| |||||||||||||
The Future of Linux is Google (PCWorld)The Future of Linux is Google (PCWorld)Posted Nov 17, 2009 21:15 UTC (Tue) by elanthis (guest, #6227)Parent article: The Future of Linux is Google (PCWorld)
The future of Linux on mobile devices is Google, perhaps. The future of Linux on the server certainly sees some Google influences, and hopefully will see more in the future as Google gets its internal kernel patches queued up for mainline.
Linux on the desktop and home theatre system space doesn't look to be very Googly at all, though. Granted, Linux isn't exactly the most popular OS by any stretch of the imagination in the desktop space and it's a little shaky on the HTPC front... MythTV is popular with home builders, TiVo and Comcast DVR use a Linux kernel, but Windows Media Center still has a very strong presence.
(Log in to post comments)
The Future of Linux is Google (PCWorld) Posted Nov 17, 2009 22:28 UTC (Tue) by Cato (subscriber, #7643) [Link]
Don't forget Chrome OS on netbooks and even some desktops ('nettops') - for quite a few people, a 'cloud terminal' is all they really need, and it would certainly help Linux with hardware support, as with Android.
The Future of Linux is Google (PCWorld) Posted Nov 17, 2009 23:48 UTC (Tue) by mikov (subscriber, #33179) [Link] Linux on the desktop and home theatre system space doesn't look to be very Googly at all, though. I wouldn't lump those together. Unfortunately, there is no Linux on the desktop, with or without Google, and from the look of things there will never be.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 0:32 UTC (Wed) by mattdm (guest, #18) [Link] I wouldn't lump those together. Unfortunately, there is no Linux on the desktop, with or without Google, and from the look of things there will never be. Wait, what's this thing I'm look at in front of me? Am I hallucinating or something?
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 1:17 UTC (Wed) by stumbles (guest, #8796) [Link] Eyez mush had taken da same pill az yoz, cuz Iz two iz luuking atz a Linux desktop.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 2:30 UTC (Wed) by JoeBuck (subscriber, #2330) [Link]
My non-technical wife and 11 year old daughter both use desktop Linux, though to be fair this would be hard to pull off without an expert in the family.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 2:47 UTC (Wed) by mikov (subscriber, #33179) [Link]
All of the LWN readers probably use Linux on their desktops, but that is not representative of anything.
There are at least a couple of things that would need to happen in theory before Linux can have a meaningful chance on the desktop:
Neither a) nor b) is going to happen for obvious reasons. Consequently, Linux has no future on the desktop.
Don't get me wrong - I am not happy about that at all.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 3:19 UTC (Wed) by lambda (subscriber, #40735) [Link] Not representative of anything? You mean, the LWN users aren't anyone? At my previous job, most developers used Linux desktops. My roommates in college all used Linux desktops (or laptops). Several of my coworkers use Linux desktops at home. My non- technical girlfriend uses a Linux laptop. Sure, it's a niche and not a mass market, but it's still pretty widely used. I'm not really sure what point you're trying to make by saying these uses are "not representative of anything" or that there's "no future." There seems to be a perfectly viable niche that is at least remaining steady, and likely will for the foreseeable future, of developers and reasonably technical users (and their family and friends, on occasion) using Linux on the desktop (or laptop). It's a niche, but that doesn't mean it doesn't exist.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 3:40 UTC (Wed) by mikov (subscriber, #33179) [Link]
The niche you are describing is not what is commonly meant by "Linux on the desktop". Linux is typically installed on desktop PCs either as a highly specialized tool for developers or by hobby enthusiasts.
(Yes, I am aware of attempts for desktop Linux deployment in schools and in governments in Europe and I applaud them, however unless the preconditions I described earlier are met, they will unfortunately remain isolated cases).
And, I would like to repeat, I am not happy about this, but I am calling the facts as I see them.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 6:02 UTC (Wed) by set (guest, #4788) [Link]
Well, Im not sure any statistical 'facts' have emerged in this thread, but I could offer up an anecdote. Of the 10 or so people I know personally that have Linux installed on computers/laptops in their home, 7 of them use it exactly as a desktop. ie. their primary or only OS for everyday computing. Ive been using it that way for 15 years now. I *am* a hobbiest, but my hobby computers are legacy unix machines, like DEC, HP, Apollo, Sun, SGI, etc.
