Slashdot Mirror
OSDL's Review of Desktop Linux In 2006
derrida writes "The OSDL's Desktop Linux Working Group has published its first year-end report on the state of the overall desktop Linux ecosystem. The report provides insight into the year's key accomplishments in functionality, standards, applications, distributions, market penetration, and more. Of great interest is the Market Growth part. Quoting from there: 'Most observers believe that much of the growth will take place outside of the United States. "It will be in the BRIC (Brazil, Russia, India, and China) countries," said Gerry Riveros, Red Hat, "because of the price and because they aren't locked in yet."'"
You don't get a free pass with a comment like that. It's easy to look at what Congress has done here, and say, "boy, is the U.S. fucked." Unfortunately, we aren't alone. Europe is taking our shiny new copyright and patent crap and running with it (and making it even worse in some respects, if that's possible.) Furthermore, there's a lot of pressure being applied to bring other countries in line, pardon me, "harmonized", with certain unpleasant aspects U.S. IP law. We're all going down the tubes together: we're perhaps a couple of elbow joints ahead of everyone else, but not that's all. Too many powerful people around the world want control of their respective economies, and one way you do that is by manipulating and suppressing technological advancement.
The higher the technology, the sharper that two-edged sword.
So they're going to say that Linux will really grow in countries like China and India, where street vendors hawk a variety of Microsoft bootlegs for less than $0.50?
I'm not seeing the appeal.
I think the interesting thing about this is the projection of the greatest growth in the "BRIC" countries. I don't think it's so much that they aren't locked into Windows, as much as Microsoft has (inadvertently on their part) pushed it along. When MS started its big "anti-piracy" crackdown, it mostly hit in these parts of the world. Add in the high cost of Windows, and the ever-increasing hardware requirements for it, and a free OS that can run on existing hardware looks pretty darn good.
The problem desktop Linux is still facing is getting more penetration in the biggest market - the United States. There are still areas where improvements need to made, and in some areas, applications to be developed. One thing that we have to recognize is that MS is not going to give up its stranglehold on the OEM installed market. The only way Linux going to be able to make any strides is to recognize that the user is going to have to do the install, and to make it easy for them. There's a project going on for Ubuntu which shows some promise, called Winbuntu - it's a Windows installer for Linux. I don't know how it'll work out, but it shows the concept.
it's already been identified that the majority of OSS gets developed outside of the USA, i think you will find america's court system and patent laws are going to result in doing software business inside the USA to become very unpopular through the next decade. you'll end up with only massive corperate entities like MS able to cope with these entry barriers, and if you think you can rely on companies like MS for innovation......
If you mod me down, I will become more powerful than you can imagine....
In years past I have always noticed that FreeBSD always makes it easy to install. Makes it easy meaning it recognizes hard drives, network cards, even 56K modems, without a problem. I installed FreeBSD with two 3.5" standard FreeBSD install disks a few years ago over a 56K modem with no problem. Like the Apple commercials say - "it just works".
I prefer Debian and Linux to FreeBSD, but Linux distros have a lot to learn from FreeBSD in terms of ease of installation. FreeBSD makes it really easy to install itself on a PC without barfing on network cards, hard drives and so forth. It was the same situation ten years ago when I was installing Slackware on multiple floppies versus my FreeBSD network installs. And from my experience last week, I see it still holds true.
The press on Vista is so bad right now, it should really be used to push Linux adoption on the desktop in 07.
I myself switched to Ubuntu on my home desktop because of Vista-fear.
Every year, I see these "Linux is ready for the desktop" articles. But it never happens. Back in 2004, WalMart offered a $499 Linux laptop. They don't do that any more. Lenovo, HP, and Dell have fooled around with Linux laptops, but try to order one on line. Search for "linux laptop" on Dell, and you get back "Dell recommends Windows Vista(TM) Business." There are some off-brand Linux laptops available, but they're overpriced.
Linux on the desktop looked closer three years ago than it does now.
KDE in its current form is quite usable for most common purposes, and those abilities it doesn't have can probably be added as widespread adoption takes place. OpenOffice has its faults but it usually does the job. I would say at this point, it's not Linux as Linux that's the holdup. It's:
1. Legacy systems, documents, and most importantly user training in said systems and documents. "If it ain't broke, don't fix it" rules when computers are the tool rather than the end goal in and of themselves, and it's hard to fault that logic. If you change your systems you're effectively "breaking" your employees in terms of their productivity, and fixing them is quite a job. It's only justified when the end benefits are worth the pain, and to be fair in most cases they probably aren't, at least in the short term. And we all know how good capitalism is at thinking long term.
