about the number of CMS on the market. The question one must ask oneself is why is it a problem to me? These people don't work for me or harm me, so why should I be concerned about what they chose to do. Who cares if there are 100 different web frameworks, 50 variations of linux and 100 CMSs on the market. Eventually it will get whittled down to the ones that are not evolutionary dead ends by a process of natural selection; and everyone got to try out their ideas. It's this healthy vital ecosystem that makes a market feel alive and exciting. If we dismiss this activity then really we're heading into the direction of the market dominators. Now commercial enterprises are going to always want to dominate the market; but domination isn't healthy in the long run - just like a mono-culture. We have Windows Vista to prove I'm not wrong.
So this incessant and sometimes minor modification might seem unsettling but it's how the human race came into existence in the first place and that is a an amazing feat in itself - little organic chemical molecules arranging themselves to create the most amazing machines in the planet. So personally I have faith in the process and accept the dead ends - because the ones that don't dead end can have interesting results - and you'll _never_ know for sure which one is the going to be the most successful branch.
All the best
Neil
projects in general. If the goal is to write for a few geeks, then the current way of many distros, i think has paid off . But if the goal is have a higher market share
than windows and Mac OS then we'd better rethink. And we should think as if we were to start a new star up that wants to have his or her solution the most used.
Microsoft has to write their own kernel, their own Boot loader, their own desktop manager, their own networking connection tool, their own paint soft
etc... and put all together and make it works. For Linux Distros, they dun need to write all those, yet still to be frank (I'm a Linux fan), the convenience of using windows
for the normal user (not the geek) is still higher, I'm talking about intuitive menus, GUI, ease of use, great apps
. Notice that if people are talking about windows on how-tos, new apps...they don't usually say this is for "Gottam windows" version 3, that is for "Tintin Windows" 7.10,
they just say this is for windows and in a few cases vista or xp. It means when you have use a windows, you can use all windows. And that it's not true for Linux, at least for the average user. Do really think if a user knows how to install apps in Ubuntu, he can jump on arch or red hat
or Slackware and install the same way? For windows, he can for any version of windows (Notice i didn't say any version of any distro of windows).
You have many hundreds of Linux distros but all of them fall into a few categories like some for pcs with limited resources, some for multimedia, some for small boxes like routers,
some for networks administrators....So why don't we have version of linux for only the few categories needed under one big open source company
. Imagine all Linux distros devs together, add all the open source projects devs (Linux kernel,gimp,gnome,kde,open office...) , they all have a common goal and way of working ,
they have one way of gluing things together (which is came from brainstorming the best way of current ways) . So when people talk about Linux, they only say Linux not Ubuntu or Gentoo. Geeks will have their version, Multimedia people will have their version
, people who wants a fast Linux versionwill have their version. And the quality of apps will be better because so many devs are available to work on it.
Let's take for example installing apps on Linux distros, you have yum for some you have apt-get for others. Frankly, for the end user, it's the same
, he just wanna install an app. What about putting all people working on yum and apt-get together to write one app using the best of both worlds. Meaning it will better than yum and apt-get
. Also, we have ".rpm" and ".deb" files, don't you think we can have only file extension like may be ".lnx". I'm sure if those companies are really for open source projects
and they want Linux to grow, they can provide a way to only have one extension for installation files.
Surely, one Linux distro is far from now, but at least we can do that for some common applications and tools
It's a free-market system. People get to open yet another restaurant or manufacture another MP3 player if that's how they want to spend their time and money. Why don't they just get a job at Bennigan's or Apple? Because they have the freedom to try to offer something better.
I expected the author of "The Open Road" to get that.
A rich rewarding life is rarely an efficient life. A rich rewarding world is full of diversity not uniformity. When people need standards they appear naturally, I'd rather have a little inconvenience and have a world of variety.
To the commenter who wants to have a market share the size of Windows, well certainly you can do that - but in doing so you will have ended up becoming just like Microsoft, by removing the diversity. So now we'd have 2 MS like companies instead of one - not such an amazing gain.
A fantastic example of how diversity can be coped with is Java; Java came to the fore because it allowed software to run on all the main Unix platforms. This is a fantastic example of natural standards emerging which cope with diversity.
And just remember some of the worlds greatest atrocities were carried out by those who seek uniformity and efficiency - and that is no coincidence - since the desire for uniformity is an extension of out egoist wish to mould the world into our own shape. Because the uniformity we seek is always _our_ preferences.
This is the Internet, not Philosophy 101.
Whoever said "The Geeks will inherit the Earth" never tried to hold a conversation with them.
Pointless flamebait.
quoth the Mattster:Fine. I'm not suggesting it's somehow wrong. I just don't see why anyone bothers. It's one of the core promises of open source that anyone, anywhere can tweak the software to fit their needs. In the case of Linux, apparently many people feel the need to tweak.
