Slashdot Mirror
Torvalds Rejects One-Size-Fits-All Linux
Barence writes "Linus Torvalds has rejected the argument that Linux developers should pool their resources behind a single distribution. 'I think multiple distributions aren't just a good thing, I think it's something absolutely required. We have hundreds of distros, and a lot of them are really for niche markets. And you need that — simply because different markets simply have different requirements, and no single distro will take care of them all.' The calls from the Linux community have been growing due to Linux's failure to show significant market share growth."
Did you ever think that he might be right?
We need a main, reliable, one size fits all DESKTOP distro. that's what we need.
and yes, all other distros should continue, for really many of them are for niche markets.
linux basically equals webserver as of now. whereas many IIS servers house 1-2 company sites (and many of them are in-house boxes), linux distros host hundreds each.
but on desktop we dont have a strong name presence so that when you name it, everyone will know. we need that.
Read radical news here
Comment removed based on user account deletion
The reason I love linux is because I have the choice. Minimal distro, server oriented distro, etc. Trying to make one big distro is absolutely the wrong thinking, it would be impossible to decide on anything first of all, and its been proved this concept doesn't work already, by a company called Microsoft.
Obligatory blog plug: http://www.caseybanner.ca/
Yet another distro. Anybody have a link where we can download this One-Size-Fits-All Linux?
No folly is more costly than the folly of intolerant idealism. - Winston Churchill
Linux Starter
Linux Home Basic
Linux Home Premium
Linux Business
Linux Enterprise
Linux Ultimate
Perhaps instead of worrying about the specific distro being worked on (and distro-specific apps), developers could unite to improve the libraries, services, and interfaces that are used universally. Gnome and KDE, for example, are the "face" of Linux to the average user. And let's face it... KDE is modern but broken in many ways, and Gnome is stable but behind the times in many ways. The specific distro being improved is less of a concern if the focus is on bringing stability, visual appeal, and new user interface innovations to the frontend of Linux itself: the GUI interfaces that the average user works with on the system. Working on that aspect would make every distro benefit.
-Vendal Thornheart
I use different distros for different tasks, because the distros themselves place different weights of importance on various factors.
For years, my servers have run on Debian plus the odd BSD box here and there. Rock solid reliability with very little maintenance overhead, but you don't get the latest and greatest stuff in the repositories.
I've got a couple of servers running Ubuntu with VMware Server on top for internal VPS work. Again, very few problems aside from a couple of issues related to kernel upgrades.
My laptop runs Ubuntu Desktop edition, which works great for me. I have almost no trouble with package management, even for cutting edge stuff, and the driver support is great.
I use a couple of live CD distros for repairing Windows systems when they get out of whack. The list goes on and on. It's kinda like programming languages; use the right tool for the job. While you *could* use most modern languages for just about any task, some are better for "X job" than others.
512 MB RAM, 20 GB disk, 200 GB transfer, five datacenters. $19.95/month.
THE obvious choice has changed over the years. THE right decision is wrong several years later. So who gets to make the call?
Never trust a man wearing a coat and tie!
Linus said exactly what was needed to say. Seems like the blogosphere (or how you call that herd of wannabe journalists) only didn't notice that there's no need to care about market share.
Agreed. And from what I remember, the size of the Ubuntu distro was set at one CD. So if they make the move to a DVD (I mean, c'mon, it's 2009 people), they could fit Kubuntu, Net Book Remix, and whatever else you'd want on one disk. Add an Android distro (Abuntu?) to the family and presto! Linux for everyone!
Tic-Tac-Toe, Global Thermonuclear War, and relationships all have the same winning move.
... and not just because he is boss (structural power).
Linux is hardly a new thing, and this is hardly a new question. There _are_ seriously differing requirements, even in the desktop arena. Also, distros come and go. Linux is eternal :) [I wonder why].
Sure, one distro makes VAR and other support easier. But open is not impossible. 'make' is pretty easy and unlike MS systems, compilers are standard.
If you want one distro, just go FreeBSD.
Although its good that certain distribution cater for different markets, the problem is the over saturation of one area with too many choices.
if linux is to really take of as an alternative it should follow the Apple model - introduce a hardware/software packadge. Give problems with getting drivers to work, and even installing the OS, it is unlikely that Linux would move much beyond the market share it already has (which at this point is limited to enthusiasts). A hardware solution (say a laptop/desktop that comes preinstalled with a linux distro) and is aggressively supported by the hardware manufacturer, if well designed, could make that problem less acute.
-- All this knowledge is giving me a raging brainer.
-netbooks w preinstalled linux
-notebooks w preinstalled linux
-cameras
-phones (google and also others)
-e-book readers
-a large number of set-top boxes.
-major pc vendors (dell) selling linux preinstalled
-network hard drives and other intelligent network hardware
-a the biggest RT operating system vendor cooking some linux
-ms support for linux
-a number of up to date embedded development boards from brand manufacturers
Most of these things have seen a strong rise in the last few years. and for nearly all of them it was mandatory that a specialized distro exists. I don't see how Linux is failing because of not having a single dominant distribution. I see how linux is succeeding because of the number os special distributions.
Did these people think about why one branch of Dell sell computers with ubuntu and the other one with RH linux?
I agree with him...to an extent. Yes, it's true that having several distributions does help fill in some niche markets, but having 100s?? That just leads to confusion. Besides, where has linux gone (in terms of market share) in the past several years? Virtually nowhere.
One big problem that newbies (such as myself) have is that not only are there tons of widespread distros to choose from (ubuntu, fedora, suse...), but each look and feel differently wrt each other (my experience, anyways). That makes breaking into linux that much harder - How do we choose? Just like the /. gripe about windows Vista having "too many" options (7?) for a consumer to have a good sense of what to get, can you imagine 100s?
I think it's time for a different approach: have a few (3? 4?) "main" distros (in order to foster some kind of competition) for the newbie (like ubuntu), the advanced version (for /.ers) and derivatives for niche market. Anyways, just my 0.02. Opinions?
Clearly you are not very familar with the linux (or OSS) community. Ever notice the wide range of opinions concerning things like design, inclusion policies, licensing, etc? Have you thought what would happen if you tried to make all those people share a distro? There are plenty of flamewars already, do that and the community would tear itself appart. New distros don't pop up for the hell of it, they pop up because people want something that fits *their specific needs*. Their needs are often unique. People need to get off this whole idea that linux is about "sticking it to the man" and that it needs to change in order to get better marketshares, just for the sake of marketshares. Linux is meant to be useful for people who want it, if it's not for you, then who cares? We're not out to become rich billionaires by toppling microsoft and apple, we're just making a nice operating system for ourselves. This is something the majority of the world can't seem to understand.
"linux is just DOS with a UNIX like syntax" -- Galactic Dominator (944134)
You could say we really only need one Desktop distro. But... People work on what they like. You can not force them to work on what they don't want too.
We have Ubuntu which has a big lead on the desktop so we have some some of those benefits. The problem with Linux is the lack of commercial software and support.
You can not call the manufacture for help or geek squad. You can not go and buy software you want. There are a lot of free packages and many of them are great. The problem is the average person doesn't know what is good and what isn't. Even when the software is really good the documentation often isn't. Out side of GIMP and OO.org you will have a very hard time finding books for FOSS applications.
I know that Click and run failed but I still think a application and media store is EXACTLY what Linux needs. A super easy built in solution just like what you see on the Wii, XBox 360, and iPod/iPhone.
See my blog http://ilovecookes.blogspot.com/ for light hearted technical information.
I could go on, but you get the idea. Don't believe everything you read.
Let us not become the evil that we deplore.
Multiple distributions are a necessary part of the Open Source ecosystem. Competition keeps the ecosystem healthy, selective pressures keep Linux evolving. Windows is built according to the direction of the Microsoft Holy Profits.
Linux grows according to what the people developing need and want. There will always be the question of which matches more closely what the average user wants and needs, but much of the strength of Linux comes from the existence of multiple distributions.