2. Compatibility with the largest possible market segment. If your customers/suppliers insist on dealing in old formats (see #1) then it's rather hard to force them to change. And every minute spent dealing with such issues is one less spent on work related to producing something.
3. Costs of retraining your IT department and switching your software/machines. Yes it will take time - hardware support, IT helpdesk training, identifying and testing replacements for currently used apps, etc. Not painless at all.
I would say Linux was "ready for the desktop" several years ago, or at least as ready as Windows. KDE and Gnome are excellent systems for most users, once installed and configured properly. (That's what admins are for - work PCs are not normally maintained directly by users, regardless of OS.) Now the problem is revealed as being rather deeper than originally anticipated - it's not JUST Linux that's the problem, it's change period.
For home use, people want to play media and install thousands of commercial specialty packages, which are all written for Windows. More legacy software issues, with no budget or interest on the part of the people writing them (why target an uncertain platform populated by geeks who give stuff away?)
The problems aren't technological now - I would say they can be more accurately characterized as inertia. It's hard to give people reasons to switch from something that works, even when the new thing is BETTER than the current one. Linux, due to legal constraints as well as not quite 100% compatibility with things like Word formats, is not and probably CANNOT become (legally) a drop-in which is better in all cases.
Personally, I think the only hope for a massive switch to an open source OS is one where the software is written in such a fashion that it can be PROVEN (mathematically) to be secure/crash proof/what have you. Such a verifiable guarantee might gain enough interest/momentum to be worth the massive shifts that still have not taken place, but I am aware of no other lack in the marketplace severe enough to warrant it.
"I object to doing things that computers can do." -- Olin Shivers, lispers.org
As a relatively new linux user, I can say that I've seen significant progress in the ease of use and functionality in just one year.
I started with Fedora Core 5, got jealous of some of the functionality of my girlfriend's Ubuntu, and I'm now extremely satisfied of Fedora Core 6 which brought all the functionality it lacked and even great extras I didn't know I needed like the desktop effects (xgl/compiz).
Am I happy with the direction the US is going? Certainly not. Be it IP laws, corporate protection at the cost of citizens, or the attorney general claiming that the writ of habeas corpus is not granted by the constitution, I'm not happy with the way things are going. But the great thing about the US is that it has a good (though not perfect) mechanism for changing the direction every couple of years. Right now, the technically inclined are noticing problems with IP. If it becomes significant enough to become a political issue, the country can change course accordingly. In many respects, our country has been in worse situations before, but we've always recovered.
With regard to the grandparent post, the US, like any other country, may fall behind in a certain area for a while. But we're not so stupid to sit on our collective ass for 20 years and allow the country to fall horribly behind the rest of the world.
I switched my Gnome desktop theme about 20 times last year, unable to decide whether a Vista-styled theme was glorifying or mocking the competition.
Then I decided that I still prefer Clearlooks.
Boy, what a year.
Do not trust this signature.
Some kind of corner has been turned for the GNU/Linux desktop in 2006.
I light off cups (that is, go to http://localhost:631/ in FF), enter th IP address of the printer in the obvious place, and stuff works.
It's a cheezy home wireless network; I really want the Dumbest Thing That Works, realizing that if there is a reset, DHCP may re-jigger things.
Trying to figure out how to set a printer by IP in that other OS has baffled me. It's an Easter Egg hunt gone ronngg. The quest for simplicity has been abandoned at a variety of levels.
At least I only have to suffer that OS at work.
Get thee glass eyes, and, like a scurvy politician, seem to see things thou dost not.--King Lear
I first tried Linux in 1997. At the time I couldn't imagine using it as a desktop. However, there were a few turning points for me:
1) GOOD package management. I started out on Redhat. Whenever anyone brings up RPM problems, they get reamed on Slashdot "RPM IS NOT A PACKAGE MANAGEMENT SYSTEM!" Well, once upon a time, there wasn't Yum or Red Carpet, and the best thing there was (RPM) was still hell to use. Now between RHEL and Gentoo, I rarely have to worry about not finding dependencies. Thank God.
2) 2.6 Kernel. The reason is because before 2.6, X under Linux always "felt" slow.
3. Firefox.
4. More expansive community, documentation. I remember in 1997 trying to get help:
ME: "I'm trying to do X and it's doing Y. Does anyone have experience with this? "
THEM: "RTFM"
ME, (looking): "The man page doesn't say anything"
THEM: "+b You've been banned, troll."
Now I look at the Gentoo install documentation and user forums now, and I am just in awe. Likewise for many of the other major distros.