So, how about your home/house/abode, Matt? Do we need *YET ANOTHER* custom(ized) House? Why even bother, you just live there all the time, using it everyday.
What about your Car? I hope you didn't bother to put a new stereo in it or any kind of aftermarket rims or accessories... why bother, you are only using it everyday.
Why bother? *WHY* *BOTHER*? That is like Einstein: Why bother with Relativity... it already does all we need it to do, we don't need to USE it or bend the rules of it.
You see, its not about *YOU* Matt, its about the choice that was made to do it. Why did *YOU* become a freelance writer, when you could still be a Staff writer? Why? Huh? Why indeed. BECAUSE YOU CAN or COULD!
Why did Bill Gates start Microsoft? Because he could.
Yes, I understand your disgust with YALD. But to be honest, if nobody would have bothered, Debian, RedHat, SuSE, nor Ubuntu would have been born. Heck if nobody would have had that itch 99% of the Software included with most distributions would not even exist.
"Did we "really" need another "content management" app? " refering to Alfresco.
As for how someone wastes their time, its no worse than playing rounds of golf or Halo.
I waste mine running Gentoo :-)
1) because "I" can. The nature of FOSS is that I can get the code and make a variant if I want to. Some want to. They may or may not get traction and may or may not care.
2) because contributing to someone else's distro is a pretty darn hard way to get compensated for "my" efforts. Yes, the FOSS model says the money is in the service, not in the "product". Cool. But if I contribute to the codebase of a distro that someone else has built the service market for, it is an uphill battle to carve out a place for me. To paraphrase very loosely Jack Welch: if you want to make money servicing someone else's distro, provide a better service or offer a better price.
Neither answers why try to make a go of it with a distro at all vs starting a project on a previously unmet need / unsolved problem or at least going after a need that is only addressed with non-FOSS today.
Why do some people favor blue, others red, others yellow and others green as favorite colors?
Not sure what your point is, if these guys were not doing there own thing do you think they would just contribute to another Distro? I think its the same as bloggers and journalists that yap every day on the net and carry on doing it even though no one reads there bull.
As a matter of fact, I was looking for the same thing, and that's how I found Sidux. I typed "debian sid live cd" into google .
Which is where you could have easily found information about why Sidux was created, by the way. I guess we should cut you some slack about such a transparently phoned-in piece, considering the holiday weekend. I picture you typing this on your blackberry with one thumb while driving home Sunday night. Of course, it doesn't help that your only source was that equally underresearched article. Why are the people posting comments in these articles always so much better informed than the authors?
I think you should consider eating your words if Sidux becomes the next Kanotix, which is clearly what it's made for, rather than "the next Fedora, OpenSuse or Ubuntu".
It never ceases to amaze me at how needless some of the "journalism" efforts are.
Sidux follows the unstable distribution. If you had spent 30 minutes understanding your subject material like any honest journalist would, you would understand why this is different.
I love Debian and found SIDUX a great choice. This post was written on a Sidux system. Sidux boots and operates much faster than Kubuntu. It ships the latest sofware.
I dont like the way the Kubuntu team has ripped off lots of useful toolbars and functionalities away from Konqueror. (I don't understand why they do this). Try viewing pictures in Konqueror from Kubuntu for instance, and then do the same thing in Sidux's Konqueror. Anybody will, I suppose, appreciate the difference.
I suppose the "journalist" Matt ASAY wanted to be really provocative. At least, I hope so. I remember my sister went to visit Moscow back in 1982 and went to a supermarket. They found : one brand of beer, one brand of water; one brand of tea, and hopefully two different vodkas.
Linux and free software gives what the user needs : CHOICE.
Many CNET readers might never have heard about SIDUX before. Thanks Matt for the advertising. You should download Sidux (there is a 440 Mo only CD available, very fast to download) and so you can try by yourself and make YOUR OWN opinion.
Long life Debian, long life Sidux !
Alejandro GOMEZ.
to talk about sidux you need to know debian, not only ubuntu.
Debian is the largest linux project, that works over 16 architecture.
There are different level of package development: a package can be considered ready for stable only after heavy testing.
Only the stable release have an installer. The testing is a pre-release, that now have an installer too to prepare the next stable release.
The sid hasn't an installer, or a live cd.
Sid is also available for 16 architectures.
Sidux is Debian Sid live cd installer for i686 pc.
The sidux team made possible we can use Debian sid, with livecd, usb installler... and the latest driver for the newest hardware.
We really don't need ubuntu, opensuse, or fedora to full reinstall the system when a new version is released, when we have the latest software... on the go.
Matt, please, try sidux live cd!