Fanatically anti-fanatical
Why not? Except the different variations would cost the same ($0) and could be tailored to different markets without locking that person into that niche. So, if Granny wants to email and web surf, give her Linux Desktop Basic - locked down, minimal extra services (Really, does she need and FTP client?), easy to fix. So now she gets into it, and decides to start using some office apps for writing letters and keeping her budget. She downloads the upgrade for $0, and when it's done it looks mostly the same but now has a few more options open.
Market segmentation is not good or bad; it's restricting movement between segments that is so crappy.
"As God is my witness, I thought turkeys could fly." A. Carlson
because we're not making money at this and seriously, who cares? Linux is a choice, not a goddamned marketing campaign.
This guy is way out there
Comment removed based on user account deletion
And where have these calls be coming from, exactly? As far as I can tell, the only one asking this silly question is some dude from DistroWatch... and it's just a question during an interview with Linus, for god sake!
Something tells me "pcpro", whoever the hell they are, is simply manufacturing a story...
How are they measuring it? Just counting how many boxes sell redhat and suse?
You have linux from bios, networking equipment, embedded appliances, cellphones, all the way up to massive clusters. Windows could be keeping taking over the old concept of desktop, but the world breathe linux in everything more complex than a calculator. What about counting market share including all of that to see more to the ground numbers?
I wonder how much of the desire for unified linux is based on technical consideration, and how much is just the standard human tendency to believe that your group will be more effective if it is purer and more unified, most frequently seen in the politics of adverse times.
In effect, the number of distros that are relevant to the alleged confused market is already very limited and fairly similar. Is Joe user actually not using Ubuntu because the existence of Gentoo and LFS and L33tNu> confuses and terrifies him, or for some other reason?
Don't we already have that?
Linux Home Basic - Ubuntu
Linux Home Premium - Fedora
Linux Business - RHEL/CentOS
Linux Starter Edition - Xandros
Linux Ultimate - Slackware
Some linux users also use agent-changers, to get around those idiotic sites with hardcoded browser requirements (that work fine in Firefox /w Linux, but display an error message unless you tell them you're running something else)
Comment removed based on user account deletion
Wasn't there just a front page ./ article complaining about the number of SKUs for Windows Vista? The complaint was that there were 7 (or maybe 8) different skus for Vista and how confusing this was for the end-user
In Linux land there are way more than 8 different distributions and options (desktop, email program, editors, etc.) within those distributions
Clearly more options can't be a disaster for Vista yet a good thing for Linux.
With Vista, a home user merely needs to decide between Vista Home and Vista Ultimate; if they need VPN, their corporate IT will probably recommend Vista Business. If they need both Media Center and VPN, then they have to spring for Ultimate. Not that hard.
Compare that to the choices faced by Joe User(tm) with choosing an Linux distribution.
Of course clones could be a great way for others to see what works and what doesn't, what feature is worth noting and etc. Of course if your product does the same as a few other products and in no way better or worse; just less used; you could consider jumping aboard a bigger distro. One could assume that certain things regarding making and maintaining a workable distro is made easier with resources and capacity.
The Long Now Foundation
Did you know that Stallman sounds like the Swedish Stålman, literally man of steel (or figuratively Superman). The nomme de guerre Stalin means (more or less) man of steel in Russian.
Clearly Stallman has the right name and the requisite facial hair and he can write GPL4,5 and 6 to enforce collectivisation of Stallix and the crushing of Kulaks like Torvalds.
echo -e 'global _start\n _start:\n mov eax, 2\n int 80h\n jmp _start' > a.asm; nasm a.asm -f elf; ld a.o -o a;
Actually I disagree with Linus. However, I can see the need for multiple too.
But imagine a single "frontman", with different "packs" that can be selected during install. So there is one "public" Linux
Need a server:
Download the Base public linux, then pick your "package" for the duty of the system
Base server pack - stable ...
Base server pack - leading edge
Base desktop pack - stable ...
Graphic artist pack ...
Setup repostiories that "are the distros"...
Here's the full interview minus the editorial: read it at distrowatch
Linux guarantees no vendor lockin. And today that applies to where you get your OS from as well as who's hardware you choose to run it on. From a customer point of view that's GREAT, as it fosters more competition and innovation from the providers.
Anyone arguing otherwise just doesn't understand business and market forces.
Of course, the millisecond that happens, the thing'll fracture into a dozen or more different, uh, "tiers" (think: Basic, Home, Premium, Ultimate, or some junk like that). Next (but still well before 1000ms has passed), every Linux developer worth his salt will immediately create his or her own, uh, "customizations" - because ONE SIZE DOESN'T FIT ALL.
So - approximately one second after unifying all Linuces as one Linux, it'll immediately breed by fission - not into two but literally dozens of new, uh, . . . (can't call 'em distros any more. Custom Linux versions?).
Should be: here
practices. When a company or government finds Gnu/Linux fits the bill better than Windows, Microsoft comes in and essentially pays them to stick with Windows. Governments like Egypt where the OLPC people had a MOU from them but then Microsoft goes over, they talk, Egypt accepts something like $50 million in stuff from Microsoft and when OLPC shows up all they get is "Does it run Windows".
And let's talk about how HP, Dell, Lenova, etc can not advertise their Gnu/Linux products. Leaked MS memo's already showed Microsoft's hand in this too. They basically said, "you can not lead with Linux" and that meant advertise and the threat is most likely to be those millions of dollars in Marketing Program kickbacks for putting those little MS stickers on everything and saying crap like "Runs best with Windows", etc.
_That_ has been what has limited marketshare growth to a large degree. IMO. Remember, we are a world full of followers so if too many start going to Gnu/Linux, the horde will follow. That's why Microsoft spends hundreds of millions to stop the switch.
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
Why are we *nux users so apt to spread our beloved OS to all the unwashed masses when what makes us great is a dedicated group of hobbyists and professionals striving to work without the constraints that mass popularity will bring? I for one am glad that *nix's have not caught on an spread like wildfire, because from a security standpoint (my work) we are a tough target because we are small, and whether one realises it or not, we are so ahead in many areas because *nix is the underdog and has to be vastly better to even be considered adequate by the *nix community standards.
If the free market decides on a standard, then that is the choice of the free market. And if you happen to look at the market share of linux, the free market isn't choosing linux: right now, it's overwhelmingly picking Apple at the high end and Vista at the low end.
The important thing is to pick something and stick with it, even if it is the wrong answer, you have a platform on which to build it's correction. If I could produce a single binary that would run on Linux 3.0, regardless of kernel options, window managers, or that computer's driver set, Linux would immediately become a serious contender in the desktop market as a platform for delivery.
You'll notice that the areas that Linux is a winner in now are things that run a platform of specific applications, like LAMP. Until "Desktop Linux" means a specific set of APIs that are standard across all installations, it will never, ever, not in a million years, become the desktop of choice for anyone but people like us. People who have time to skim man pages and forums for three hours to figure out how to do simple things, like get their wireless card up and running.
With labor costs going up, Linux has even less of a chance. Vista may suck, but I can make a program that will open on every single version of it without any tweaking at all. That should be the goal, if Linux users are serious about taking over the desktop.
Yet I find it funny that Microsoft is bashed for their choice to offer multiple versions of their Operating System.
The greatest revenge in life is massive success.
developer mindshare and market share You dont have to stop work on other distros to get a winning distribution. If you like one, just stick to it and as much as you can to get it better. /. read the mythical man month, it would be unreasonable to expect that a distro will get better by pooling programmer resources. As for monetary resources, they will eventually follow the best amongst all these distros.
I really don't buy the idea that pooling resources will make things better. As I would expect most people on
You speak London? I speak London very best.
Emperor Skull joked about Linux Starter, Linux Home Basic, etc above, but I think the concept of "Linux Starter" would be a good one.
If the community could settle on one or two distributions to recommend to new users, fewer new users would be confused/disheartened by the number of selections. Just make sure a search of "linux for new user" comes back with a page pointing to the two easiest distros to get up and running.
Would that be hard?
One single distro definitely would be a bad idea, but I wouldn't mind seeing one package management system that all of the distros agreed upon.
