Submitted by “Freedom of choice is one of the great benefits of Open Source Software in general and Linux in particular. This freedom gives consumers the ability to select, without fear of litigation, what software they will use and how they will use or modify it. As a principal, this freedom is extremely valuable. However, a couple of announcements this week seem to indicate that market value of freedom of choice has dipped considerably. The biggest hurdle Linux adoption faced this week wasn’t Microsoft, it was an enemy from within: Linux fragmentation.”
Of course it’s a strength and a weakness to be so open.
But unification would make it stronger. POSIX and the Linux base aren’t enough. I’m thinking about package management especially that duplicates work and incompatibilities accross distros.
That’s right folks; Linux has got where it has because it is too weak to survive!
Film at 11.
Avoid.
Where has it gotten on the desktop?
Only place Linux is really going right now is into the data center, where companies hire people to make sense of everything for them.
It’s a real shame too, because so much about Linux is supperior, it’s just too confusing for normal home users.
I really hope Ubuntu continues its popularity, and becomes what people base all Linux desktops from. It’s already beginning – with Linspire and MEPIS – and it would be far easier for ISV’s to support.
Where has it gotten on the desktop?
Even if only 2% of people use it on the desktop, that represents 15 million out of 300 million computers. Hardly small fry.
Only place Linux is really going right now is into the data center, where companies hire people to make sense of everything for them.
On the contrary, I think the fact that Dell felt it necessary to do as much as they did (and the fact that Linux was number one on its list) shows that the best is yet to come on the desktop.
It’s a real shame too, because so much about Linux is supperior, it’s just too confusing for normal home users.
Windows can be plenty confusing, too. Those who don’t have a guru to hand either must be very flustered with their systems, or limit themselves to email, web browsing and maybe games.
I really hope Ubuntu continues its popularity, and becomes what people base all Linux desktops from. It’s already beginning – with Linspire and MEPIS – and it would be far easier for ISV’s to support.
No thanks. This is the biggest problem with the “we need a Linux monoculture” theory: Just about every distro you choose to base this mythical “Linux for everyone” on is going to have its rabid fans, its ardent detractors, and the rest of us caught in the middle. Personally, I think Ubuntu is crap.
I am hardly a rabid fan… and I also don’t care much for Ubuntu either…
However I do have several family members running it, simply because it’s easier. Hardware is well supported, their gagdets work fine, and they don’t have to worry about drivers or virii etc, so they’re very happy.
Ubuntu is stable, it’s clean… and it has plenty of documentation. What more can you ask for in a distribution aimed at non-geeks?
Ubuntu is stable, it’s clean… and it has plenty of documentation. What more can you ask for in a distribution aimed at non-geeks?
Ubuntu is stable, it’s clean… and it has plenty of documentation. What more can you ask for in a distribution aimed at non-geeks?
My first hand experience of Ubuntu indicates that its hardware support is crap…As for the documentation, I hear more complaints about it than praise.
What were you trying to get working?
My Mom can understand the documentation, and use it to help her… so I have no such complaints…
She’s about as technophibic as it gets…
May I refer you to the entry in my blog:
http://latedeveloper.org.uk/2007/02/colour-me-brown-and-unimpressed…
In this blog:
I’m not sure why you’d want /boot to be anything other than ext2 anyways… seems kinda overkill?
I can’t speak about the LiveCD installed… I still use the alternative install media (debian-installer).
Doesn’t say much about actually giving the system a real shot though…
Doesn’t say much about actually giving the system a real shot though…
No, but then if the system isn’t compatible with the hardware, why bother trying to?
Sorry to hear about your (obvious) Ubu installation tortures.
I on the other hand, went out and bought a brand new 80Gb. Western Digital HD,
and gave the whole danged thing to Ubu 6.10 – and have had no problems!
I’ve always preferred to keep difrnt Operating Systems on difrnt hard drives:
– none of this partitioning nonsense for me.
Call me simple-minded, but call me.
I’ve always preferred to keep difrnt Operating Systems on difrnt hard drives:
– none of this partitioning nonsense for me.
Yeah, good idea if you have the money and the space!
Call me simple-minded, but call me.
Heh 😉
What “documentation?” I’m not saying it’s hard to find info on ubuntu, but I think you’d be hard pressed to find an official ubuntu handbook online _ I could easily be worng, but I’ve ceratainly never seen one. Sure, there have been several books released about running ubuntu, but that isn’t really documentation.
Look to the BSDs if you want to know what documentation looks like. Other than Gentoo and redhat (and perhaps SuSE – I don’t know since I haven’t used it since 8.2) I’d say that, in general, Linux documentation bites the bone. And, yes, my OS of choice *is* linux.
I am hardly a rabid fan… and I also don’t care much for Ubuntu either…
However I do have several family members running it, simply because it’s easier. Hardware is well supported, their gagdets work fine,
and they don’t have to worry about drivers or virii etc, so they’re very happy.
Ubuntu is stable, it’s clean… and it has plenty of documentation.
What more can you ask for in a distribution aimed at non-geeks?
EXACTLY! – Sheesh..with a bit more ease-of-use, this Ubuntu thing could actually get to be semi-Widely Popular. Ubuntu 6.10 managed to grab me pretty well, and I’m a lot dumber than you guys!
2% of 300 is 6, but anyway…
I’m curious to see how that Dell thing works out. I suspect that 90% of those who asked for Linux pre-installed on Dell workstations are linux fans who have no plans on buying Dell workstations with Linux pre-installed.
I’m curious to see how that Dell thing works out. I suspect that 90% of those who asked for Linux pre-installed on Dell workstations are linux fans who have no plans on buying Dell workstations with Linux pre-installed.
True. Though I suspect it would be a different story if they had chosen a different distribution, such as Ubuntu. (I would replace Ubuntu with something else, but I’m not giving any money to Microvel.)
“I’m curious to see how that Dell thing works out. I suspect that 90% of those who asked for Linux pre-installed on Dell workstations are linux fans who have no plans on buying Dell workstations with Linux pre-installed.”
Absolutely, but you should ask yourself the reason why.
Dell has had an interesting relationship with Microsoft in the Past. The suspicion that you can by the same product *with* windows for cheaper, and Dell would loose there add-on market.
The reality is though with *any* Linux distribution installed. I would buy it and put *my* distribution o choice on becuase, my biggest problem with choosing dell for Linux is simply becuase I don’t know whether it would work or not.
At the end of the day unless Microsoft is a *paid* add on for a dell machine not part of its price, that is simply enough to remove or replace why would I be *interesting* in subsidizing Microsoft?
“Even if only 2% of people use it on the desktop, that represents 15 million out of 300 million computers.”
Although you think “Ubuntu is crap”, at least it comes with a working calculator…
class comment 😉
Even if only 2% of people use it on the desktop, that represents 15 million out of 300 million computers. Hardly small fry.
That depends on who you talk to. 2% of the market isn’t enough to catch the attention of large software makers or hardware makers, for example.
Ubuntu 6.06 may be crap, but Ubu 6.10 is the first Linux flavour in SIX YEARS of dabbling, that I could get my head around for longer than one frustrating week(and then giving up..AGAIN)
I’ve not booted my XP drive in about a week – that says a lot (in fact my XP drive is phukt..I couldn’t boot it even if I wanted to – and I’m NOT worried!)
Ubuntu is OKAY, says this Windoze user of 6 yrs.,
(and Mac user for 5 years, before the conversion to Windoze)
Ubuntu clearly works well for many people, and I’m not suggesting no-one should use it. All I’m suggesting is, it’s no closer to being the mythical LinuxForEveryone than any other distro.
There is no such thing as an OS that is not confusing for a normal home user and that includes having it pre-loaded. The only operating system without confusion is a brick.
Where has it gotten on the desktop?
Right up there with Apple? All with a fraction of the resources, experience, and time?
Still there is no Ableton, Photoshop, Premiere, Aftereffects, InDesign, Poser, Bryce, etc version of popular Apps that are available on OS X and Windows despite an equal marketshare as OS X has and despite not having realy competing oss apps.
This could very well be due to the fact that coompanies don’t feel like supporting 100+ incompatible Linux Distributions.
Install Cinelerra to see why diversity doesn’t always have to be an advantage.
I don’t see what would be so horrible about at least settling for a common installation mechanism.
And just where is Linux today? Except for the coffee shop around the corner and one of my own PC off and on, I can’t remember seeing another computer running it. Time for a reality check. Fragmentation is not strength.
And just where is Linux today? Except for the coffee shop around the corner and one of my own PC off and on, I can’t remember seeing another computer running it. Time for a reality check. Fragmentation is not strength.
I’m not going to be drawn into another argument like the one in the article. Your argument has been bandied about since time immemmorial, and has no more merit now than it ever had.
Some would prefer that it thrive, not simply survive.
Just where has Linux gotten? It’s available for no cost, yet it still is on just a tiny percentage of systems. People still pay money for Windows and OS/X rather than running Linux.
If you were giving cars away for free and General Motors still had a bigger share of the market, wouldn’t you be concerned about the quality and/or public perception of your cars?
Look, you people keep replying to me on a subject I’ve already given my thoughts to. Come up with something new or they aren’t changing.
Some would prefer that it thrive, not simply survive.