Now that wireless is going smooth, the only thing I have to complain about is no matter what I do, font rendering is inconsistent and often ugly. But as of two years ago, I am a happy full time Linux user! Take this for what it's worth, I just wanted to share my experience.
Legacy systems, documents, and most importantly user training in said systems and documents. "If it ain't broke, don't fix it" rules when computers are the tool rather than the end goal in and of themselves, and it's hard to fault that logic. If you change your systems you're effectively "breaking" your employees in terms of their productivity, and fixing them is quite a job. It's only justified when the end benefits are worth the pain, and to be fair in most cases they probably aren't ...
Funny how Microsoft has gotten away with just that. Every version of Windoze has a few pointless GUI changes and little real improvement, yet the Dells of the world push it out. Vista and Office 2007 mark the largest GUI change in a long time. Legacy software is broken. Where does that leave the user's "faultless" logic?
Free software interfaces are more stable. Window maker, is a Next clone and it's basics have not changed in fifteen years. There are several others, like the fvwm or olvwm, and Enlightenment, that have been just as rock stable. At the same time there have been many other excellent interfaces that have grown up. All of them are extensively customizeable so that you can have as much change in each as you like and they all work together, so you can mix and match. The same performance from Microsoft would have Windows 3.1 GUI be adequate, customizable still available and easily interchangeable with a dozen other excellent window managers. Right.
The same arguments apply to file formats and hardware. Vista is bringing with it .DOCX, the M$ "open" format with a 6,000 page spec. It's also going to obsolete 54% of exiting computers and 94% of them are not really "premium" ready, so their users will soon be disappointed by an upsell that degrades their actual performance. DRM promisses to make it all that much worse.
The real hope is that Vista goes nowhere. XP did not move hardware and it had much better driver and legacy application support at launch. It took four long years for it to be majority. People want new hardware and it's time for it to move. There are major improvements that are good for both performance users and people who want something small and quiet. If Vista's changes are so bad that it actually harms sales, look for Dell, HP and others to follow Lenovo's lead to make up the difference. That would break the M$ monopoly once and for all and then we would not have to worry about this upgrade train nonsense.
Vista - the Ow is Now.
Friends don't help friends install M$ junk.
You forgot to name the "counterexamples."
Europe is taking our shiny new copyright and patent crap and running with it. . .
It's called the Berne Convention Treaty. America was shoved into coming in line with it; and ran with it.
America's shiny new copyright and patent crap is firmly rooted in the monarchial grants of absolute right and trade guildism that America's founding fathers firmly rejected.
Too many powerful people around the world want control of their respective economies, and one way you do that is by manipulating and suppressing technological advancement.
And adoption of the Berne Convention Treaty was one of the first signs that America was heading down this path. We gave up being the industrial might driving the economy of the world for being a bunch of paper traders.
If they can manage to hold themselves together as a nation China wins. If not India wins. In any case, we lose as our bits of paper become worthless on the international market.
Jesus we used to make some good stuff.
KFG
Cana-fuckin-da.
The role of the writer is not to say what we can all say, but what we are unable to say. -Anais Nin
Jesus we used to make some good stuff.
Amen. What's ironic, is that as I drove around certain parts of Western Pennsylvania a few weeks ago, is that it's not as if we don't have the capability to make "stuff" anymore -- the machinery, the productive capacity, is mostly all still there, albeit rusted, and the workforce is there, albeit unemployed and twenty years out-of-date -- it's just that the desire to do it disappeared and moved elsewhere, by virtue of some pieces of paper that swapped hands and certain handshakes between heads of state.
We have a government run by the "paper traders," as you put it, for their own kin; they have sold off the economy, piecemeal, to the benefit foreign interests and themselves, despite the obvious outcome: you cannot maintain a first-world economy and standard of living, when you are competing in a labor market with a billion-plus Chinese and Indian peasants. It just isn't going to happen, it's unsustainable: either the first-world country's costs and standards of living are going to sink, or the third-world's are going to rise, and the former is a whole lot easier and a lot more likely than the latter. (Think of it in terms of economic "mass," and of two bodies orbiting around each other; it's a lot easier to move 300 million people down towards the level of a billion poor ones than it is to move the billion up to meet the 300M.)
When the shell game is done, the U.S. is going to become a nation of aristocrats: the same paper-traders who have run the place into the ground, and thus knew from the beginning where it would end, and have moved their wealth into hard currencies; and everyone else, who will be stuck with their savings in a currency suddenly not worth the paper it's printed on (it's already not worth the metal its minted with), and forced to buy everything from abroad (since the country has long since ceased to produce anything of value), who will be stuck with the bill.