Now, none of these people have ever purchased the Linux they run, so they are hard to count. Their ages span a range from 40 something *cough* to gradeschool, and they all live in midwestern US states.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 9:32 UTC (Wed) by dmk (subscriber, #50141) [Link]
linux on the desktop has a future. or at least the possibility of a future outside the niche.
in my opinion, the adoption of the linux desktop is hindered atm because of the peer-to-peer support thingy. people want to talk in their coffee-breaks about there problems with the machine. and they need their co-workers to give them tip's like: "when my machine stutters i just wiggle the mouse two times left and then two times right and all's well...
the typical desktop user at the moment (those which have, for example, problems of grasping the simplest concepts, like "programs"... yes, there a different entities on that screen!) is getting more and more replaced by a generation that is used to computers and internet.
for those it will be more natural to get support on the internet via mail, web-search and social-networks. this will benefit the linux-desktop.
on the same time, software development in one firm has serious scalability problems. look at how big ms is. this is not sound. they are bound to have overhead-issues and inefficencies.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 9:30 UTC (Wed) by fb (subscriber, #53265) [Link]
Pretty much any mobile phone shop now sells (locked down) Linux phones running a Open Source environment (Android).
However, all computer shops in the city center of where I live will _not_ sell any laptop or desktop with Linux pre-installed. Heck the only straightforward and sure way I know of avoiding the "Microsoft tax" on a new laptop is to pay "Apple tax".
[...]
BTW, I disagree that the Linux desktops niche has been steady.
The Linux desktop lost a lot of mindset. Some years ago, the majority of people I knew doing academic research used Linux in their laptops. A minority used Windows. Nowadays all those folks that had Linux migrated to Apple.
The biggest player in the desktop market had a huge flop with Vista. Golden opportunity for the growth of any alternatives. My impression is that Linux desktop presence didn't gain much during this flop. Do you think it did?
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 10:58 UTC (Wed) by nye (guest, #51576) [Link]
Where *are* all these people that are using Macs?
I've always wondered why the majority of commercial software, including almost all the worthwhile games, get Mac ports, but very little is ported to Linux.
In the last decade, excluding the CS department when I was at university, I've seen a grand total of *one* Apple laptop, and *zero* apple desktops. Compare that to one Linux laptop and three or four Linux desktops. Plus probably a couple of hundred Windows machines.
If we include the CS students, the Apple count goes up by one laptop, and the Linux count by one or maybe two.
Everyone seems to know of Macs, and think they have some idea about them, but so few people have ever actually *used* one that I don't understand why they're so well-known.
Perhaps there are more of them in the US, whence we import all our technical news and opinions?
who has a Mac? everyone! Posted Nov 18, 2009 11:24 UTC (Wed) by tialaramex (subscriber, #21167) [Link]
Every meeting I've been in, for several years, has been dominated by Mac laptops. Some of these people are hackers, some are decision makers, some journalists, they all buy Apple laptops. They work for the government, for private businesses, for pseudo-state entities like the broadcaster or railways. Macs are everywhere.
At work, we write software which runs on Linux (some of it is Free Software) and Java web applications which run on servers that in turn run Linux. Our 6-7 person technical team consists of one person with a Linux laptop (me), five with Macs including the CTO (who is also a perl hacker when he's not busy in meetings with suits) and one guy with a Windows laptop.
I'm the only one who can run our most important new software natively on the machine he uses to read mail and the web, but today that doesn't matter, everyone can have a VM, or just connect to a development server over the Internet, and they're more comfortable in OS X.
who has a Mac? everyone! Posted Nov 18, 2009 14:18 UTC (Wed) by nye (guest, #51576) [Link]
Your experience is so vastly different from mine that I can only assume it is either regional, or because you mostly work with/know people in the technology sector, where I would expect Macs to have a higher market penetration. My (unfounded) perception is that Mac users are more likely to be technology enthusiasts who have made a deliberate choice.