Like me for instance.
Just where has Linux gotten? It’s available for no cost, yet it still is on just a tiny percentage of systems. People still pay money for Windows and OS/X rather than running Linux.
Windows comes preinstalled on PCs and OS/X on Macs, therefore most people think of it being free.
You know what I think Linux needs to become an even bigger success? A good marketing strategy. It does NOT need to merge in order to survive, any more than we need toasters to come only from Toasters, Inc.
If you were giving cars away for free and General Motors still had a bigger share of the market, wouldn’t you be concerned about the quality and/or public perception of your cars?
If we were giving away cars for free and General Motors still had a bigger share of the market, it would be a pretty big indicator that GM had a monopoly on cars.
Concerns about the quality of Linux relative to Windows? Gimme a break.
“If you were giving cars away for free and General Motors still had a bigger share of the market, wouldn’t you be concerned about the quality and/or public perception of your cars?”
Well if that actually happened….I’d be concerned about GM being in bed with the oil companies and how they changed and patented a new fuel fill hose design that is available only on GM cars. Because when you get down to it the GM cars are designed pretty much like any other car, but if you can’t get fuel so you can take your car where you want to go you’re forced to use GM cars (or carry an OSF “open standard funnel” to use to refuel).
I dont really expect too many companies to commit to pre-installing Linux at this point. The real hurdle that linux needs to be overcome is getting vendors to support linux friendly hardware. Being able to buy a machine that is completely linux compatible and doesnt come with Windows preinstalled is huge. Also, I can see linux coming pre-installed as part of an appliance type device from ISP’s. I think this would make them a lot of money actually.
with the bad.
I think the good far outweighs the bad. I also feel that the competition amongst Linux vendors if healthy for Linux as a whole.
Hopefully the Linux community continues to stride for innovation, and not sacrifice itself for short term adoption. <Yes I’m hinting at the Novell deal here.>
Today, Dell ships their non windows boxes with FeeDos on separate media, i.e. it is not preinstalled.
They probably would sell more of these boxes if they replaced FreeDos with any version of Linux. Preferably some free distro containing only free drivers. They could then list what functionalty that worked out of the box with that distro. They should also provide a hardware a chip level hardware spec.
That way people would have a good idea what was working or not regardless what distro they prefer to install once they have bought the box.
freedos is very popular bundle with many computer dealers i seen (in europe too), which is kinda weird imo, since linux would make more sense.
When Dell goes to setup a PC for a Linux users there’s one central concern: That the hardware is supported, preferably in the kernel but at least fairly easily.
If you get the OS pre-installed it tells you that the devices all work… Otherwise they should be telling you “everything but this works.”
As the article stated, Linux (the part that supports hardware) isn’t fragmented: This is the part that we want Dell to get working, especially on notebooks.
They don’t have to please everyone with their distribution choice, they just have to pick one and support the hardware…
As usual, people are using fragmentation as an excuse where it isn’t one. It makes much more sense to use it as an excuse for not porting to Linux: Their actually is a lot to keep track of to support client applications on Linux; but that also keeps you honest where some Operating Systems have had a hell of a time getting their developers to install and design programs in a proper way (*cough* annoying UAC dialogs over bad applications that assume administrator rights for things they don’t need *cough*).
There may actually be fewer major distributions today than there are versions of Windows Vista (major is pretty subjective though).
Mandrake (maybe)
Ubuntu
Suse
RedHat/Fedora
Debian
Gentoo
A couple others…
Of course, the differences between different versions of Vista is usually so minor it’s almost insulting.
Why is Gentoo considered here?
Gentoo will never be mainstream, because no normal sane user wants to compile everything – or even understand why it is a good idea.
No disrespect to Gentoo, for their audiance, it’s a great distro…
Also, Red Hat and Debian don’t target desktop use at home… so they shouldn’t be included… (Debian happens to be my favorate distro however.)
Mandriva is insignificant these days…
Really it’s SUSE vs Ubuntu … at least for home desktop use.
That should probably be SuSE vs Ubuntu Vs Fedora.
Personally I use Slackware, but would probably recomend Ubuntu to anyone who asked for a good linux desktop distro.
Fedora aims to be too pure…
I consider it more of a hobbyist/geek distro than anything… they have a lot of features that really aren’t ready for home users (SELinux for instance over-complicates things.)
Really it’s SUSE vs Ubuntu … at least for home desktop use.
I agree that those two are the big players, but what about companies like Linspire and Xandros? They have been in the Desktop Linux camp since day-one.
Linspire is moving to Ubuntu…
Xandros hopefully will join suit…
Really.. it’s SUSE vs Ubuntu … at least for home desktop use.
Yup: that’s about it!
Compile? What am I? – a Ph.D in Computer Science?
I’m a dumbass Windoze Refugee!
I want to use the computer…not have the stupid thing use me!
It’s even fewer than that. If you’re talking about device driver support, the only thing you need to validate is the Linux kernel since distros stay pretty close to it. Pick a kernel that’s the oldest of Debian, the last Red Hat Enterprise Linux, and Novell SUSE Enterprise Linux and test it. If your hardware tests passes with that kernel, declare that your hardware supports kernel versions X or greater which should be in most Linux distros (contact your distro if any incompatibilities).
If your hardware tests don’t pass and you can’t find substitutions that do pass, then either state that the hardware is not compatible with Linux (it should save some headaches), or better, test out with more modern versions of the Linux kernel and if you find one that passes, state which version you need and have the caveat that it’s not widely available in all distros so it’s your responsibility to confirm.
It’s that simple. The Linux crowd is more technical than the Windows crowd, so they’ll be able to handle this level of detail and understand the caveats and issues involved.
As time goes on, they may want to preinstall a distro like Debian or Ubuntu to show give confidence to their “it runs Linux” statement, but that’s not needed initially.
he biggest hurdle Linux adoption faced this week wasn’t Microsoft, it was an enemy from within
The article refers to Dell’s “reasoning” why they won’t pre-install Linux, however the claim that there is no single best choice for a Linux distribution is not an argument at all, it is a pretty neat excuse.
True, quite a lot of customer will likely install a different distribution than the one which came pre-installed, but any other distribution is a very good starting point for this:
– one can check which kernel version they are using and which configuration they have it compiled with
– one can check whether there are any patches applied to the kernel source
– one can check which modules are contained in the initial ram disk
– one can check which hardware loads which drivers
– one can check how the configuration for all parts of the pre-installed system look like, e.g. parameters used in xorg.conf
– assuming a minimum installation of the pre-install distribution is kept as a kind of dual boot, you can get technical support that can be directly mapped back to your own choice of distribution but still have the customer support deal only with one distribution’s tools
Well, we all knew that would be resorting to this standard excuse, didn’t ww?
Maybe DELL has other reasons to say this (which Microsoft has nothing to do with, of course :/ ), but a lot of us live in a world where we don’t have time to make sense of that list you just posted. We have businesses to run or jobs to do, and the computer is a tool to that end. It is not the end. I have to keep up with government regulations and developments in my own field. My computers need to make my life easier and more efficient. I don’t need more obstacles to overcome.
I used to run Linux (SUSE) as my desktop in the 90s. But I’ve reached the age where I don’t want to learn about the ins and outs of different distributions. Life is just so much easier (on the desktop) now I’m using Windows. The distro market really needs an 800lb gorilla that becomes the de facto standard, that everybody can rally around.
When you’re new to Linux the fragmentation is exciting, but after many years it just becomes a pain and a bore.
But I’ve reached the age where I don’t want to learn about the ins and outs of different distributions. Life is just so much easier (on the desktop) now I’m using Windows. The distro market really needs an 800lb gorilla that becomes the de facto standard, that everybody can rally around.
Yes I have reached that age too. I just run Ubuntu LTS, works better than XP. I am happy, seems like you are living in the past as far as desktop Linux goes.
Ironic…
LTS is almost a year old
Ironic…
LTS is almost a year old
And it will be supported for four more years
Not that I intend to keep running it that long 
too bad it’s not supported by oss devs. Forget about finding some fresh oss software for it. True especially for non-core desktop apps.
Supported means backporting security holes, not newer versions of sw that fix bugs making you pull your hair off.
Edited 2007-03-05 20:43
too bad it’s not supported by oss devs. Forget about finding some fresh oss software for it. True especially for non-core desktop apps.
Not exactly. I have just updated to the latest version of Zim (a fantastic original application BTW and much better than Tomboy) from the old one in the Dapper Universe repository.
I have also recently installed Tracker which was not even in the original Dapper distribution and I am about to install the new Mirage viewer again not in Dapper. There are repositories for these that provide binaries for the latest stable versions for both Edgy and Dapper.
Of course when Feisty comes out the Dapper support for new apps will rapidly begin to fade. That is one of the reasons I will upgrade to Feisty when it is released.
Edited 2007-03-06 06:00
OK, but can you honestly state that the effort and expertise required to find and integrate those components was realistic for nontech user? (knows that software is installed but not much above it)?
OK, but can you honestly state that the effort and expertise required to find and integrate those components was realistic for nontech user? (knows that software is installed but not much above it)?
Yes – it’s not as if I am suggesting he compiles them from source.