First off, I'll be the first to say i'm no expert on linux, but I don't think one distro would be a good idea (let alone feesible). However, maybe narrow the playing field a bit? I mean, hundreds of distros? You'll always have the application specific OS (Phones, stand-alone devices, etc) but I think there should be some way to convey to the average joe that "these distros are a good place to start". Windows has different "distros" so to speak (not talking versions like 95/98/xp/, but XP Home vs XP Pro, Server 2003 Stanrdard vs Enterprise, etc). I believe Linux kind of accomplishes this via Kernels, so there is some common ground between them (or maybe i don't fully understand it). But even knowing what I know, I still don't have any idea of what distro would be good for what application. Maybe establishing some kind of "master list" of distros and what applications they're good / better suited for would be a place to start? From the outside looking in I don't know where to even start with Linux. I'd like to get into it more because its obviously a great OS with plenty of great applications, but where do you start?? It's like walking into a food court with 20 different restaurants that all look the same from far away (except for their name), but you gotta walk up and read each menu to figure out what they have to offer. Then as you're walking around, you see other hallways with even more restaurants! Pretty soon, you just say forget it and go back to the restaurant you know best, called Microsoft.
If Linux wants more market share it needs a face, it needs something people can point to and say "Linux".
That's the way this sort of thing works.
From someone who is no longer in the tech community but is a big geek and has geek friends, we have the wrong aim, people.
We need good communication, not analytics. Yes the OS should be improved, but in order for lots of people to adopt it we need to communicate a fairly unified, confident idea.
If Ubuntu is that face, great! Let's work on that.
If KDE is that face, great! Let's work on that.
If Gnome is that face... you get the idea.
People love consistency. We geeks want to analyze and pick apart everything, change it and tweak. Your average person DOES NOT CARE! They want something that works. Until we get that through our oversized brain/ego/whatever then Linux is not likely to take off in a really big way.
Will this sacrifice a few things? Sure it will. However, since it is Open Source those little niche OS's can still exist! That is the problem with the big players now. We can still tweak things to make them better.
Geeks like to be RIGHT and not make mistakes. I think it has something to do with smarts, or not being hugged enough as kids, or something. Their confidence/power comes from analyzing and making the "right" decision, which is why science is an analytic's passion. We can be "Right". People do not always want that. They want something that makes them feel good, simple and easy that they don't have to think about. If that thing is windows or Mac for them then so be it!
The bright spot: What if we did this and got more market share, huh? We'd be in the spot where software SHOULD be. The geeks run things behind the scenes, tweaking and improving, altering and modifying for their user base, while the average person (99% of peopl out there) can use an innovative, slick interface that runs on cheap hardware. When they want to use their special application it works! When they need software or processes tailored to what they are doing, it will be easier. Businesses will run better since there will be less down time on the user side (i think), fewer upgrade$ to the newest Mac/Windows Neon Bloat Fantastic, and fewer headaches with techies trying to make programs/systems work together.
All we have to do is learn to set aside our infighting because we want things to be scientifically "perfect) and market some form of Linux, anything, and unify behind it for the user base at large. Yes there should lots of distros for niche markets, but a general distribution would be very helpful.
-
I use Linux on the Desktop. There is no sign of anything even remotely Microsoft on any of the 4 boxes in my house.
I know Linux. I use Linux. I love Linux, etc., end of story.
While I readily admit that anyone with even an iota of common sense and responsibility to anyone else on the internet should use Linux, I really don't care. Linux doesn't have to be more popular for me to justify its use. Other than being part of a safer internet, more use of Desktop Linux has zero effect on me.
I'm not changing. I do believe that sooner or later, the rest-of-the-world will see the light, and join me in the bliss that is Linux - the ability to exercise virtually 100% control over my system, but I'm in no rush. It doesn't have to be tomorrow. Or in 2010. It'll happen (and I'll be here to say "I told you so" to all those former MS users....)
I use Arch as a one-size-fits-all distro. pacman is awesome. I have built desktops and servers dealing with many different tasks from the same ISO. It really is a benefit if you take time to learn it.
Ubuntu is a necessary evil. For some reason, we need a distro for the Windows masses. But it's better that than Arch mailing lists spammed to the brim with "How do I listen to my MP3s? Linux is dumb."
While a one-size-fits-all distro might sound like a good idea in theory, in practice it's very bad. Unless you want a Linux-based Windows.
No, they didn't spend millions, since copying a CD only costs a few cents.
Of course, Microsoft counts the donated software at the full market price, but they just provide cheap copies.
Microsoft is just counting on the fact that they are the first ones, but we'll see if this strategy will continue to work once everybody will be more fluent with computers.
Thing is, most of the time when I see these "market share" figures it turns out to be measured by revenue from new sold units during the most recent [quarter|year|whatever].
Someone erasing their "Windows 2000" system and turning it into a Linux server doesn't show up at all on this measure. Someone who has to "upgrade" their Windows server repeatedly while their Linux box sits and runs without needing any additional spending on it distorts these numbers, as do the people who spend twice as much on each server due to software licensing fees.
This is going to be even more distorted if they're specifically talking about non-server "market share", since it's so hard to find pre-installed Linux desktop systems most of the time. I have a suspicion that a lot of Linux desktop machines - even the NEW ones - came with "lowest-common-denominator" Windows OS and were subsequently wiped and replaced with a Linux of the installer's choice rather than showing up as an explicit "linux desktop" purchase somewhere.
Hacker Public Radio is our Friend
A sizable portion of the linux community are not using mainstream distros and have no desire to do so. In my experience, most of those calling for unification are Microsoftee's.
When a comment like this is hidden to -1 and thus both censored from view and censored from archival, the Slashdot forums are officially dead. Disagreement has officially become trolling, and echos of "conventional wisdom" bounce about exponentially, like an array of ping-pong balls on mouse traps.
One thing is for sure, however: Desktop Linux has not gained significant institutional usage. That is a fact. Downmodding that fact will not make it go away.
I don't mean the Editions of the Vista...but besides Windows Server, they make Home Server, Small Business Server, Essential Business Server and Windows HPC, they all based on the Windows Server.
YEAH! WOOO!!! Ok.. how do I install these RPMs? You know.. for the handful of things that I run that ONLY come in RPM format?
Trust me.. we still need those niche dists. Maybe not all of them, but many of them.
Now don't get me wrong. I use ubuntu in a few places, but as long as Cpanel is RPM based, I will continue to run RH/Cent boxen too.
Can all fish swim?
to the individual distros. a single distro is not a good idea for the reasons Torvalds gives; however it may be beneficial in particular instances to merge together distros with the same goals to reduce replication of work. actually merging two communities together is another story.
I'm getting really annoyed with all these call for greater market share. It' almost as if someone has discovered your awesome fishing spot and decides that it would be a great idea if they built an interstate highway next to it so everyone can fish there. I'm not against greater market share, but market share as a goal is dumb.
Apparently "Calls from the linux community have been growing" translates to "an interview question where someone asked the question on Distrowatch"
I can see how the submitter got these confused, the two being so similar and all.
From tfa: "And for the last few years, it has been Fedora," he reveals."
That's the revelation?
Move along. Nothing to see here.
is not whether a single distro survives as the dominant one, but whether there can be an established standard, a specification for ISVs to refer to when they want to deploy on GNU/Linux distributions. The fact that many distributions have different configurations/places to store their prefs, have different package management systems and so on, makes it difficult for ISVs to test their stuff and deploy on a platform. For instance, if I wanted to release Photoshop on Linux what do I tell my clients? Requires glib v2.4 and above, GTK 2.12 and above, GNOME 2.22 or later, blah blah... isn't it easier to say Fedora 11 and above? or is it better to say requires LDB (Linux Desktop Base) 1.1 (like LSB only more desktop oriented). The LDB should even define some way for the app to list its dependencies so they can be automatically installed by packagekit, aptitude, or whatever. I could then, just test the product on a single LDB platform and presto! I have a single platform to support, no problems (ok I am simplifying this a little). Anyway, say no to a single distro, say no to a single desktop; diversity is the strength of libre/open source .. there just needs to be some work on interoperability and we are done.