Or possibly, given the cost, they're mostly owned by people who are rather outside my social class.
It would be interesting to see how Apple rate their market share across different countries and demographics, but I'd be surprised if they published such information - it would similarly be lovely if Google's zeitgeist still included data on OS market share. I guess the data-mining opportunities are to great for them to want to share that.
(One final point which may or may not be useful depending on what you think his opinions are worth: even Steve Ballmer thinks that Linux is more popular and a bigger threat to Microsoft)
who has a Mac? everyone! Posted Nov 18, 2009 18:18 UTC (Wed) by mikov (subscriber, #33179) [Link]
My experience is exactly the same. All our products (server and embedded) are Linux and Java based based, but I am the only one who actually uses Linux on all his desktops and laptops and can run them natively on my machines. (Even though the software is Java, it is still OS-specific to a large extent with JNI, native system tools, etc).
Most technical consultants and businesses I have dealt with have Mac laptops (with a copy of Windows running in a VM).
We _tried_ to convert to Linux a couple of years ago because in theory it sounded good - after all our products are Linux based, it would be cheaper, no viruses, etc. The experiment failed miserably and it gave me a lot of insight in the subject.
I believe these are the actual objective reasons of why Linux cannot grow its market share despite being free and technically better. It is not only because Windows comes pre-installed on PCs:
- It is practically impossible to develop closed source commercial desktop software for Linux. (Which distribution? Which package manager? Which desktop environment? Which version of all the before?)
- It is extremely difficult from a business perspective to develop Linux drivers for hardware devices. Which kernel version do you target (or which distribution specific version)? How do you distribute bug fixes and updates?
who has a Mac? everyone! Posted Nov 18, 2009 19:18 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]
answering your objections in reverse order
for hardware support you target getting your driver in the kernel.org kernel. all the distros pull from that (and most of the backport drivers)
as far as developing closed source apps for linux, this is not as hard as you make it out to be.
you don't need to target a specific desktop environment during development unless you _really_ are worried about making the look-and-feel match the desktop (and if you are doing that you need to support skins/themes anyway) look at freedesktop.org for how to do things in a desktop agnostic way.
I agree that when it gets to final compiling and testing of the application it's more work to test with the different distros. but in most of the time the problem is more in packaging dependancies than actually needing different binaries (you can get unlucky with one distro updating a library in a way that's not backwards compatible, but this is uncommon)
if your entire QA effort is manual then this can be backbreaking, but if you have automated testing of your application (which is becoming more and more a requirement anyway), then adding additional platforms to test on is relativly cheap.
it is different than latest-version-of-windows-only development, but nowdays even windows development requires testing on multiple platforms (XP Vista, 7 at minimum)
who has a Mac? everyone! Posted Nov 18, 2009 19:28 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]
also on the windows platforms you need to test, don't forget that you need to test with every service pack
so in windows you have three 'distros'. XP, Vista, Windows 7, possibly 98
each of these have several releases (plain, SP1, SP2, etc)
in Linux you have a few families of distros to support RPM distros (RHEL, Fedora, Suse) and deb distros (Ubuntu and Debian) not much worse than the windows
each of these will have more than one release version out at one time (2 for Debian, 2 for Fedora, 3 for Ubuntu + LTS versions, etc), again, not much worse than windows
who has a Mac? everyone! Posted Nov 18, 2009 19:45 UTC (Wed) by boudewijn (subscriber, #14185) [Link] At the company I work for, we develop an application that has binaries available for Windows, OSX and Linux. It's partly open source, but there are closed source plugins, and in any case, given our development model, nearly everyone uses the binary installer anyway. I'm not sure whether the app is within the definition of commercial software -- but there is no volunteer work done on it. We use Qt, of course, otherwise we wouldn't be able to develop the software with 3 to 5 developers. And on Windows and Linux we use bitrock to create an installer. We only target 32-bits linux, though some people have compiled the open source parts on 65 bits linux, mostly gentoo. It works perfectly fine on my OpenSUSE and Kubuntu systems. We have about 60,000 Windows users, about 5000 OSX users and about 300 Linux users. Despite having a lot of automatic tests, all operating systems give us pain: on OSX it's hard to support 10.4, 10.5 and 10.6 with one binary, and besides, we've only ever got access to the latest Apple hard and software, For Windows, side-by-side errors are bad. One version of Visual C++ actually creates corrupt manifests, The differences between XP, Vista and Windows 7 are particularly time-consuming. For Linux, the biggest problem is running the autotests: our test machine doesn't have X11... And we had a lot of problems with different versions of gstreamer. All in all, if you develop your commercial desktop software using Qt, use the Bitrock installer and have decent automated tests, providing your application for 32 bit linux as well doesn't cost a lot of extra effort.