I used to run Linux (SUSE) as my desktop in the 90s. But I’ve reached the age where I don’t want to learn about the ins and outs of different distributions. Life is just so much easier (on the desktop) now I’m using Windows. The distro market really needs an 800lb gorilla that becomes the de facto standard, that everybody can rally around.
FreeBSD would fit that shoe perfectly.
FreeBSD would fit that shoe perfectly.
Yeah, apart from the fact that the hardware support is even worse than a bad Linux distro’s, and the BSD licence allows freeloaders to take proprietary advantage of any work you might want to contribute to the project.
Yeah, apart from the fact that the hardware support is even worse than a bad Linux distro’s, and the BSD licence allows freeloaders to take proprietary advantage of any work you might want to contribute to the project.
“Oh woe is me, someone might take the code I contributed to the community and use it for something, boo-hoo. Someone might use the work I gave away without bending over to kiss my copyright, how will I ever live with myself!?” Get over it, you put the code out there for people to use, otherwise you would have just changed it on your home PC, quit crying about it. This “I have to fan my ego with my code” crap is annoying, I hate you all.
That aside, my computer has better support under FreeBSD than Linux. RT2500 USB support built-in, rather than having to download an outside open-source driver (hard to download something for your non-working network card >.<), SATA support is far more stable in FreeBSD (on my chipset anyway), USB/PS/2 mouse integration has worked together for far longer than it has under Linux (yes, it always worked, but was a far greater pain to configure in 2.4.x), file system performance, last I checked, was roughly 2.5x higher on UFS than ext3 (again, this is my system, drives, chipsets, and tuning I’m sure make a difference, as I said SATA support for my card in Linux is flaky).
FreeBSD has the disadvantage, of course, that you have to actually know something first (i.e. You need to read that handbook first, which takes all of 15 minutes. 15 MINUTES!! OH NO, THE HORROR!) and know which hardware you have (lspci/pciconf -lv anyone?) but, yeah, having to knowing your ass from a hole in the ground is totally “worse hardware support than the worst Linux distro”, if by worse you mean worse for lazy know nothing gophers. Come on, do you really want to run a kernel with SCSI support required for SATA cards? Really?
But, yes, that is a correct statement I suppose, if you change it to “crap hardware has crap support in any non-crap kernel.” People, if you are going to buy crap shoddy hardware… then you shouldn’t expect it to work under *BSD. That’s why they only had really good support for SCSI for so long. Would you buy a $1 air filter for you car and then wonder why mechanics won’t bend over backward to make the rest of your car perform well with it? (hint for the slow: You have to reduce horsepower and generally make things chuggy and slow so that it will perform reliably with something bad. Same idea, only with “software-stability/code-bloat” rather than “air-intake/fuel-efficiency”)
If that’s an example of something you think worth reading, I hate to think what you consider as crap.
Two small points however:
1. When I contribute code, I expect it to be used, and improved and/or paid for. I DON’T expect it to be mashed into some proprietary crap.
2. I have installed FreeBSD before WITHOUT reading documentation. That’s right, it is THAT easy. However, no amount of reading is going to make hardware work that does not work due to support limitations.
Edited 2007-03-05 18:38
But, yes, that is a correct statement I suppose, if you change it to “crap hardware has crap support in any non-crap kernel.” People, if you are going to buy crap shoddy hardware… then you shouldn’t expect it to work under *BSD. That’s why they only had really good support for SCSI for so long. Would you buy a $1 air filter for you car and then wonder why mechanics won’t bend over backward to make the rest of your car perform well with it? (hint for the slow: You have to reduce horsepower and generally make things chuggy and slow so that it will perform reliably with something bad. Same idea, only with “software-stability/code-bloat” rather than “air-intake/fuel-efficiency”)
Blah, blah, rant rant. Is £700-worth of hardware (~US$1400) crap? In the real world not everyone can afford to pay £2200 for the very best hardware, if that’s what’s needed to run FreeBSD. Most people would be much better off getting hardware they can afford (yeah, I know, strange concept) and running Linux on it.
However, I will say this for FreeBSD: I like it much more than I like many of its users.
>the fact that the hardware support is even worse than a bad Linux distro’s
Some urban legend is the credo in every discussion about the disadvantages of Linuxism? OMG.
I could tell you about the quality of mere hacks or real quality, but this would be end in bashing.
With *BSD, you have to think. You should know what you’re doing.
>and the BSD licence allows freeloaders to take proprietary advantage
This I call *real* freedom. I would be a hypocrite telling people about freedom and denying something at the same time.
After many years… I’d hope you’d have settled on one…
I hope you’ll forgive me for asking but the first thing that struck me about your statement was, if you “used to run Linux (SUSE) as my desktop in the 90s” then why would you need to learn about the ins and outs of different distributions?” I mean, Windows is absolutely a legitimate choice and that’s fair enough, but you had already settled on a distribution in the first place.
Or maybe I missed something. That happens to me sometimes.
Edited 2007-03-05 17:46
I hope you’ll forgive me for asking but the first thing that struck me about your statement was, if you “used to run Linux (SUSE) as my desktop in the 90s” then why would you need to learn about the ins and outs of different distributions?…
It’s in the nature of the Linux enthusiast to always keep an eye open for alternate distributions just in case the grass is greener elsewhere.
It’s in nature to not reboot…
How exiting to see all this differents answers by people tired of managing distros internals. One chose Ubuntu, the other freebsd, the third Suse … I ve personnaly chosen Kubuntu.
Isn’t it about Linux fragmentation ? What s interesting here is that we ve almost all chosen Commercial supported Linux distros but at the very begining we had the possibility to tweak and choose, break and learn.
The LSB made all of this more easy and with us people going more mature, Linux will become the same.
The fundamental problem is simply that you as a user need to have intimate knowledge of your system to make it function.
This reared its ugly head in the most fundamental way to me the other day. I was installing Sun’s Glassfish on a CentOS machine. Glassfish is 99% Java, but Java 5 and earlier have, simply, a terrible console interface. Specifically it can’t read from the console with “noecho” for things like passwords. In this specific case, the Glassfish installer has a JNI native library bundled to allow it to read the console with no error.
And what did I get when I ran this? An error message that it couldn’t find stdlibc.
So, here we have a top tier software comapany delivering software, running on, I assume, a reasonably close “top tier” Linux distribution (CentOS ~= RedHat), and it couldn’t find something as fundamental as stdlibc. The core of the system. Something that EVERYTHING needs.
On Windows, this would not have happened. On Mac OS this would NOT have happened. On *BSD, this WOULD NOT have happened. On Solaris this WOULD NOT HAVE HAPPENED.
Period.
On a popular Linux distro from a knowledgable vendor (no Sun conspiracy rants please), a simple install against a fundamental library failed.
That’s insane. And it’s a waste of time.
And events like this continue to sour my experience with Linux distributions. I would rather be a casual user of the operating system than have to be an expert user.
On a popular Linux distro from a knowledgable vendor
I wouldn’t say CentOS is a “popular distro”. Fedora, Ubuntu, SuSE and Mandriva, sure…
Have you tried to figure out where the error lies? As far as I know, the standard C library is in the same place for *all* of these distributions. Such an error is quite surprising…it seems to me the problem here is with Glassfish, not with the distro.
But of course, it’s *so* much easier to blame Linux than blame the ISV for this, right? This is not a “conspiracy rant”, but to use your own words, if a “simple install” from a “knowledgeable vendor” cannot resolve a “fundamental library” link, then I think either the “knowledgeable vendor” made a mistake, or the problem is situated between the chair and the monitor.
(Oh, and as far as Windows never having issues with Java…I’m sorry, but I have to laugh!)
If centos declarations about compilance with RHEL are true it’s equal to largely the most popular distro for the enterprise, that is where you’d like to install an application server.
“you as a user need to have intimate knowledge of your system to make it function.
This reared its ugly head in the most fundamental way to me the other day. I was installing Sun’s Glassfish on a CentOS machine. ”
This is not an end user writing…. Or, not one that I ever met.
We can say the same about Windows users.
Windows95 users, Windows98 users, WindowsME users, WindowsNT users, Windows2000 users, Windows2003 users, WindowsXP users, WindowsVista users and that was without taking care about its subversions like WindowsXP HomeEdition users, WindowsXP ProfesionalEdition users, WindowsXP MediaCenter users and so.
Doesn’t Dell got confused by that?
Edited 2007-03-05 17:01
No, you should rather compare it to Ubuntu Breezy, Ubuntu Edgy, Ubuntu Feisty, etc… Which are versions of the same distro.
Is Dell, or anyone else, pre-installing Win85, 98, NT, ME, or 2000 (or even XP)?
How many editions are there of Windows TODAY?
How many distros are there of Linux TODAY?
Try again.
“Is Dell, or anyone else, pre-installing Win85, 98, NT, ME, or 2000 (or even XP)?
How many editions are there of Windows TODAY?
How many distros are there of Linux TODAY?
Try again.”
Yes and DOS
About 15 maybe more including CE, embedded XP and a couple of ‘N’ editions, retail, OEM, upgrade. Depending on how Crippled; Additional Craplettes; Restricted you want your License.
Very few that fit the bill i.e. Widely Used; Recognized; With Support; Complete enough…I can think of three. Although I would be astonished if Dell couldn’t pick a partner that best suited them, or even create their own Distribution, based of a couple of package Managers.