Thanks
There can't be one size fits all in Linux. The very essence of Linux is have multiple distros and people creating offshoots. That may be the sole reason in the end that Linux may not completely replace Windows! However, that is also the strength of Linux. One good example of why one size fits all is a bad idea, is that in today's world most people will discard the old P3 with 128 MB RAM. However, Puppy make it crisply working system, and in style too. Some students who come from a village to study at my Univ in Delhi, bought an old P3 with a 2 GB (yes, no typo) hard disk (running windows 98), for a paltry Rs 2000 ($ 40). They couldn't even use their pendrives on it, and could not install any useful software in that little space, whereas their friends made fun of them. After I handed them a remastered Puppy CD, they are thrilled to bits. They use internet, watch movies off pendrives, and listen to mp3 music. The friends who were laughing at them, stare at Abiword springing up immediately on the icon click. And believe me, in our country there are many who can only afford such a system. Which size will fit this category, except Puppy (or DSL)?
Well if you compile from source it usually works across distributions.
If I ask if something works, and you say, "usually", then you're telling me that it doesn't work.
There is no perfect solution. On the one hand choice is good, competition is good, concentration on specific needs is good.
On the other too much competition is very costly, too much choice is confusing.
I hope you know what I mean. Besides this there aren't a lot genuine distros. If I am installing ubuntu, change the application, name it mybuntu and save this, I don't have created a real new distro. A added so little value that it does not count in my book.
Rockin regards,
Marco
It's allmost big enough for me.
I mean: I can get the software.
Nice, interesting stuff.
I can get support.
Most stuff even works!
So...
You are so right. I wish I had mod points. When I first read your parent post, I thought this will be modded up. But no, you were flagged "Flamebait". That is really wrong. Just so you know, I think a lot of people would think you are indeed right. Desktop Linux is not were it could be...
"Sockets are the standard networking API, also useful for stopping your eyes from falling onto your cheeks" zeromq.org
I was thinking pretty much the same thing... Except I did not include openSUSE, because it has the Stink of Novell on it. Fedora and Ubuntu are first-class, first-rate Linux distributions, widely available, and widely used. Either Fedora or Ubuntu can be simply installed onto most any modern PC, and provide functionality for free that would cost (literally) thousands of dollars for the commercial equivalents.
I think that this year (really!) will be the year for mainstream Linux desktops. Microsoft's missteps with Vista, and the general state of the economy will combine to get more people to look around for reasonable alternatives.
The real reason Linux has not better penetrated the market is the lack of high-end applications, ala Photoshop, that run on Linux. However, we get closer and closer with reasonable alternatives that meet 80% of the users needs: The Gimp, and Open Office, for instance.
there are 3 kinds of people:
* those who can count
* those who can't
Mostly untrue. Linux does change, but from distro to distro and release to release the changes are in line or in order with progress toward ease of use and standardization. If you don't see it it is because you don't want to. Like the monkey with his hands over his eyes.
If a desktop doesn't appear to meet business needs for the next 3 years it is because it was never given a fair look. Clearly the statement above where you state you have never used Ubuntu demonstrates that. I've worked in IT for a number of years and watched management make decisions with a lack of knowledge which ultimately hurt their bottom line, but what can anyone say.
Ubuntu, in most forms, is stable and has solid performance. It also has LTS (long term support) releases every so often, which means it will be supported in that version with little modification for a number of years.
The unfortunate fact is that Linux has a ton of documentation. For instance, when you look for help on the web you are inundated with support for Linux. So many articles, so many Wiki pages, so many man pages, so many blogs, it's just hard to find out of that what you want.
All my videos play on linux (all of them).
Domain controller and codebase is a different animal than the desktop. That's apples and oranges.
Most of the high end SQL databases are available on Linux and some even started their life on Linux.
Most everything else is just business complaining and not doing anything about it. A couple years people such as you complained about drivers, etc. Those are now non-issues. Linux on the desktop is an elegant and capable solution.
You can lead a man with reason but you can't make him think.
The problem with multiple distros is that you cannot support them. Let's say Intuit decides to release Quicken for Linux. They will be getting support calls from Fedora users, Ubuntu users, Debian users, Mint users, Suse, Yellow Dog, Ygdrassil (or however that's spelled) etc. Suppose further that Quicken needed a minimum of 1024x768 resolution. Where do you change that? It's not a simple matter of right clicking on the desktop, selecting properties, and moving the slider. It's not even as simple as opening /etc/X11/xorg.conf in an editor. Or let's say you needed to open a port in the firewall.
Linux is not going to make significant progress onto the corporate desktop until software companies start publishing linux versions of their software. They will be reluctant to do that because the cost of support will be so high.
When our name is on the back of your car, we're behind you all the way!
The installer cd and desktop skin of the platform is irrelevant to the underlying functionality. This is precisely why we should pool our resources around the best distribution. Does Linus weigh in on what boot loader people should use? Or what X11 based gui?
Embedded systems, and rare platforms will take care of themselves. They will always be supported in the same sense that any platform is supported because this is open source.
I don't believe that concentrating efforts on a single distribution does anyone any good. In fact, I believe that the only thing that it could achieve is concentrating the desktop linux effort in a single point of failure. For example, let's focus on the main distros and ignore the later forks and offshoots.
If that distribution was Debian then the problems that have plagued that project would've hindered all progress that we've benefited in the desktop front in the last 10 years. That's not good.
If that distribution was Red Hat then, as the company is focused on the server then we wouldn't have much of a desktop to begin with. That's not good.
If that distribution was SUSE then... Well, desktop linux would be in the hands of Microsoft. That's not good either.
What other distribution is left? Slackware? Gentoo? Arcane commands and compile everything from source? There's no doubt that those may be very positive selling points for the refined geeks among us but what about mere mortals?
So, as I see it, there may be no ideal distribution but the best desktop options available today are here only because someone saw an itch that needed to be scratched and had the courage to get off his ass to do something about it. The same progress would never exist if that same person had to jump through hurdles imposed by a pre-established project to get his idea up and running. As in nature, diversity is the single most important source of progress and evolution.
Slashdot, fix your code or at least hire someone who is competent at it to do it for you.
Comment removed based on user account deletion
What Linux really needs is mainstream applications to be supported. I bought a Mac so I could run Photoshop, Aperture, Final Cut Express, Logic Express and iTunes. And YES I do use all those and NO there is nothing comparable for Linux.
I do use Linux full time at work as my main desktop but then I do software developments and Linux is well suited to that. I started with Linux back in the pre 1.0 Kernel days but I do like to use Photoshop, Aperture, Final Cut Express, Logic Express and iTunes.
Come on! It's not about Linux, it's about the free desktop. It doesn't really matter if it is Linux or BSD or HURD, it doesn't matter if it is Gnome or KDE or a mix, what we need is to have an standard Free Desktop for desktop developers to target.
Consider the success of Windows, by far the most successful OS, why? because of the desktop. Linux can already hand its ass to Windows in the server and the embedded market, small devices and big devices, they are all covered.
Where the lack of penetration really hurts our (my?) daily life is in the desktop, every time I have to install a windows software, every time flash behaves weird, every time I visit an IE only site, every time one of my relatives system gets infected with viruses, I wish for a wide spread free desktop.
My full support is with Ubuntu, they have the right goals and the capacity to reach them. They say "stop trying to turn Linux into Windows" but we (I) *NEED* a Windows replacement dammit!!
But... the future refused to change.
but linux is still comparatively hard to use, full stop. Many computer users have never even installed an operating system, never mind trying to get their home scanner folder shared across their house using samba. Also, Ubuntu is ugly, a minger, a dog, beaten down. That has to change too. As to the specific post, linux is better with competition methinks. The lack of competition is precisely why Microsoft get away with being so trousers at everything.