Take a look around: you're witnessing the decline of one of the world's great empires, which, like many before it, was brought down not by invaders from afar, but from mismanagement and greed from within.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
Sounds reasonable. I'd trade my old copy of W98 for half a dead chicken.
Sounds like a bad trade to me. You can make soup out of that chicken, but what are you going to do with those Windows discs? Even sauteed, I think they'd be pretty tough.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
I think that desktop linux is not ready because it still plagued by a problem of text configuration files. I'm perfectly OK configuring my debian box from various files in /etc directory, however most of the users e.g. normal people aren't, and as long as proper GUI configuration tools, like Control Panel in windows, are absent from KDE/GNOME desktop environments I don't think that majority of people would like to use it. And these tools would not be there for some time, because a few distros currently support common location of configuration files, and LSB itself is a joke: "LSB compliant system is the system that supports RPM package management". So until these things are going to get sorted out Linux will get mainstream, which is, I hope, just a couple of years from now.
And, yes I'm aware of Red Hat system-config-* stuff, the problem is, these utils aren't that great, an they not that well organized.
Your comment is correct. I spent two years in Haiti, the poorest in the western hemisphere. Did anyone run Linux, no. They used old hardware everywhere. Old hardware that would not run linux. I tried replacing the pirated copies with linux and failed! Bandwith is very expensice there at least when compared to Income levels. So downloading linux "for free" is actually much more expesive than the 50 cent Devils own copy of windows. You'd have to find an older version of linux that would run on the hardware ( unlikly in most cases) and then compare it the the windows equivalent. I'm sorry I love linux, but redhat 5 doesn't compare to win 2k on the desktop for new users. No, DSL linux didn't work doesn't matter beacause the oss applications running on top neccisary for real work ( openoffice or abiword) perform terribly on older hardware.
Technology flows from the first world to the third. They will be the last to get linux on the desktop, not the first.
Well.. maybe. Or Maybe not. But Definitely not sort of.
Is STANDARDIZATION. Linux will never make it until there is a hell of a lot more standardization. Not 50000 formats for releasing packages, but ONE that works everywhere. JoeSchmoSoft doesn't want to install 20 different distributions to test that their packaging works and to create, host, and support those 20 different package formats.
/bin, /usr/bin, /usr/local/bin, /opt/bin? What about the conf file, is it in /etc, /etc/progname/, /opt/progname, /opt/progname/etc, ...? This may be my being 'used' to Windows, but I prefer each program has its own directory where all of its files are, that way I know exactly where to look, and I can have a nice overview of what is installed.
In Windows, one installshield package does everything on any Windows version.
Nor does Grandma Gertrude want to download all 20 to figure out which one works on her system if all she knows is shes running Linux. Even if she does grab the correct package, she certainly isn't going to be able to open shell, su to root, and dpkg -i or rpm -Uvh the file. She will double click it, see nothing happens, and give up.
Something like VFW would be welcome, I install a video/audio codec, and it works in all applications. Also, some cleanup of the video code in general, why is there no decent video output, opengl has tearing issues even with __GL_SYNC_TO_VBLANK = 1, Xv is really pixellated, I forget the others now, but none look nearly as good as the standard overlay in Windows.
Standardization of directory hierarchy. Does that executable go in
Some just general stupidity as well, like I was bored and decided to try SuSE, so I pop in the 10.2 minimal cd which lets you install over the network. I boot and first off it takes like 2 minutes to boot into the installer, and its running like I'm on a 386, and if you've ever booted off the minimal cd, you'll see it has no reason to run like that. Anyway, I suck it up and figure it'll be okay once everything is on the hard drive. So initial question is where do you want the installer to get the files from, I was on a laptop, so I hit network, wireless, enter my WPA key, and off it goes, grabs the installer files, launches the installer, grabs all the packages, installs, and reboots to finish the configuration. Except... this time it doesn't ask for my network settings, nor save the ones I entered earlier, so I end up hitting skip on a bunch of files it needed for ending configuration because it couldn't see my freaking network anymore.
So I figure oh well I'll just configure it by hand when it boots, so I reboot and after about 5 minutes of waiting for it to boot up to a login screen I just hit the power button and boot off the windows CD and remove all the partitions and reinstall Windows.
Note: It is not the hardware, I had a Gentoo install on it ever since I bought it that worked fine, I just got tired of waiting for crap to compile all the time, and was hoping for a 'no-hassle' installation.
Other general stupidity...
- I have to install like 300MB of libraries in order to run Firefox if I'm running KDE.