“Windows95 users, Windows98 users, WindowsME users, WindowsNT users, Windows2000 users, Windows2003 users, WindowsXP users, WindowsVista users and that was without taking care about its subversions like WindowsXP HomeEdition users, WindowsXP ProfesionalEdition users, WindowsXP MediaCenter users and so.
Yet you can create one installer and it will work on all of those Windows versions. Try doing that in Linux.
”
There are only *two* of main package managers that are used by distributions.
I’m sure you are aware of the vast number of different installers available on the Windows Platforms.
Windows95 users, Windows98 users, WindowsME users, WindowsNT users, Windows2000 users, Windows2003 users, WindowsXP users, WindowsVista users and that was without taking care about its subversions like WindowsXP HomeEdition users, WindowsXP ProfesionalEdition users, WindowsXP MediaCenter users and so.
Yet you can create one installer and it will work on all of those Windows versions. Try doing that in Linux.
No problem. We just need to write a new installer framework.
It’s not a hardware/driver issue.
Dell would have no problem brewing up a batch of hardware that worked with most major distributions. The problem is, if Dell pre-installed Linux, it would have to support it. I do not believe Dell would be ready for this. Most Dell (and other company) technicians I have met are Windows-only folks. They are proud of their Windows skills. Same for the “Geek Squad” at Best Buy. It would be hard to staff up with Linux-savvy technicians. In addition, they really should have people skills, and not be caffeine (etc) addicted moles who live in dark corners!
I believe this can only change when schools (especially high-school and college) routinely use Linux, so that the next wave of technicians enters the game with established Linux skills.
“In addition, they really should have people skills, and not be caffeine (etc) addicted moles who live in dark corners!”
Yeah, because the Windows “experts” at Dell and Best Buy are paragons of social interaction, right? Right?
Please try to avoid needless stereotyping, thanks.
Well, I do believe from my experience in both worlds (Window Developer / Linux Enthusiast / LUG member) that there is some truth to this problem. In the Linux world, things tend to be a little more of the “free and easy” mentality. Windows developer meetings tend to be very structured, while LUG meetings tend to be more disorganized (but a lot more fun!).
My intent wasn’t to label people, but to point out a challenge for the Linux enthusiast community. We are going to have to get a little more organized and give up some of our wild-n-wooly freedoms if we want to have an impact in the larger world. I’m one of those caffeine-addicted moles, but I am starting to learn some of these basic truths. They aren’t fun, but they are necessary.
If Linux is to make it at the Dell pre-installed level, there must be more “regular” folks who use it, and not just us fringe folks who don’t mind using disklabel to setup our OpenBSD laptop (which I do!). I do beleive the Linux crowd does currently attract a more elitist, DIY crowd.
As I mentioned, this will change when Linux becomes a more mainstream OS. Also, as I mentioned, this will only happen when Linux becomes a normal OS to use in school, so all kinds of people get familiar with it, not just the geeky folks (of which I am a proud member!).
Edited 2007-03-05 17:40
So, not to be an eternal nag about this, but why does everyone seem to need the grandeur? If it works for you, fine, it works for you, but you don’t *need* dell to pre-install it, you *can* do it yourself.
You don’t *need* linux to gain acceptance, or “get out there in the world” as if the number of people who use linux has any affect on you at all.
So, your point is “if more people support linux commercially we will have better hardware support!” right? I know you didn’t say that directly, but “we need other people to use linux” comes down to either that or “I’m insecure in my choice and need someone to validate it for me!”, and I assume that is not your stance.
In any case, waiting for commercial support is a stupid thing for you and everyone else to wait for. Instead, go out of your way to only support people who already support it in one way or another, don’t expect the people who don’t to magically change their ways.
ATI may eventually have non-crap support now that AMD bought them out, but for the time being it is both easier, wiser, and more productive to just buy something with an nVidia chipset. Then, you can send off a lovely letter to ATI about why you didn’t buy their product. Same idea with every other bit of hardware. If Dell won’t support it without you asking for it then, you know, don’t buy a Dell. It’s crap hardware with crap support anyway, consider it a blessing.
Here, I’ll show you:
Dimension C521
$409
Sempron 3400+ (crap), Nvidia Geforce 6150 integrated (integrated=crap), 512MB DDR2 533mhz (DDR2 is crap anyway, but I guarantee it’s bottom barrel memory), 16xDVD-ROM (not even a burner), 80GB SATA hard drive (SATA 1, so crap chipset and slow drive).
Great, crap, crap, crap+crap, old/less-useful, crap
Compared with some random parts I pulled out of my bum (all with FreeBSD support, which pretty much guarantees linux)
Athlon 64 3500+ (25% faster on benchmarks), Motherboard with SATA2 (supported chipsets, SATA2, sound, etc), 512MB PC2-6400 (800mhz Samsung), Geforce 6600 (256MB), case with 450W power supply (Dell usually ships with bare minimum, i.e. 280W in a P4 computer), 16x DVD burner (with lightscribe), and an 80GB SATA drive… all of it 100% supported under FreeBSD/Linux/whatever, all at least 25% better hardware, and it comes to a whopping $352
$352, without my reseller discounts, straight from some other guys store. That leaves you $50 to pay the local geek kid to assemble it for you.
So, why are you all here babbling about Dell support again? You could always, I don’t know, support your local computer shops that sell better hardware and can make something that already works for you, instead of asking dell to do it?
Oh, that’s right, that requires thinking for longer than 8 seconds and not buying on impulse just because dell puts a $100 mail in rebate on their crap in a box.
What you are saying condenses down to “it works for me, so forget what others are doing”.
Well, I don’t agree. I actually think switching to Linux, and the Free Software philosophy underneath would be good for society, and especially good for our children.
Also, if this country doesn’t get out of it “big software in boxes” mentality, I’m afraid this country will slip behind much of the rest of the world.
Condensed version of my post: “I _AM_ my brother’s keeper”.
I understand your point, but Linux has already started to move out of this area. I myself have never been to a LUG meeting, and though I am certainly a geek, you’d never guess it by looking at me. Not only that, but I work with a lot of developers working on Windows and they look just as geeky/weird as the typical Linux stereotype (all right, I work at a video game development studio, but still…)
I do agree that the geek image is hard to shake off, but the times are a-changin’…
I have a confession to make: I didn’t read the article.
Because articles with titles like these are almost always irrelevant.
“If only there were fewer distros Linux would gain more traction!”, they say.
Yeah, and if we could eat rocks we could use them for food.
Bemoaning the level of diversity, and even debating the costs and benefits of that diversity, is a complete waste of time.
Because no matter what conclusions you come to… *its not going to change anything*.
The Linux community is what it is because that is its nature.
Forget it. You can’t change the terrain simply by saying that it is or is not the best terrain. The terrain is the terrain.
So the real question is how to best take advantage of the terrain just as it is, and as it likely will be on into the foreseeable future.
That is where the conversation should start.
The rest is useless hand-wringing.
Bravo. This is pretty much exactly what I was going to say. These types of discussion are sterile. If Dell wants to provide Linux, all it has to do is pick a distro…the fact that *one* distro will be supported will make it easier for other distros to be installed anyway.
Personally, I think Mark Shuttleworth should give Michael Dell a phone call. Ubuntu’s got the best brand and one of the best desktop distros out there, those are non-negligible factors to consider…but even if they’d decide to go with Novell/SuSE it would still be nice to have Linux-friendly Dell machines. As a Ubuntu user, I could live with that, and would seriously consider buying a Dell as my next PC.
Right on. Diversity is what drives innovation on the Linux Desktop. We’ve see how monocultures encourage stagnation and lead to the consumer paying over and over again for the same old crap. The proprietary software industry has been trying to replicate the revenue model of the mainstream media content industry, but piracy trends demonstrate that these models are flawed.
The value of software isn’t in the bits or LOCs. The value is in the fact that it works properly and that you can get support when you need it. Software has become so central to our daily lives that it is inconceivable to use software for which you can’t get support. Even the technical elite wants an active community to provide insight into some unusually hairy stuff. Providing guaranteed software support is where the value lies.
With razor-thin margins on hardware, you’d think that Dell would jump at the chance to sell support services to relatively savvy Linux early-adopters. There is currently no major consumer OEM that sells a Linux system will comprehensive single-source software and hardware support. This is a potentially HUGE market! But instead, Dell prefers to let the Linux vendors sell the software support, which means less money for them and more for the Linux vendors. Too bad for Dell. I’d rather the money go to a company that directly invests in Linux anyway.
EDIT: And what kind of editorial director for a Linux-specific publication doesn’t know how to spell Ubuntu? Or describes Linux distributions as “flavors of Linux,” as opposed to Linux, which is apparently something you can run on its own if you wanted (without init or glibc I presume)? These kinds of articles seem to be written by one of three kinds of writers: partisan hacks, know-nothing hacks, or hacks that don’t even know how to write a coherent article. This guy is the know-nothing sort.
Edited 2007-03-05 19:08
It’s not imminent to OSS.
BSDs are oss too and do *much* better job dwarfing fragmentation.
Besides there are 2 kinds of choice and freedom.