He's wrong. There's no way that software developers are going to develop software for a fragmented platform. It's a nightmare to develop and support. It won't happen. And until big manufacturers start making stuff for the *nix platform, Linux will never be a viable option for many, many people (like me!)
whats the name of that coverter again? alien?
comment first, facts later. http://chem.tufts.edu/AnswersInScience/RelativityofWrong.htm
What if you want one that somebody spent a lot of time and money on making a consistent look and feel? Don't mistake me, I run Debian myself, but if I wanted to set someone up who isn't a computer enthusiast, I'd probably go with something that has a more polished or unified look like gOS or Ubuntu. That's the point I was trying to make with this comment: that there is no distro that satisfies everyone's needs so there's no use pretending that there could be, but I was modded troll as you can see.
Gentlemen! You can't fight in here, this is the war room!
Market share is the number of copies sold as pre-installs (e.g., netbooks) and retail boxes. For Linux, this number is really immaterial.
The number of interest is the 'installed base', which is the number of copies installed on hardware. For Linux, this number is hard to get. Some of the larger distributions have started making (low-ball) estimates, but even they admit the numbers don't really reflect the number installed, for various reasons.
Another question is whether or not to count the number of embedded Linux copies. If my TV, DVR, PMP, MP3, PDA and other devices run Linux (they do), should those count toward the installed base? Or should we be counting general purpose computers only?
Can You Say Linux? I Knew That You Could.
Suppose someone creates a very minimalist linux distro which includes a very good package management system. Suppose this package management system includes nearly all popular linux software packages.
Now suppose it were rather easy for anyone to install any number of those packages, bundle them together into one meta-package keyword, and call that a distro.
That's pretty much the stated goal of Debian, it aims to be the universal operating system. You can do just that with APT.
Nick
Seriously, in virtually all cases where you see instructions to "apt-get" you really want to use aptitude instead. Aptitude manages package selections far better including remembering that you installed library x simply to make package y happy. It'll automatically uninstall x when y gets removed!
Nick
Given two choices, in some cases, one option is good and the other bad.
In other cases, both options are bad.
In other cases, both options are good.
But in none of these cases is there necessarily a reason for the option to exist. Things as basic as "how application foo stores its default configuration" should not be something that varies from distribution to distribution, for example. (just an example)
Different distributions only make sense when the choices they've made make sense. If it doesn't make sense for there to even /be/ a choice, then there's no reason for multiple distributions.
I think a lot of the "little things" which fit into that category are what annoys people about switching from one distribution to another. First step: those things should be eliminated.
As for "niche distributions", I think a lot of those could also be combined in some way. For desktop use, it's clear that not everyone needs every package, so they have easy ways of choosing what is relevant, and how it fits together, and making sure that all the shared parts are secure and bug-free. I don't see any reason why this wouldn't make sense for any other niche, even if it's something horribly meta like "a distribution for building distributions".
What is really so incredibly niche that it can't pull from the same server as everything else? Don't throw your weight behind one distribution, if the one you'd support doesn't exist yet.
-- 'The' Lord and Master Bitman On High, Master Of All
or, you know, to have alternatives that aren't horrible. Or to stop pretending that the alternatives are non-horrible.
-- 'The' Lord and Master Bitman On High, Master Of All
No they're not. They're like very naughty boys.
Windows does precisely this. It tries to hide behind the whole Home vs. Professional vs. Ultimate, but really it's one Windows distribution.
The whole point of multiple distributions is to target different markets and needs. What might be a good idea is to find out where distro overlap occurs and consolidate based on that need, but to get behind one single distribution that "fits all" sounds like a recipe for a piping hot cup of fail.
You never worked with Linux or indeed any OS have you?
The linux kernel ain't 30mb. For instance mine are around 4.5mb so you are not even close.
Oh wait, you are talking about the complete source? Yup that is 30-40mb. Why should you wish to put the complete source code on a mobile phone? Anyway, modern phones can easily have several gigabytes storage so even then it isn't that much.
Oh and while my kernel is 4.5mb, that is because it includes a LOT of drivers that are not needed but I am to lazy to remove. A mobile phone maker would compile the kernel with only the options that are needed for that piece of hardware. The proof? Mobile phones with linux running on it.
So basically, you are suggesting a fix for something that ain't broken. Oh well that is slashdot for you. What next, you claims Vista is slow because you have to install it from DVD?
MMO Quests are like orgasms:
You may solo them, I prefer them in a group.
Kernel devs focusing on one distro would not force anyone to use it, nor would it prevent competition or niches, since it is all still open source (unlike Windows). It is a preference, not an ultimatum.
What it would do is go a long way toward establishing standards and increasing flexibility which are sorely lacking. Making that one distro (or meta-distro) more flexible (to accomodate derived distros) and standards based is far better than having thousands of rigid and incompatible options.
The community can, of course, still evolve new standards through competition. They just won't have to do it in parallel for each major distro (or desktop).
Platforms do.
And except for Android, I know of no Linux-based platforms aimed at normal users and/or app developers.
Distros are too fluid and there are too many of them anyway. This situation makes coding-for and independently distributing PC applications very confusing.
The only things that would rectify the situation would be to create a fully-spec'd out and vertically-integrated (up through the GUI) platform like Android, or have the community get behind something like LSB Desktop. The latter does not seem to be happening though because it it being marketed to neither users nor app developers AFAIK.
Notice there was no mention of LSB in the article -- There's almost zero awareness of it.
I would like to point out that Linus is against forking the kernel, and his group essentially demands a unified kernel and toolchain (with different distros having different configurations of these pieces).
But when it comes to higher-level stuff that end-users require, they complain about one-size-fits all. Frankly, that attitude says to me that the audio and video architectures in Linux-based desktops will continue to be slipshod and wobbly (unstable performance and unstable APIs), and you can forget about widespread adoption at the consumer level until either the Torvalds mentality dissipates or an Android moves into the desktop space.
I think Torvalds & Co. are hypocrites who prefer showing off to their coder pals, users be damned. Even worse, they're foul-mouthed trolls who regularly make personal attacks on people they dislike while insisting on civility being directed towards themselves.
Linux will continue to act as repellent to ambitious application developers looking to make their mark or a buck. We'll have to be content for the forsee-able future with ham-fisted G-, K-, X- apps that are usually mere shadows of what they imitate.
Alas, even excellent software like Firefox doesn't get major UI flaws (like radio buttons always disappearing) because of this situation... Mozilla doesn't even bother packaging their apps for "Linux" anymore... you gotta unzip it to /usr and make all the correct linkages and icons yourself.
The other great FOSS app, OpenOffice.org, is fairly complex to install/upgrade even with rpm/deb packages... and proper desktop integration will be either absent or badly broken. Again, SUN/OOo would rather attempt a fit-and-finish on proper platforms like OS X and Windows than play the bitten-by-a-hundred-repository-hackers game.
A single distro is not necessary. I think people here have made that clear. But I can say what is necessary is a standard base library file list so that people creating commercial software can release software without having to provide source code. Now I know that some of you will say they can release the source its better. But in some cases it is not. Especially if you have trade secrets you want to protect. Commercial vendors are never going to agree to releasing their source code. Something similar to install shield would be ideal. Something that detects the distro and provides the bindings to the correct libraries. Maybe even downloads the dependencies for you.
One kernel
Actually, nexenta (opensolaris kernel + ubuntu userspace) is looking pretty good...
I mod down anyone who says "I will be modded down for this", regardless of the rest of their comment
You have those that will pay for vendor to support it, just another "UNIX" like operating system to them, and those that use it because it is open and free and want to roll their own. Server rooms don't show up on usage stats. Linux will never gain a significant desktop share in it's current incarnation or with the current direction. It has become quite obvious that those looking for a UNIX desktop have chosen Apple, that was the impetus for the early Mac adoption that I saw. Time after time I saw people that had tried running Linux on a laptop but like me they got tired of having to massage things and so they got a Mac notebook instead. I love my MacBook Pro and no Linux laptop/desktop has ever come close to it.
but the evolution of Linux begins at the base. With new releases being a norm every six months, new developments come and go. The linux kernel is the fundamental base. The extensions surrounding the kernel should have a fundamental to reshape how the kernel is exploited. There needs to be a market leader to invest time and resources to simplify how the kernel should be exploited. It will change the dynamic for Linux distributions while maintaining the integrity of the opensource community. The fundamental of the kernel isn't the problem, it's the fundamental of the extensions surrounding the kernel that is.