- I have to wait for all those libs to load every time I open Firefox off of a fresh reboot.
- Total lack of standardized advanced GUI tools. What tool is good for administering what programs start when the system starts up? What about when the user logs in to X? How about something that tells me what video codecs are installed, what audio codecs?
- No apps seem to be 'lightweight'. Look at my two favorite windows media applications, Foobar2000 and Mediaplayer Classic. Foobar2000 is a 1.6MB download, supports every audio format under the sun, and loads almost instantly on a cold start. The closest thing in Linux is AmaroK, which is a 20MB download, and loads slow as holy hell, and doesn't offer nearly the range of audio support foobar does. Now MPC is a 1MB download (3mb? uncompressed s
In terms of being cryptic and user hostile, I agree that editing the registry and browsing to some random port on localhost are about the same.
However, there IS one important difference--it's quite easy to screw over a machine by mucking around in the registry, either by accident or because the instructions you found were incorrect. I can't compare that to simply browsing around localhost. What's the worst you could possibly do? Hit the wrong port and get a screen full of crap from chargen?
You know, I think I'm going to bookmark this post, because it describes the situation with every government office and employee that I've ever encountered in my life, to perfection. And to a lesser extent, most large corporations.
I'm not sure you know how right you are. (In fact, I hope you don't; and if you do, I feel your pain.)
The "training problem" is something that most technical people fail to appreciate, because it almost universally doesn't apply to them, because they generally have some conceptual understanding of how their software and hardware operates. Once you have that conceptual understanding, it's nearly impossible to imagine how it would appear without it. It changes the way you think about the tools you use, on a fundamental level.
Unfortunately, imparting that type of conceptual understanding to someone who isn't interested in learning it, is nearly impossible as well -- even when in the long run, it's almost certainly to their benefit to have it.
"Ladies and gentlemen, my killbot features Lotus Notes and a machine gun. It is the finest available."
The OS is a commodity, $9.99 at Walmart with Office is the Windows future.
I wish this were true, and it probably is true in the enterprise market. But I think it's unlikely in the home or consumer market in the near term. Items become commodities (in the non-pork-belly sense) when there are many suppliers, producing nearly interchangeable products, competing mostly on price. We're not there yet in the OS world. There are still only a few major players: Windows, Linux, various flavors of Unix, and assorted niche OSs. Price doesn't seem to matter much to consumers and OEMs: Linux is mostly free but Windows still has huge market share.
Even more important, the OSs aren't yet interchangeable. With a commodity like wheat or gasoline, it doesn't matter what kind you buy, because they're all basically the same (marketing nonsense like "Techron" notwithstanding). With computers, the OS still matters: it affects the user interface, security, training, and the applications you can run. There's also the network effect, where people tend to use an OS, or any other kind of software, because all their friends use it and they can get free support.
For what it's worth, I use and like Linux (primarily Ubuntu) at work. I mostly develop in languages like Java and Python, where "write once, run anywhere" is now finally true. However, I still use XP at home. I'm almost to the point where I can dump it, but I still use Photoshop occasionally (I hate the Gimp's UI), plus a few other tools (EAC, Quicken, some MP3 tools, a few games) that either run only on Windows and Mac or don't have easy to use Linux equivalents.
You must be one of those people who descend on Amazon and "tag" every single Microsoft product with "defectivebydesign".
No, but that sounds like a good idea and I'm glad that someone is doing it.
I wish you would all stay here and here only.
Hmmmm, you don't like me and I don't like you. Let's make a deal. You go away, your friends leave computer vendors, ISPs and everyone else alone and our paths will never cross. How about that? You get to pay Bill Gates for permission to use your computer, I get to use my computer and both of us are happy. Best of all, you can avoid advocating M$ shit to a free software web forum.
Now that would be super, but you've got a job to do don't you? Suck it up and get back to work.
Friends don't help friends install M$ junk.
That's exactly what I did, installed Slackware from diskettes in an old notebook with 16MB memory and 1.3GB HD. Runs Abiword and Gnumeric fine, what you need for the majority of office work. It originally had windows 95, but do you know what is the newest version of a Microsoft OS that will install in a machine with 16MB RAM? And how would you fit a Microsoft OS plus Microsoft Office in a 1.3GB disk with space left for user applications, unless it was w95?
The advantage Linux has is that you don't need a 1995 version of Linux to run in a 1995 machine. There are distributions made specifically for small machines
The only thing that keeps Linux from being widely used in the poorest countries is the same factor that keeps it from being more widely used in the USA: ignorance. The tragedy of it is that the poorer a country is, the most it would gain from switching to Linux.