Developers choice of using whatever toolkit/framework/library is sexiest at the moment and whatever version is around and freedom to be carefree about such mundane tasks as maintaining binary compatiblity.
There’s also user’s freedom to user whatever software s/he needs on the os of choice without many hurdles.
Also the freedom use and maintain his/hers system without running too problems requiring expert knowledge (useless to non-it people) that are typical to “house of cards” systems that majority of linux distros are.
These two are in striking contradiction.
Really, if linux cannot live up to the task of decently supporting non-it-expert user with easy to use and maintain,efficient, dependaple and open system it’s time for all people who believe it’s their mission to move on to something else. Maybe there are other systems that could use community engagement better for achieving such goal.
Edited 2007-03-05 21:17
Well, I specifically used the term “Linux” and not “OSS” for exactly that reason. However, now that you have brought it up, I will address it.
A certain degree of diversity, which some would call fragmentation, *is*, in fact, inherent in the OSS model.
*BSD (and note the wildcard at the beginning) is not all of OSS. The BSD’s are a few players among quite a number of other players that make up the OSS community. Even if you consider them as a whole to be one player, they are still *adding* adding to the “fragmentation”, not reducing it.
Furthermore, from the standpoint of popularity and getting themselves noticed (and I know I’m likely treading upon thin ice, here) they have not been nearly as successful as Linux in this ecosystem. Limiting their diversity internally has not been an adaptive strategy for the BSDs *if* the goal was to compete for popularity.
Do you see that your post illustrates, very nicely, why OSS is as it is?
You are saying that you know a better way than what is being done by Linux today; That the *BSDs would do it right if only they rose in popularity and name brand recognition.
Now, you may be right, and you may be wrong. I don’t care. And it doesn’t matter, because the world is as it is, not as we think it should be.
The important thing to note is that if the BSDs *do* start gaining more traction, and more name brand recognition, and become a choice that people have to consider right along side Linux… you have just substantially increased the “problem” of fragmentation in OSS.
The road to fragmentation is paved with the projects of people who have wanted to end fragmentation once and for all, and have started their own be all and end all project to do it. Each thinking that his idea was so good, his project so compelling, that everyone would agree and start doing things his way.
And *that* is why OSS is as it is and will continue be so.
And arguing against that, in my opinion, is about as effective as arguing against the second law of thermodynamics. The two have a lot in common.
You assume that software wins solely on the base of its technical merits.
MS example shows that it is simply not the case.
Marketing matters, even in community driven world where its role is filled by hype, word of mouth and grass roots journalism.
Better alternatives can pop up any time (and I believe have already) but without the whole hype machine they can simply live and die unnoticed as everybody (that cares) sits there waiting eagerly for the year of linux desktop to come.
There are various motives for people to commit their time for producing software. In many cases they are based on vision of providing free, open and component desktop OS. It is where (perception of) suitability of linux for such purpose really matters, regardless whether it is inherent in it or not.
Besides (developers) community does not have the power to make something popular and recognizable outside of itself. Big corps may have a chance (but may fail too).
With both community and corps turned back to something (theoretically) superior it has 1% chance of gaining recognition by itself.
The important thing to note is that if the BSDs *do* start gaining more traction, and more name brand recognition, and become a choice that people have to consider right along side Linux… you have just substantially increased the “problem” of fragmentation in OSS.
I think that their value to fragmentation (one platform per one kernel) ratio justifies their existence. Beside they are just a demonstration that inbred fragmentation is not inherent in OSS system development, not necessarily examples of overall superiority.
The road to fragmentation is paved with the projects of people who have wanted to end fragmentation once and for all, and have started their own be all and end all project to do it. Each thinking that his idea was so good, his project so compelling, that everyone would agree and start doing things his way.
It takes maturity to understand that. I believe such maturity will emerge in OSS world some time.
First of all, appologies to all for posting my last message twice. I’m not sure how I did that.
“””You assume that software wins solely on the base of its technical merits.”””
Absolutely not. In fact, I thought that what I said implied that there were many factors. Much of the OSS community dismisses the value of basic marketing smarts… like not naming their project “Podfuk” or something. (Yes, that is a real project name.)
Linux is awful with marketing in a lot of ways. But I suspect that it has been less bad at marketing than the BSDs, judging by relative popularity. (OK, this is a terrible metric, but I will point out that Googling for BSD gets about 35 million hits and Googling for Linux gets about 281 million hits.)
There’s a lot of low hanging fruit available to projects that understand at least the basics of marketing and assign value to it.
And I’m not implying that those projects which take advantage of marketing have an unfair advantage. The same lines of thinking that encourage good marketing practices also encourage placing value upon solid usability testing, etc. which are very valid strengths for a project. (I am speaking generally here, of course.)
“””
It takes maturity to understand that. I believe such maturity will emerge in OSS world some time.
“””
Ah, but the average age of project leaders in the community is low, and there are more contributors being added all the time. The community is a different thing than an individual. I have a different outlook at age 44 than I did at 18. But I’m not sure it works that way for the community.
Besides… do note that I tried always to put words like “problem” and “fragmentation” in quotes. I’m not at all sure that OSS’s current dynamics are, on the balance, a negative. (Obviously, its a mix.) I’m not sure that us old folks can keep up with all these youngsters with the new ideas and confidence that only comes with youth. 😉
But back to my original point… OSS is what it is for now, and that is what we have to work with. Even if you are correct, and maturity will come to it in time, there is no magic incantation or plan which can imbue it with maturity any sooner than when it is ready (much to the dismay of many parents, I’m sure). 😉
“””
BSDs are oss too and do *much* better job dwarfing fragmentation.
…
Really, if linux cannot live up to the task of decently supporting non-it-expert user with easy to use and maintain,efficient, dependaple and open system it’s time for all people who believe it’s their mission to move on to something else. Maybe there are other systems that could use community engagement better for achieving such goal.
“””
Well, I specifically used the term “Linux” and not “OSS” for exactly that reason. However, now that you have brought it up, I will address it.
A certain degree of diversity, which some would call fragmentation, *is*, in fact, inherent in the OSS model.
*BSD (and note the wildcard at the beginning) is not all of OSS. The BSD’s are a few players among quite a number of other players that make up the OSS community. Even if you consider them as a whole to be one player, they are still *adding* adding to the “fragmentation”, not reducing it.
Furthermore, from the standpoint of popularity and getting themselves noticed (and I know I’m likely treading upon thin ice, here) they have not been nearly as successful as Linux in this ecosystem. Limiting their diversity internally has not been an adaptive strategy for the BSDs *if* the goal was to compete for popularity.
Do you see that your post illustrates, very nicely, why OSS is as it is?
You are saying that you know a better way than what is being done by Linux today; That the *BSDs would do it right if only they rose in popularity and name brand recognition.
Now, you may be right, and you may be wrong. I don’t care. And it doesn’t matter, because the world is as it is, not as we think it should be.
The important thing to note is that if the BSDs *do* start gaining more traction, and more name brand recognition, and become a choice that people have to consider right along side Linux… you have just substantially increased the “problem” of fragmentation in OSS.
The road to fragmentation is paved with the projects of people who have wanted to end fragmentation once and for all, and have started their own be all and end all project to do it. Each thinking that his idea was so good, his project so compelling, that everyone would agree and start doing things his way, once they saw how great it was.
*That* is why OSS is as it is and will continue be so.
And arguing against this truth, in my opinion, is about as effective as arguing against the second law of thermodynamics. The two have a lot in common.
Which is why I say we may as well leave square one behind and move on to square two… and figure out how to operate most effectively in this existing ecosystem.
Edited 2007-03-05 22:07
with distros like ubuntu that take a good stable base and wreck havoc on it then push it out the door, market it and call it the ‘best thing out there’ it no wonder that fragmentation is rampent. they took the debian base that had several distros that were all compatible at the core, and made a distro that is winning over newbies left and right yet i think TRIES to break compatibility. ubuntu is a virus at best.
Being a Linux advocate myself I can say the following over the years has transpired.
Startup distro’s promise the world with a nice retail box and screen shots only to be run afoul by bugs and non working software with no support.
I am not trying to be negative just the way end users response was I myself never gave up and I kept trying. Then at work it started to take over in around 2001 or so and from that point on I was more interested than ever.
In my opinion with SLED/Red Hat and of course Ubuntu the options for an AWESOME desktop distro are more now than ever. But marketing is the key to conquer the desktop and pre-loading it with working hardware and software.
Plus the fact on laptops it does not use as much battery or make the fan run all the time…
How many number of fanboy posts we see saying Dell or anyone else feel confusion on Linux market is wrong.
Yes ignore everyone’s complaints. Act like Linux is superior. Good for Microsoft at least they listen.
I many times think i will write software for Linux but given the vast amount of incompatibilities in Linux, unless you want to write open source it is impossible for single developers to test their stuff on various chaotic mess of distribution.
To me FreeBSD is better because at least the branding is unique and there are not 100s of FreeBSDs.
The sad fact is SuSE is sitting on a desktop/laptop GOLD mine waiting to be discovered.
SLED could be the world class desktop distro leading MS by the hand but it will never happen because it is hidden by lack of marketing!
“If you look through the comments on the IdeaStorm website you’ll see a number of people seriously suggesting that Dell should just create their own Linux distribution that they can easily support.”