In my belief, there needs to be a distribution that bridges the gaps between other distros, then another one to break down dependencies exclusively. It doesn't necessarily require totalitarian consolidation but there needs to be a community effort to focus on variations of the types of consolidation.
Fedora has done great efforts with security but has some fault with stability for certain features. It's cutting edge just not as stable as I would like it to be at times. OpenSuSE has stability, ease of use and great insight into stable features but with encrypting the root and swap partition by copying files over to another partition, repartition, then move the files back is a bit of a stretch for the average user. The system updates could be more aggressive but hardware support for nvidia and ati drivers is perfection. Debian is great and stable, very proactive in terms of updates but in terms of getting it to meet up to current technological standards without sacrificing it's open integrity is the only thing that keeps me from switching over fulltime. Ubuntu needs stability. Most of the updates are no more than scripts to get things by. You see this when trying to do a base server install and then attempt to piece together a decent windows manager on top of it. If a system had the security of Fedora/Red Hat, the stability and ease of use of OpenSuSE/Novell, not to mention the integrity of Debian, you would have one hell of a distribution. If anybody has the time, they could learn a lot from Linux from Scratch. I'm doing my part but I'm just one person. Anybody out there willing to work on a project like this, I would be more than willing to help out.
Here goes some karma...
Frankly, Linux isn't growing because it has fundamental problems, even in its "core competencies".
Last year, I was volunteering for for the local congress critter. They needed a number of PC's with web browsers to access their canvasing tools. Where I work, we have a parts room full of 2-7 year old desktops, so I borrowed a few with the intention of installing Linux and Firefox and loaning them out. I downloaded the most recent version of ubuntu, and openSuse and installed a mix. I would say in large number of the cases the install was _NOT_ smooth requiring some text file editing or worse to get the machine to a usable state. That was ok, I did that before donating them. The problem is when I loaned them the desktops they were plugging their own monitors in (donated by someone else). In probably about 50% of the cases the resolution/refresh rates could not be set by the GUI utilities even though in 100% of the cases i saw data was available with the DDC display utility. One guy requested I reset the refresh rate because it was hurting his eyes. The look I got when I ended up hand editing the xconfig file summed up the why Linux isn't anywhere near ready for general consumption.
This is acceptable if you have an "expert" pre-install the machine, and setup every piece of hardware and software that will be used. At that point the linux machines need big stickers that say "no user serviceable parts inside." I end up hand patching and compiling stuff all the time, I frankly cannot figure out how a normal user is suppose to figure this stuff out.
Far and aside from the fact that community and volunteer developers wouldn't just surrender their favourite project for the sake of the Linux "brand", a bigger problem would be the corporates.
A massive part of the hard-graft development is funded/sponsored/carried out by the big corporations in the FOSS world. All big companies only ever get involved with projects where they think they can make a profit out of it. If we had a one-flavour "official" Linux, most of their business models would be gone. Without a good business model they wouldn't get involved, and that means less resources backing the Linux project as a whole.
Take Red Hat. Their core business is in selling their official boxed version of RHEL, and selling support services to the companies and regular users that adopt both RHEL and Fedora. If there was just one variety of Linux, they wouldn't have their own product to sell- the best they could do would be to sell copies of Mono Linux (which is what I'm calling it now), where they'd be competing with every other Linux vendor (including those distribution it for free). Similarly, without their flagship products, their support services suddenly have no big selling point. Red Hat contribute lots to all manner of Linux-related FOSS projects- without them, theres less money to be had.
Similar goes for Canonical- they started a brand new company so they could develop and promote a new distro, with an eye to making a profitable, sustainable business. If they weren't able to create their own distro for some reason, what incentive would they have had to get involved with Linux in the first place?
Take away the best of the corporate support, and add in the troubles you'd have getting the community to support Mono Linux, and GNU/Linux wouldn't be what it is right now. Not to say it couldn't work, just that it'd be something else entirely.
tl;dr- If you want Linux without lots of distros, try BSD.
If people want to grow the Linux market share which would be good for everyone involved in Linux then people at least need to settle on certain standards or get behind one.
More importantly to get somewhere a distro needs some marketing and it would be better to pool resources behind one considering even then paying for marketing is bound to be a money losing situation for some time.
It would not hurt to get behind one distro for a short period when the return is that more developers put focus linux and companies start giving it better support so people don't have to spend their free time making things work that should just work and can instead focus on innovation rather than making printers and web cams work.
"I'd prefer to see App Bundle distributions similar to OS X"
Such a thing actually exists:
http://klik.atekon.de/
The people that are jumping up and down about Linux not gaining market share simply don't get Linux. It's like market analysts trying to analyze a block party, not realizing that its a block party not a stock exchange. Linux is not about market share. It's about creating an OS that works. If people like it, they use it. If they don't like it (or don't understand it), they use something else. There is no marketing plan. There is no "market expansion". There is no "competition". People who want it, use it. People who want to make it better, develop for it. People who don't, don't. That's it. Even Linus himself said that the destruction of Microsoft will be a totally unintended side effect. Yes, he was be sarcastic, and he was only talking about one company, but it's basically true. There was no intention of competing with anybody. Anyone who tries to make it out to be simply doesn't understand how Linux came about, why it still exists, and how it will continue to grow and thrive despite other people's warnings of "not gaining market share".
Why do X-Servers have the graphics drivers rather than the kernel or HAL? The X-Server should only be a consumer of graphics services!
You must be a Linux-only kind of guy. There are a dozen other kernels out there that can run X as well, if we move graphics drivers out of X and into each kernel then we end up with more OSS fragmentation than you just set out to solve.
I went to eat some animal crackers and the box said, "Do not eat if seal is broken." I opened the box and sure enough..
Fair enough. If I spent the afternoon trying to count them, the number would probably be around 5. So I exaggerated a bit. Although each of those had its share of sub-headaches.
./configure, it may very well be that the two things I'm trying to run together are not capable of running together. Or maybe they are. I never know, but I always have hope. Hope that strings me along for 30 hours of trial and error, at the end of which I'm fairly convinced that it's not going to work, but I also think that it still *could* work. So I keep plugging away at it, or I admit defeat and feel like an incompetent fool (and perhaps I am).
./configure doesn't work, it tends to be isolated to my particular strange configuration or ignorance, and help is not to be found.
I think it's a psychological irritation more than anything else. If I MUST use an application and MUST use Vista and the two will not work together, then it's a done deal. I fail. I can't do it. I can then sheepishly tell the client that it's a no-go. We move on.
With
So while it's not really a happy situation, I am significantly less crazy when confronted with a 100% guaranteed "this will never work" failure immediately that I can move on from, than a "this should work, but it's not working" that strings me along for days. As an added bonus, if enough people are confronted with a 100% guaranteed "this will never work" and are all pissed off about it, chances are someone will find a way to make it work and try to sell it to me in the near future. If a
-- I prefer the term "karma escort."
We're not out to become rich billionaires
I am. But only because I like the offspring of Debian and the color brown.
Just as a counter point, Ubuntu (at least, while using Gnome) doesn't have "Configure Desktop" at all on the menu when you right-click the desktop. It does have "Change Desktop Background", but that only lets the user change the background image (and other visual preferences), but nothing dealing with the screen resolution.
That was kinda the point of his argument. In that hypothetical situation, Quicken tech support would have no idea how the user's system varries from an expected Linux install. Unless a standard was enforced across all distros, each user would be have to be treated as an individual, potentially different from the previous one, drastically increasing the time tacken to support him or her vs. if there was a unified standard.
Who doesn't give a flying fuck if Linux gains marketshare or not?
Don't get me wrong, I hate Windows, and I wouldn't impose it on my worst enemy ... ok maybe I would.
