OSDL's Review of Desktop Linux In 2006 200
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."'"
not to mention (Score:1, Insightful)
not encumbered by patents or opressive DMCA type laws
US kids in 20 years will ask: what happened ?
and you will say: well, we where in court arguing semantics and business methods while the rest of the world just got on with it
rAC
Re:not to mention (Score:5, Insightful)
Other countries (Score:3, Insightful)
I'm not seeing the appeal.
It's not happening (Score:5, Insightful)
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. [walmart.com] 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.
Desktop linux is in good shape, now it's users (Score:5, Insightful)
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.
Re:Other countries (Score:3, Insightful)
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?
Any you pay $179....
Or should you pay $98 or $95...
The OS is a commodity, $9.99 at Walmart with Office is the Windows future. Lets face it a $5999 dual core system not to long ago goes for $599 today. Everything in personal computing has gone down but Microsoft. Eventually the cost and perceived benefit is going to change. I even predict Microsoft Linux some day. Maybe 2010. M$ isn't dancing with Novell for nothing, they are milking corporate America for what they can get before the big switch. Makes the CIO/CIO feel like they know something to see a familiar "Windows" screen in the data center, OK, make that 2012.
I call this PONIIC, Price Of Not Investing In Change.
But when push comes to shove, and the PACRIM $500 3 TB database appliance comes through the door with Linux inside it makes a poor case for $1999 of W2003 and MS-SQL and you haven't bought the hardware yet. Think - more people in China, India and Russia are learning Linux than there are users in North America. This will make the CEO question costs of PONIIC.
Microsoft knows where the growth is. That is why it is $0.50 in other countries an not in the US. Hold it off as long as possible. Nice way to beat anti-trust too because in theory it should be the same price in Taiwan as the US. Not technical reason, Windows is loaded with enough DRM and remote control a simple update could shut the pirates down in a shake.
Re:not to mention (Score:3, Insightful)
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.
Linux has come a long way (Score:5, Insightful)
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.
Re:not to mention (Score:2, Insightful)
Re:not to mention (Score:5, Insightful)
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
Re:not to mention (Score:3, Insightful)
Cana-fuckin-da.
Re:Printing (Score:5, Insightful)
If you want to add a new printer there should be an "add new printer" tool somewhere obvious, like under the System menu. Bonus points if it already detects the attached printer for you, and if the system can be configured to pop up the add-printer dialog any time you plug in a new printer.
Overlooked... (Score:3, Insightful)
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
Re:Outside the US for now (Score:5, Insightful)
If you change the steps, the order of the steps, or the location of the steps, and they're LOST. Not only are they lost... they're angry, unhappy, less productive, complaining, and in need of re-training. It doesn't matter if the new software is better or "just as good". It doesn't matter if the platform is better. They know how to be productive when they're following these specific procedures. If changing the software changes the procedures, you have to re-train them in the new procedures, and you have to deal with all the productivity lost while they learn and adapt. And that doesn't include the pushback from the ones who resist change out of fear or inertia.
Any exec or front-line salesperson who uses ACT!... never going to switch platforms until ACT! supports that platform. Seriously.
And that's where your hurdle comes in. Change is neither easy nor painless. Imagine a pain meter on a scale from 1 to 10. Let's say that Windows is a 5 and Linux or Mac is a 2. But the adjustment of switching is an 8. People will opt to stay with the 5. They know the 5. They know they can tolerate the 5. Because even though the 2 is promised, the 8 looms large in the immediate future.
What's going to prompt people to switch is when the combination of Microsoft arrogance and aggressive bad guys raise the pain of Windows to a 6.5, while the efforts of Linux and/or Apple developers lower the pain of switching to a 6.5 or lower. When switching is no more painful than staying the course (or possibly even less painful), you'll see the needles start to move in bigger ways.
Re:My 2006 report (Score:3, Insightful)
About three years ago, I ran into a server (HP, if I recall correctly) that had the strangest problem I'd ever seen: Neither Gentoo or FreeBSD would run on the thing. You could install it, but the thing would randomly kernel panic within 5 minutes of being up, and you couldn't trace where exactly the crash was.
If, on the other hand, you shoved Windows XP onto the thing.. it would run perfectly fine.
To this day I'm still not quite sure what was wrong with that box.
My point? Nothing's perfect. The problems you described having aren't necessarily indicative of problems with the distro-- of course, you didn't really go into any real detail on WHY the installs failed-- and for every report of one of a "Linux just won't work on this machine" type of trolling (not saying you are, but I see those all the time around here) in favor of BSD or whatever there's a thousand reports of it just being fine on a given hardware.
Your mileage may vary, I guess...
You can lead a manager to water.... (Score:4, Insightful)
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.
OS as a commodity (Score:3, Insightful)
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.
Re:Outside the US for now (Score:3, Insightful)