They could also create their own FreeBSD distro, or have PC-BSD as a pre-installed option. Then the Nvidia cards they ship would have drivers that work.
Excerpts from the article bolded:
As a principal, this freedom is extremely valuable. However, a couple of announcements this week seem to indicate that market value of freedom of choice has dipped considerably.
This is the wrong attitude. While freedom means ten million different distros, many of which don’t have a sizable community behind them and will probably fizzle out, freedom also means a few dozen prominent distros that have become hugely important.
Communities will naturally rally around the distros that they perceive to meet their computing and support needs. Having lots of other options doesn’t have to take anything away from the consumer.
The other announcement this week was IBM’s refusal to certify Oracle’s Unbreakable Linux. Again, it’s an issue of too much choice.
I think there’s probably a lot more to IBM’s decision than “an issue of too much choice”. This is more like posturing among white shirts, and forcing Oracle to prove how serious it is (or is not) about its new venture.
In other words, certification is handy but it’s certainly not pre-installation.
Certification from major sellers is a good start, and it’s a hell of a lot better than where things stood just three years ago.
The fragmentation argument is a shield. I suspect there are a number of people within the linux community who are less interested in linux as a platform, and more interested in linux as a solution to an anything-but-Microsoft agenda.
These people embrace the “openness” of linux compared to Windows, but then turn around a propose everyone settle on a single standard for DE, package management, distribution, etc. in order to be more competitive. There’s a bit of a dichotomy there, you can’t extol the virtues of an open platform while insisting that openness should be diminished by arbitrarily setting platform standards everyone should comply to in order to gain acceptance.
Linux is not Windows. It’s not a replacement for Windows. It will never be either. Insisting that Linux can succeed by emulating Windows simply won’t work, because we already have Windows. It simply reduces linux to being nothing more than a freeware version of Windows, and assuming that people are using linux mostly because it’s free as in beer doesn’t send a positive message to commercial vendors about the willingness of users to invest in commercial software.
The real issue is the viability of desktop linux. As long as there are no concrete metrics to measure both a realistic number of desktop linux users and their willingness to purchase commercial software, it will remain difficult for mainstream companies to justify investing in development for linux as a platform. A community of users saying Ubutnu rocks! doesn’t equate into any sort of market study or potential market assessment.
The bigger issue is market lethargy. Even within the linux community you find many people dual-booting or switching to Windows because there are applications they require. That being the case, since even the linux community is showing that they will still use Windows when the situation dictates, why invest in linux? The only vote users have is with their wallets. If you’re committed to desktop linux and want to see it become more of a commercially viable option, then sacrifices need to be made and you need to search high and low to seek out those vendors currently supporting linux and support them with your hard earned dollars. Whether talking about hardware or software.
The issues of disparate distros and standards will easily be dealt with by the ISV’s *if* they are convinced the market exists to justify the investment. Until that time, they’ll continue using fragmentation and competing standards as an excuse, and the anti-Microsoft camp will rally around that and insist we sacrifice flexibility for arbitrary choices.
Linux must succeed on it’s own merits if it is going to truly succeed. It won’t succeed by trying to turn it into something it’s not.
>These people embrace the “openness” of linux compared to >Windows, but then turn around a propose everyone settle >on a single standard for DE, package management, >distribution, etc.
DE, package management, distro file layout, lib versions these things are fascinating for developers but most other people couldn’t care less. They want them to be transparent. The variances here are only unnecessary distractors.
The best way to move such details out of the picture is to standarize it. Combination of those components (becouse their developement is happent outside of distros) It’s really not the the place where innovation beneficial for the user happens.
>Linux is not Windows. It’s not a replacement for >Windows. It will never be either. Insisting that Linux >can succeed by emulating Windows simply won’t work, >because we already have Windows.
Too bad because world badly needs replacement of MS’s licensing strings, overprices forceware and “screw ya” attitude. I for one once thought that linux could have made the difference here. Seemingly I was wrong.
> That being the case, since even the linux community is showing that they will still use Windows when the situation dictates, why invest in linux? If you’re committed to desktop linux and want to see it become more of a commercially viable option, then sacrifices need to be made and you need to search high and low to seek out those vendors currently supporting linux and support them with your hard earned dollars. Whether talking about hardware or software.
I could turn the question around: Why the casual user should start investing his hard earned dollars when the community isn’t trully commited to sacryfying some of it’s freedom to deliver someting that works for him.
>The issues of disparate distros and standards will easily be dealt with by the ISV’s *if* they are convinced the market exists to justify the investment. Until that time, they’ll continue using fragmentation and competing standards as an excuse, and the anti-Microsoft camp will rally around that and insist we sacrifice flexibility for arbitrary choices.
These issues are far from easy. Read autopackage docs and mailing list for more insight. The is ballance. The cost of retesting and certification makes the barrier for entry unacceptably high for a commercial vendor compared to meager returns from fragmented market. The first step belongs to community.
>Linux must succeed on it’s own merits if it is going to truly succeed. It won’t succeed by trying to turn it into something it’s not.
It have had more that 10 year to succeed and it did in some niches. Desktop just happents not to be among of them, which makes the quest for liberating it still open. That’s really the core conclusion of the summed up experience.
Edited 2007-03-05 21:53
“we just do things differently around here”.
(oops… mistyped ‘fragmentation’. Silly me)
Edited 2007-03-05 18:49
The only way that any Linux distro will ever become mainstream or succeed commercially is if 3rd party companies can make money off of the end users. In a world where money is the be-all end-all, it is as simple as that.
Unless there is a substantial market for games in the Linux arena, big name game developers will not make them.
Unless there is some way to make money from music enthusiast, mp3 player companies will not support it.
Until protected content can be played by any Tom, Dick, or Harry, Linux will not become mainstream.
Until the corporations are #1 in the distro’s eyes and the users merely a source of revenue, Linux cannot replace Windows in the majority of homes.
It makes me angry, but it is true.
Unless there is a substantial market for games in the Linux arena, big name game developers will not make them.
I wouldn’t hold my breath for this…the PC game market keeps losing ground to the console game market, year after year. Apart from a few big companies (Microsoft, iD, Valve, Blizzard, Bioware…) there is increasingly little “major” game development for the Windows platform. Most major games are now developed for the Xbox360, the Wii and the PS3, PC versions being an afterthought.
This is both good and bad…for most gamers, that means they can use Linux or OS X on their work PCs, and get their gaming fix from the consoles. The bad part is that games for either Macs or Linux computers are even less attractive than they used to be.
Unless there is some way to make money from music enthusiast, mp3 player companies will not support it.
Doesn’t really matter, pretty much every mp3 out there works under Linux (iPod support under Amarok, for example, is actually very good). The only thing that’s really missing is iTunes access, and I don’t see Apple supporting Linux any time soon (if you’ve noticed, they act as if Linux didn’t exist, because they like to position themselves as *the* alternative to Windows…)
Until protected content can be played by any Tom, Dick, or Harry, Linux will not become mainstream.
This one’s a bit more complicated…reading DVDs in Linux is trivial, but technically against the law in the US…that said, it *is* possible to buy commercial DVD software for Linux (I believe it will be available through CNR). As far as HD-DVD and Blu-Ray are concerned, it’s still too early to tell…we’ll see.
Until the corporations are #1 in the distro’s eyes and the users merely a source of revenue, Linux cannot replace Windows in the majority of homes.
That’s one way to look at it, although one could argue that more Linux PCs in the workplace will lead to more Linux desktops at home.
You make excellent points, and I agree with you for the most part. However, about the games… I personally own a few games systems, but I like playing RTS mostly. Not just because of the games, but because of the mod’s that can be used. That is the thing that will keep PC gaming unique.
The mp3 players may work, but they do not say that on the retail box. And let’s be honest, my if mother were to buy my kids or niece one, she would only be looking at the box. We who are more inclined towards geekiness would know to look for open source or community driven projects to support our chosen player. Speaking of which, I need to find out if Amarok can support my Zen Vision W.
More Linux desktops in the workplace would definitely help out the home usage. Not only for the familiarity of it, but because of the access to friends or IT nerds at work to ask for help. (I say nerds because I am one.)
Thanks for the reply, you make some good points.
I personally own a few games systems, but I like playing RTS mostly. Not just because of the games, but because of the mod’s that can be used. That is the thing that will keep PC gaming unique.
Oh, the PC game market won’t die. There are few genres (like Simulations, RTS and “casual” games) where it still reigns supreme, and others (FPS) where it is a strong contender. However, it will continue to be marginalized with regards to console games.
I agree with you as far as MP3 players go, labeling them as being compatible with Linux would help. We’re still at the point with Linux where you need to check if equipment is compatible before buying it (just like you would for a Mac). Most peripherals are, but some aren’t, or require some amount of tweaking to make work. This is changing, albeit slowly, and there’s no real way to speed things up.
Linux can be like electricity it is going to take the path of least resistance…
There is no place you can run and hide from your problems not up on the highest mountain or the deepest cave they will still be with you.
On those two notes Linux has come a LONG way since for instance Red Hat 6.0 when getting it installed and working was a party and something to brag about. Now it functions on most hardware and it is VERY stable with no crashing from bugs or lockups.