But I hate Windows, and I love Linux and all that, but at the end of the day, you run whatever the hell you want, and I'll run whatever the hell I want, mmmkay?
look at how many different names you mention while counting. gnome and apt, ubuntu, kde, yast, opensuse, yum, fedora.
how do you expect telling a non tech (even linux) savvy average joe/jane all about these ? they will get confused even before you start explaining them in detail.
we need one name prominent like what's firefox now :
- hey need an alternate browser ? why, try firefox !
everyone gets this.
Read radical news here
sentences apart into two halves that can't be understood on their own, and the subject isn't descriptive of the body.
Sorry to pee on your sugarcane, but it's really annoying... And sorry for not being above it myself.
its for people who have to use Windows or Macos.
its for people who cant set out to create their distro according to their needs.
basically its for average joe and jane, in middle of nowhere, to do simple tasks.
Read radical news here
we'll see if this strategy will continue to work once everybody will be more fluent with computers.
I think "everybody" will be more fluent with computers only inasmuch as young people start to replace those who are old now.
Take your in-laws. Take your old English class. Take your band, or the stamp collectors' club. Of those, take all between 15 and 35. How many run Linux? They're the "everybody" who will inherit the earth.
I might redact this in 35 years when I have grandkids ;-)
There, fixed that for you.
Linux already has significant market share. Look at the web and the Internet infrastructure: the vast majority of it is powered by Linux, technologies that are based on Linux, or utilize Linux in some indirect way. Linux is gaining traction left and right in the embedded market. No competent system administrator hasn't at least fired up a LiveCD to see what all the fuss is about.
It's true that Linux isn't as strong on the desktop as many advocates would like, but that's mainly because there's not yet any big company throwing their weight behind it to leverage business deals and spend billions in marketing to the consumer. (Canonical is trying, but they're still pretty small fish at the moment.)
As I've written repeatedly, ever since the very beginning Linux has had steady but slow growth. This isn't a good thing, nor is it a bad thing, it's just how it is. I think what we're seeing now is that more and more people are looking at Linux and open source and saying, "now, how can I make a quick buck off of this?" and realize they really can't and then spend all day lamenting about it in their blog.
Why don't all the people that keep complaining about the variety of distros, simply pick one and unite behind it, then help out?
FOSS = Choice. Make one.
-Viz
Don't kid yourself. It's the size of the regexp AND how you use it that counts.
Did he learn nothing from the '80s? How many Unix distros existed then and who profited? Microsoft.
Those who cannot remember the past are condemned to repeat it. - George Santayana
Because earlier he said, "I think that people who argue for splitting desktop kernels from server kernels are total morons, and only show that they don't know what the hell they are talking about." (source)
Any sufficiently unpopular but cohesive argument is indistinguishable from trolling.
Welcome to the "wisdom" of crowds.
The difference between the layers of different distributions is _profound_
And that's part of the problem. Linus may be right that there are good reasons for multiple distros to exist. But that doesn't mean there are good reasons for them to store the same files in different locations.
One of the downsides of multiple distros is their tendency to do *everything* differently. At one point, that was probably a good thing. Experimentation and competition may be the best way to learn how to do things optimally. And in any case, in FOSS, there's no way to prevent it.
But at some point, certain things have been shaken out to the point that they're more or less equivalent, and it would be a good thing for the multiple desktop distros (after all, that's what we're talking about here) to agree to agree.
Beyond that, where competition is still producing some benefits, diverge away. But at least try to limit it to situations where it makes sense.
Posted from my Android phone. Oh, I can change this? There, that's better...
100's causes fragmentation and wasted resources and confusion. But i agree some choice is good.
---- Booth was a patriot ----
I always liked to have the possibility of choice. Even if i don't need it.
Having many distros means theres something for everyone. I use Ubuntu (Im new more or less new to GNU/linux) because its easy,
has apt-get and an update manager (And other stuff).
But maybe someone else wants to personalize his stuff to the max (A thing thats difficult on MS-Windows) use gentoo or something like that.
So i think having many distributions is good.
By reading this you agree to give me (Noxn) 1 dollar.
from what I've seen coming out of US high schools and colleges, most are still pretty darn computer illiterate. They know "The Word", "The Excel", and "The Powerpoint" but they know nothing about the basics of what the "tool" a computer is. They are taught to click buttons and not understand what is going on and I'm not talking about coding. Ask anyone what a print spooler is and they'll be clueless. Yet, when they click the print button, that file is sent off to a spooler somewhere and will remain there until the printer finishes. This is a standard part how to use a computer.
So hoping that a more educated society is going to open peoples eyes to Gnu/Linux and open source is dreaming and a very long way out. IMO.
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
"One linux to rule them all" reeks of Architecture Astronauts. Yes oneday we can achieve this, but until you can fit the same desktop on my mini cell phone, you just can't do it.
The limitation is the hardware, not the software. If this wasn't true why is Intel even creating the Atom cpu platform? Why can't the iPhone run true OSX (do you Linux geeks really want to take on the BSD people?)
While we are here talking about all the different versions of Linux, why don't we get rid of all the different models of cars, who buys Ferrari's pfft, we can get rid of those for starters, what about Porsche, just over priced germen cars with the engine in the wrong end for over payed yuppies. Everyone should just buy a Ford or a Toyota 4 door saloon, imagine the money everyone would save on parts and the time wasted on looking for a car thats suits you.
For those with no sarcasm challenged, YES!! I am being sarcastic..
Windows 7 could fail miserably but there's no guarantee that such a failure will help Linux. Perhaps it could finally be the year of Apple on the Desktop!
It's a trade-off. Do we want more flexibility and competition, or do we want to push free software?
I do agree that competition accounts for a lot of innovation. I wonder if this innovation could be maintained if we merged into one desktop distro. We do still have a couple of other heavyweights to compete with.
I think Ubuntu's model is good. They have their main distro, and other side distros to deal with more specific cases. They have a lightweight distro, a totally free one, one for multimedia enthusiasts, one for schools, etc. With this all under the banner of Ubuntu, once the choice to use Ubuntu has been made, all the user needs is to decide what they want to use it for. It also allows them to switch easily, whenever they want. If Ubuntu were the only major distro, then it's actually a far easier decision than trying to decide between the different Windows editions, because they all come fully featured with no lock-in.
Not saying Ubuntu should rule Linux. Just saying that this method has a lot of potential. As it is, I feel Ubuntu neglects their side distros a bit too much. If it was decided to put development behind one distro, then effort could go behind these different flavours, and could result in something truly great.
Of course, this kind of chokes the free software model. Nowadays, if someone has a great idea, they can showcase it in their very own distro. They also have dozens of opportunities to get their idea adopted by existing distros. This would have to be addressed, but I think the boon to free software could be worth it.
It would not be an easy transition, but with care and a clear model of how things would work, it might be doable. Luckily, a thread on Slashdot is pretty much guaranteed to incite change, so we should be seeing this by late next week.
Given the fact that Linux is based on a 30 year-old OS it's not exactly the poster boy for innovation either.
Why can't they use a unified packaging system between all of the distros? In the end, the packaging doesn't matter at all, but rather the content and that it works. These arguments are rather silly.
I also find the argument that Linux doesn't need market share rather short sighted. If there is no public interest due to irrelevance, then Linux will die a long quiet death. You must have interest from both vendors, businesses, and the general user community. Market share gets you that and it is a good indicator of how well you are doing gathering that interest.
I also have to laugh at these Windows jibes regarding all of their 'distros' when in fact, they are all the same base OS, with only the packages installed differing. That is what I see as a critical difference between Windows, OS X, and Linux. Any version compatible software will install and be expected to run on Windows or OS X. The same is not true for Linux.
The community needs to stop stroking themselves so heavily and unite behind a few basic standards like common interfaces for configuration, hardware, networking, packaging. Set a standard that developers, vendors, and users can always assume will be consistent, and then customize on top of those standards.
When the Linux world can't agree on a single standard for filesystem hierarchies (see the huge differences between Suse, Red Hat and its derivatives, Debian and its offspring, or the multitude of contradictions and alternatives in the documented "standard" at pathname.com), or whether or not to use standard Unix tools for controlling which services run at boot (google "chkconfig debian")...