Like I said earlier Novell is sitting on a gold mine of proportions they cannot comprehend. If SLED was marketed as a dual boot option at least the minimum people would be exposed to what a slick operating system it really is and the ‘wow’ factor would knock Vista back to the unused portion on the hard-drive.
sp
>On those two notes Linux has come a LONG way since for instance Red Hat 6.0 when getting it installed and working was a party and something to brag about. Now it functions on most hardware and it is VERY stable with no crashing from bugs or lockups.
The one thing that haven’t really improved though is fragementation and unstable platform interfaces. And that what the article is all about.
I think i can speak for
many people out there,
At this point, linux is not
yet ready to replace xp
or 98se on my computers. Check out
fry’s electronics, linspire
is preinstalled on some
of their computers.
when people i know buy
them, linux comes off
windows xp goes on.
Browser: Mozilla/4.0 (compatible; MSIE 4.01; Windows CE; PPC; 240×320)
Why are windows apologists/users so defensive towards Microsoft and so ready to put down Linux progress. For all its shortcomings Linux has made amazing progress all without marketing or multi million dollar budgets.
People who say that they “use whats best for the job” or deride linux for being inconvenient are the same ones who are ready to go through inconveniences for something they feel ethically motivated towards (like fairtrade coffee or being against polluting vehicles). Then why the defense of Microsoft products? Why deride something that is made by the community for YOU and your benefit by ppl not asking anything in return. People who say that Linux does not support hardware.. dont they realize that Linux supports most hardware WITHOUT any support from the hardware vendors.
Sorry if I am ranting but why would anyone defend a large organization over something made by the community even if its not so perfect is something I dont understand.
They defend Microsoft and Windows because they believe the hype.
They believe Linux is communist, anti-american (anti-us really as there is more countries in the American continent than just the USA), hippy, anti-business etc etc etc etc
I live in the UK, and one of our major media outlets is the BBC. Over the years the BBC has been a staunch supporter or Microsoft.
However, over the past year almost all the BBC reports about Microsoft, Windows or Vista have been extremely negative.
Why is this ?
The BBC decided to look behind the hype and examine the products themselves. Microsoft told the world that Vista would be the most secure Windows, BBC decided to wait and see on that one. today it posted this;
http://news.bbc.co.uk/1/hi/technology/6418965.stm
Previous bad articles are here….
http://news.bbc.co.uk/1/hi/technology/4748257.stm
http://news.bbc.co.uk/1/hi/technology/6407925.stm
http://news.bbc.co.uk/1/hi/business/6407419.stm
and so on and so forth.
Oh I understand that people believe the hype but in todays internet culture that really does not make sense. People can find out information at the click of a mouse. Also I can understand inertia etc among individuals but why among businesses? As I understand it businesses are there to make money right? Any money going to another company is definitely a no-no. Then why are businesses so defensive?
Now people will say 2 things to the above:
1. Businesses have software that requires windows
2. Training on a so-called “unfamiliar” system
3. Support is non-existent for Linux
And I say fine if that is really the case, but do you mean to tell me that in a company where many of the users are not power users or use specialized software that businesses are even considering any shift to Linux? or even OSS. If not why not? Surely the cost in licenses for Windows+MSO is going to be less ?
Training is a strange thing.. I read articles where just because the start button is not the same as in Windows the case is made that users will require extensive training. Now in case of Vista + Office 2007 the interface is completely changed, so shouldnt businesses at least evaluate some software which has a familiar interface and costs less to boot?
Support we all know is available.
Addressing the article, its all bull****. Dell just needs to support ONE distribution nothing else. And include it in the terms that installing any other distribution voids the support contract. Thats all that is needed. Surely dell will not support you installing Win98 on an XP laptop? This is a non-existent problem.
No, my main gripe is this: Most businesses will go out of the way to evaluate “n” number of vendors for ANY of their products including paper-clips before deciding on one. Its only in the case of Microsoft that they have a blind spot. It is a bit strange that in these competitive times when every penny saved is an advantage there are so few businesses not trying to do what seems to be common business sense and try to keep control of their own inventory.
Obviously there are people who want to buy a box labeled “Linux Professional Edition” from the Linux Corporation. Ignore them.
The Linux product is made by and for the FOSS culture. That is the nature of the beast. Don’t expect it to transform itself for recognition or recompense that will never arrive. “Market share” gives nothing to the developer. The joy of creating something -new- does.
Its late and I’m irritable.
Linux is a kernel. This is important, as it *can* be *one* small part of the ecosystem of a distribution.
A Distribution is simply a set of packages put together that the Distributer thinks should go together. Without going into advantages of a specialist distribsion. i.e.
1) One that works on old hardware.
2) Geared towards Video editing
3) One that can be used to Partition a Hard Drive
4) One that shows of Open Source Gaming
etc etc.
A distribution is normally…
A kernel+Shell+Package Manager+Windows Manager+Desktop Environment+Web Browser+Office+Other Applications.
And are often distinguished by which packages they choose although there is only really a couple of main choices in the *main* categories.
The difference between distributions is often how cutting edge vs stable these main packages are.
Sometimes there is a focus on licenses, about what is more “free” vs larger support etc.
*BSD distributions are exactly like Linux distributions apart from the fact that Linux *the kernel* has been replaced. Thats all!? If I run *BSD I would still run BASH+X+XFCE+OpenOffice+Firefox…Hell if I run Vista I would still run the last two.
Anything to do with BSD is off-topic. *BSD is just another distribution.
*BSD distributions are exactly like Linux distributions apart from the fact that Linux *the kernel* has been replaced. Thats all!? If I run *BSD I would still run BASH+X+XFCE+OpenOffice+Firefox…Hell if I run Vista I would still run the last two.
Anything to do with BSD is off-topic. *BSD is just another distribution
I’m not going to get into a discussion about Linux fragmentation. Quite frankly, I don’t give a damn. But with this comment, you sir, have derailed. I don’t believe you’ve ever used any BSD, and if you did you sure as hell weren’t paying attention. There is far more difference between a BSD fork and a Linux distribution than you can obviously comprehend.
http://en.wikipedia.org/wiki/Image:OpenBSD39snapshotxfce.png
is that a screenshot of xfce on openbsd.
http://www.freebsd.org/ports/shells.html
is that a list of shells I notice bash is on there.
http://lists.freebsd.org/pipermail/freebsd-current/2004-July/032267…
ooooh is that when freebsd moved to X.org
ftp://ooopackages.good-day.net/pub/OpenOffice.org/FreeBSD
is that…is that Openoffice…no!?
I fully accept there may be differences but they are sooo well minor. There are differences, but 99% of users *shouldn’t* care. I don’t care what *nix I use.
I’m more that welcome for you to point out the differences, although I suspect we both know the truth.
Solaris; *BSD; Linux don’t care.
Now if your running a headless server with a dedicated job the answer *may* be different. For an alternative to the Microsoft Platform on cheap hardware by dell. I’m more concerned whether they use Gnome or KDE.
Not this one again – same hatchet is being pushed out again and again and again.
Look, there are three main distributions; OpenSuSE, Fedora and Debian – everything else is based off those three core distributions.
Whether its Ubuntu based off Debian, Novell SuSE Enterprise Linux based off OpenSuSE or Red Hat Linux Enterprise based off Fedora, the simple fact is, the red hearing of ‘fragmentation’ is just that, a red herring.
I mean, if you want fragmentation and and incompatibility between versions, one doesn’t need to look any further than Windows! I mean, not to beat Microsoft because some of the changes they’ve made, in regards to fixing win32 problems, needed to be done – but to say that some how the so-called problems of win32 are dwarfed by those of Linux fragmentation is ignorance at best.
Another thing these doomsayers forget is this; opensource means that that the software companies who write software for Linux can actually get *INVOLVED* with the development of the operating system; fixing up deficiencies that they might seen when developing their software, fixing up performance issues with libraries they might rely on – it gives them a flexibility than they did with Windows, where they’re at the mercy of Microsofts willingness as to whether or not to address their concerns in a timely manner.
Another thing these doomsayers forget is this; opensource means that that the software companies who write software for Linux can actually get *INVOLVED* with the development of the operating system
Very good point!
Unfortunately they seem to prefer getting things shoved down their throats and finding workarounds later on.
Maybe it is a bit like what Agent Smith said in Matrix: they don’t want it perfect, they need the suffering
Whether its Ubuntu based off Debian, Novell SuSE Enterprise Linux based off OpenSuSE or Red Hat Linux Enterprise based off Fedora, the simple fact is, the red hearing of ‘fragmentation’ is just that, a red herring.
Without knowing what “based” means in any case this really doesn’t mean anything. Besides those three root distros are the types of “allways in flux” software that is an oxymoron of platform. The fact that they breed lots of spinoffs in based of various snapshots doesn’t help a bit. Actually by lowering barrier for proliferation it actually worsens the fragmentation.
I mean, if you want fragmentation and and incompatibility between versions, one doesn’t need to look any further than Windows! I mean, not to beat Microsoft because some of the changes they’ve made, in regards to fixing win32 problems, needed to be done – but to say that some how the so-called problems of win32 are dwarfed by those of Linux fragmentation is ignorance at best.
Still market for easily installed binary software exists on overpriced windows and not on free linux.
(by market share figures if linux was equivalent technically its market would exist in apriorately lower scale, but it doesn’t).
Another thing these doomsayers forget is this; opensource means that that the software companies who write software for Linux can actually get *INVOLVED* with the development of the operating system; fixing up deficiencies that they might seen when developing their software, fixing up performance issues with libraries they might rely on – it gives them a flexibility than they did with Windows, where they’re at the mercy of Microsofts willingness as to whether or not to address their concerns in a timely manner.
They have flexibility but not dependablity. They can contribute whatever they want but they can’t influence in what shape their contribution find their way to final products (distros). And they can’t fight fragmentation. MS at least gives them roadmaps, announces its (arbitrary, true) decisions in advance and sells tools that produce code that works on it’s systems.
Without knowing what “based” means in any case this really doesn’t mean anything. Besides those three root distros are the types of “allways in flux” software that is an oxymoron of platform. The fact that they breed lots of spinoffs in based of various snapshots doesn’t help a bit. Actually by lowering barrier for proliferation it actually worsens the fragmentation.
Again, who is forcing them to support those distributions? pick two/three large distributions and only support those – problem solved. As for the spin offs; Enterprise Linux is directly derived from Fedora, and there are enterprise distributions that are direct pull downs of sources from Red Hat enterprise minus the branding.
Still market for easily installed binary software exists on overpriced windows and not on free linux.
(by market share figures if linux was equivalent technically its market would exist in apriorately lower scale, but it doesn’t).
The reason why they don’t port to Linux is simply this; if they ported their applications to Linux, they would be forced to compete head to head with opensource applications – people would be asking Adobe to justify $1399 on Creative Suite when a opensource tool chain of Gimp/Scribus/Quanta and the likes could achieve the same sorts of results.
That is what commercial application vendors are scared of – a group of users finally demanding that they justify their steep pricing for very marginal improvement over their opensource counter part – meaning, they would either have to dramatically improve their product OR lower their price.
Why don’t they have a problem on Windows? Because its easy, there is someone/something they can bash on a regular basis, Microsoft, who seems to be universally hatred amongst end users and allows the likes of Adobe and Symantec get away with murder – when all the shit hits the fan, who do the bash? easily, they bash Microsoft; when Microsoft proposes a superior replacement to PDF with more liberal licencing, who plays the victim game? Adobe. When Microsoft decides to lock vendors out of the kernel and provide interfaces for kernel monitoring functionality via a user space API (Defender API) rather than requiring a kernel driver to be loaded, who starts to spread lies and half truths about it? Symantec.
These software vendors *NEED* Microsoft because it provides them with a whipping boy when all things go down the gurgler – perish the thought of them actually having to stand behind their product and the decisions they made when writing it!
They have flexibility but not dependablity. They can contribute whatever they want but they can’t influence in what shape their contribution find their way to final products (distros). And they can’t fight fragmentation. MS at least gives them roadmaps, announces its (arbitrary, true) decisions in advance and sells tools that produce code that works on it’s systems.
Oh pulease; Oracle have made major contributions to the Linux kernel to benefit their own applications; there is nothing stopping Adobe from helping out with GTK/Cairo in respects to PDF implementation and scalable interface, and fixing any possible problems which they might find on their journeys.
I am sure that the Fedora community wouldn’t mind Adobe or Corel contributing to Fedora’s development by providing new libraries that are licenced under acceptable terms.
Don’t blame the distro’s if these software companies *CHOOSE* to be anti-social – all they have to do is sign up to the mailing list and air their dirty laundry if they want, they fact they choose to hide themselves in an ivory tower and bash organisations behind PR spokes people speaks volumes for the lack of back bone there of, that many of these software seem to lack.
Again, who is forcing them to support those distributions? pick two/three large distributions and only support those – problem solved.
For your engineers yes, for your sales, not really, definitely not for the users.
As for the spin offs; Enterprise Linux is directly derived from Fedora, and there are enterprise distributions that are direct pull downs of sources from Red Hat enterprise minus the branding.
Redhat actually managed build sort of an enterprise platform with RHEL. That’s why it captures most of the commercial support support linux can get nowadays. Those clones are not really quite the same as derived distros.
The problem is also incompatibility beetween versions of single distro.
The reason why they don’t port to Linux is simply this; if they ported their applications to Linux, they would be forced to compete head to head with opensource applications – people would be asking Adobe to justify $1399 on Creative Suite when a opensource tool chain of Gimp/Scribus/Quanta and the likes could achieve the same sorts of results.
That is what commercial application vendors are scared of – a group of users finally demanding that they justify their steep pricing for very marginal improvement over their opensource counter part – meaning, they would either have to dramatically improve their product OR lower their price.
Gimp is available for windows for some time. I don’t know about Quanta and Scribus but they should arrive to Win32 world as soon as they are ported do KDE4. Generally most OSS worth a dime get ported sooner or later.
Oh pulease; Oracle have made major contributions to the Linux kernel to benefit their own applications;
Hey, Oracle is 800lb gorilla. A platform in itself. You can’t compare that to an arbitrary gui app that is installed by a family man, not integrated by an army of dedicated engineers. Besides, Oracle also supports only a handful of enterprise distros.
there is nothing stopping Adobe from helping out with GTK/Cairo in respects to PDF implementation and scalable interface, and fixing any possible problems which they might find on their journeys.
Only that they have to do it >2 years in advance (compared to product shipping) and pray the changes will consistently make into majority of used distros (including LTS variants) by that time.
Don’t blame the distro’s if these software companies *CHOOSE* to be anti-social – all they have to do is sign up to the mailing list and air their dirty laundry if they want, they fact they choose to hide themselves in an ivory tower and bash organizations behind PR spokes people speaks volumes for the lack of back bone there of, that many of these software seem to lack.
Halleluyah, bro! Don’t buy software from companies you consider evil.
I for one, would love not to have to use Windows. Unfortunately, linux community by not agreeing to provide a platform is giving in the field to MS through walkover.
Dell is just like most companies today. They want to survive using the same techniques and methods that they have always used. Its only when their backs are against the wall that they try to think of the best way to serve the customer. Does anyone recall Dell ever asking for customer feedback before ? Its only now after their poor financial situation that they have suddenly woken up.
And then of course there are the corporate apologists like the author who can never find any wrong with what the corporations do. Its always the community that has to bend backwards, change the way they think, behave or simply are to accomodate the companies. Keeping in mind that it will be DELL that is going to benefit from this by not being beholden to MS and gaining new users who are not being served by any other vendor.
I really find the article strange. Why should the community change? Dell became what it is because it was willing to change to meet the customers requirement. Then of course the customers became less and less important. However they asked for feedback and their customers gave them feedback. Its up to Dell to satisfy them. Its NOT the communities job or nature to satisfy Dell. Its Dells to satisfy their customers. Why? Because the community has never received anything from Dell. Its owes Dell nothing. But Dell has risen to where it is because of its customers. Those customers have changed and are demanding Dell should change. Its up to DELL to meet the challenge. The community is fine as it is.
This is what Ive been saying all along…universal package manager that makes installing software across distros as easy as it is in windows would bring new linux users flocking.
Fragmentation is normal. There are a couple (not many) major distributions of Linux. There are a couple versions of Windows (home, professional, data center, small business, etc.). There are a couple flavors of BSD.
Linux is a set of tools available to build different products, and operating system as an example. Linux distributions are products built on those tools.
What good is a modularity if one can’t combine components into various products ?
I think that unification people are about some kind of world domination by one OS. If Bill Gates was Linux user, he would be a supporter of unification.
DG
The key is to standardize (and move out of picture) what’s boring and innovate above it.
Finding the right balance is hard, but pretending that the problem does not exist is not the right way to do.
I think of distributions as separate products. Novell, Ubuntu and Red Hat are managed by single entity. They have some kind of company standards and guidelines, there is no doubt about it.
I believe that it is enough. Majority of Linux users are covered by those distributions, so they are some kind od de facto standards.
DG
Well redhat (RHEL) is nonexistent on the desktop and Novel is directed towards paying enterprise customers. So those two aren’t really positioned to cover the majority of (potential) desktop users.
Ubuntu has lax abi policy (between releases) and openly declares being more an appliance OS than a platform. Third party of the shelf apps aren’t much welcome there.
When talking about these 3 distros as roots of derivatives nobody defined which part of them is a “platform part”, what can and what cannot be customized to preserve “compatible” label. Besides, no derivative clearly documents what changes it introduced from the “platform” standpoint. When developing an app, or certifying your hardware for it you cannot assume to what extent derivatives will be compatible with the root distro and know whether your testing will have to be repeated for each deriviative.
Fragmentation is rather normal, but a fork of BSD is something different, than 1000+x distros in Linux because there isn’t something to fork of.
And if you do have a base, you can do a quality fork, but you’re doing just “something” without a base. There is no need for unifications, there is just a need for a stable and reliable base system.
>If Bill Gates was Linux user, he would be a supporter of unification.
And if someone is really mature, he will be using *BSD
A agree that BSD situation is not exactly like Linux situation. I just want to say that it is hard to create one OS for everybody. Good example is PC-BSD and FreeBSD. PC-BSD offers minimal system administration, while FreeBSD offers flexibility.
BTW, I am using BSD (FreeBSD).
DG