There's no chance you'll ever settle on a single "one size fits all" distro, or even a single "web server" or a single "desktop" distro.
Until you can at least come up with a single credible standard for where to find things in the system, there will always be a need for a sane system like FreeBSD :)
Where is the wisdom we have lost in knowledge?
Where is the knowledge we have lost in information?
Linux is a play ground and a professional opaque arcane tool. Simple and sad but it's a toy or it's an arcane server product that only gurus use. It will never be on a desktop as a mainstream product unless and until there is only one Linux.
Flexibility is great but when you can bend so many ways that you can't tell which end is up you lose all the perspective that you need to be a tool that a desktop user can just use. Take that last run on sentence for instance.
Mac OS X is not that far from Linux, as it's based on something called BSD (I forget which one now) but there is only one that I can use on the desktop. There is a server version but I don't need that on my desktop. So as far as a user is concerned there is only one. I buy software for Mac OS X. I Install that software and I don't have to worry about additional packages too much because in production packages they actually take care of that for you.
Windows is similar but it now has too many heads as well. However there is only one type of executable (I'll ignore 64 bit crap for now) to install and its again all in one package.
Open source is great I use quite a bit of it. But generally I go looking for a commercial package when I want something to just install and run. If I'm in the mood to play I'll try to install an open source package and hope that one of the packager type install things actually works to get all of the dependencies but it usually will not happen that way.
Don't even try to tell me that this tool does it right or that one does it right, I want to work not spend all day looking for all the parts of the one true tool to install everything. You won't be there next time I want to install something and I won't remember this conversation.
Why bother
thinkers rare
doers rarer
thinker-doers rarest
unity of single distro is practically and politically impossible
Aptitude manages package selections far better including remembering that you installed library x simply to make package y happy.
...As does apt-get, since quite a few versions ago. Alias "apt-get remove" to "apt-get autoremove" to get it to automatically uninstall x when y gets removed.
What's purple and commutes? An Abelian grape.
they don't just throw CDs of Microsoft software at them, they tie in training and hardware and all kinds of things. Whatever it takes to keep Windows on those computers.
LoB
"Anyone who stands out in the middle of a road looks like roadkill to me." --Linus
No, they didn't spend millions, since copying a CD only costs a few cents.
Of course, Microsoft counts the donated software at the full market price, but they just provide cheap copies.
Microsoft is just counting on the fact that they are the first ones, but we'll see if this strategy will continue to work once everybody will be more fluent with computers.
Does that work for corporate tax writeoff purposes? Yuck.
You're all such a bunch of nerds, thinking that "fundamental operating system components" and "very good package management" have anything to do with the success of a *desktop* OS. Windows is proof that you only need to be just good enough to be the king.
Many distros = uncertainty for the average consumer. That's the bullet that sprays Linux' brains all over the floor in the desktop popularity war.
Of course, this all assumes that the desktop is the battle Linux wants to win, which I seriously doubt. It's strength seems to be in filling all the niches others can't.
The wrong questions/comments are being made. The problem should not be addressed by "standardization" in the sense of forcing everyone to use one program, in the sense of rules and requirements to only accomplish X. The "standardization" should occur instead by the best meaning of the term, which is to use planning and technology to solve problems, and to establish common communication methods, and for those to then be adopted naturally, without force. APIs/ABIs are the answer, not "only support 'distro' X".
All the problems that Linux has as far as fragmentation, immobility, lack of choice, "unmodularity" basically, can be solved by creating the frameworks/APIs/structures/standards which allow programs to get along, and which allow users to have the freedom to use the software they want, easily. Once those great programs/interfaces/protocols/etc are in place, once the system for running, installing, and developing software for Linux is much more smooth and easier, Linux will be much, much more successful.
Promote true freedom - support standards and interoperability.
The point is that you should be able to have whatever you want. Everyone should. But, the key is that to accomplish that, you don't have to force everyone to use software X. All you have to do is work more on the frameworks/standards/APIs/ABIs/etc etc, methods for common communication, so that you can help streamline Linux. In other words, reducing fragmentation should be accomplished through the use of better programming and planning and communication, not through forcing everyone to do something a certain way.
Once Linux is easier to program for, easier to install software on, and in general more streamlined and gives everyone more freedom and simplicity, Linux will be much, much better and bigger.
Promote true freedom - support standards and interoperability.
I don't see the multitude of distros as being the source of the marketshare issues. It seems to me the interface is really still the problem. It has come leaps & bounds since the 0.9x days, but it's still really lacking.
The apps seems to be coming along and so forth, but it still lacks the polish of a modern GUI. And trying to emulate Windows layouts isn't an improvement. There really needs to be a group of HMI people working on a well thought out GUI standard for it. Realistically, since there are still so many UI options, it's not too late to completely rethink the UI.
That's the killer app I'd like to see on Linux. A really cutting edge interface with a superior set of tools to target it.
That and the installs can still be a pain... I found Debian to be relatively painful from an 'understanding what to do' standpoint. At least as a netboot. I never did manage to install an archived version via internet install. Almost made me wish for my 11-floppy slackware install from back in the day.
The kernel is not bloated, it's just that it comes with drivers for a shitload of hardware.
That _is_ the definition of bloated, since many people use a limited subset of PC-compatible hardware.
Is this a troll? The definition of bloated is useless features you will never use and needlessly wasteful code. Except that ...
The number 1 issue raised against Linux here by the 'turfers and trolls is that it has more limited support for various devices and systems than Microsoft Windows. Generic kernels for distros are going to include drivers for as much hardware as they can just because it makes sense. They want it to boot on as many systems as possible.
Last year, it could be said that Linux ran on more different kinds of x86 systems than Microsoft Vista, not sure what it's like now.
If that's a bother for you, rebuild your own kernel and only include the stuff you absolutely need. Modularize everything (which the distros do anyway) and you can get a small kernel footprint.
Stock Linux *can* be run on embedded systems and the developers go after waste in the core system with a vengeance.
A kernel that will not boot a certain system is definitely 100% "bloat". If that's what you want to call it.
It's nice that you can just stick in almost any piece of kit and Linux detects it and runs with it. It would be nicer if all that cruft was cut out of the base kernel and drivers were available for download on demand rather than shipped with it.
The exotic drivers in Linux do not contribute bloat - they can built as loadable modules. The size of modern kernels first came from TCP stacks (my second home Unix box, System V/R2, had a kernel around 100k and although you could do all kinds of things with it, you could not plug a network card into it, I forget what size the kernel was on my first Unix box, probably about the same).
I've been through this sort of thing personally. When I started off as Mr. XEmacs, it was widely criticized (aside from the usual anti-Emacs venom) for being even more bloated than Stallman Emacs. So, I stripped everything not essential for basic editing and doing connections to a network out (so users could download on demand the emacs lisp modules they needed). The FIRST demand I got hit with was to have everything available in one download, which sort of defeated some of the purpose - every available unused emacs lisp package *does* consume resources.
You cannot please everyone and it's not worth it to try. I'm comfortable with the model that Linus has chosen, but I never had any say in it. He's endured flames from the very beginning for batching together all the Linux kernel source code in one package and resisted attempts to break it up. The way that Linux kernel development has turned out, I would say he is correct (cf Greg KH's recent introduction of half-baked drivers into the kernel.org's .src.tar.gz). I think that's a good thing.
On the other hand, I've had plenty of comments from people who were most happy they could trim their own XEmacs down. You can do the same thing with Linux too, the level of difficulty is about the same, but with a lot more risk than with an editor.
Oh and note for Twitter: I'm not a Linux guy simply because it's the kind of system I've dreamed about and supplied new code and bug fixes for most of my adult life, it's also because I've tracked development on linux-kernel off-and-on for over a decade now and I not only respect Linus for his coding skills, but also for his managerial skills. I am even more amazed that he has learned to balance his life and get married with children (and stay! married) and *continue* to lead.
I agree, multiple distributions are a good thing. Even if they do cause arguments!
http://spamsnake.x-fusion.co.uk