who has a Mac? everyone! Posted Nov 18, 2009 20:06 UTC (Wed) by mikov (subscriber, #33179) [Link]
How does Bitrock handle .so dependencies (even Qt itself) in different distributions?
The (only?) positive thing about Windows is that binary compatibility between OS versions is an explicit goal. While I still was a Windows developer we had no problems moving software (including our kernel drivers) between NT4, W2K and later to XP. Of course I don't claim this is universal, because our packages were pretty self contained. Also, i realize that Vista introduced more significant changes and breakages compared to, say, XP.
who has a Mac? everyone! Posted Nov 18, 2009 20:25 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]
window's binary compatibility is like Java's source compatibility. it works all the time, except when it doesn't
yes it's a goal, and yes it works a lot of the time. but it doesn't always work.
for linux systems, you can take just about any binary from one distro and run it on another distro. the .so files are all named the same way so if they are there everything works.
the cross-linux-distro problem is more a matter of how to make sure that all the right packages are there, and where are all the config files than anything else.
there are a lot of companies selling software for linux, many of them charging lots of money for their software. some of them take a hard line attitude like 'we only certify RHEL 5.0, if you upgrade to RHEL 5.1 you are on your own', but many of the recognise that the distros are all more the same than they are different and instead say 'we have tested this on RHEL 5.0, it should work on anything else, contact us if you have problems'
the companies that take the first attitude are very slow to support any new distros or releases, frequently when a customer says 'I will buy your product if you support distro X' (I've done this to a few products).
companies that take the second attitude tend to be much faster to add 'supported' distros to their list.
who has a Mac? everyone! Posted Nov 18, 2009 21:30 UTC (Wed) by boudewijn (subscriber, #14185) [Link] Basically, you package everything you need, Certainly all the Qt libraries you use, but also other libraries, we're not using the system Qt. It doesn't solve everything: the gstreamer problem I mentioned above was because we didn't want to package gstreamer as well, and the gstreamer on ubuntu and gstreamer on OpenSUSE weren't binary compatible, so when we packaged the Qt Phonon library, which uses gstreamer, we had problems. (Of course, Phonon couldn't play mp3's on XP, but could play wav's while it was the other way around on Vista.) The set of libraries we package is almost the same on Windows, OSX and Linux, by the way. On Windows we have to package the c runtime, as well, that is the main difference.
who has a Mac? everyone! Posted Nov 18, 2009 21:18 UTC (Wed) by dbruce (subscriber, #57948) [Link]
Two observations regarding laptops:
1. In my "real job" (surgery) - at least 95% of folks use Windows laptops, with the rest using Mac. I've never met another MD in real life who uses Linux, and perhaps 5% have even heard of it.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 11:32 UTC (Wed) by fb (subscriber, #53265) [Link]
I meant to talk about academic research staff (not the students). As I said, right now most in this group use Macs or Windows.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 16:27 UTC (Wed) by shmget (subscriber, #58347) [Link]
Where *are* all these people that are using Macs?
Google is your friend:
http://itc.virginia.edu/students/inventory/compare/
Just an example among others....
Yes it is US-centric. Mac is making much less in-road on the other side of the Atlantic for instance...
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 17:43 UTC (Wed) by drag (subscriber, #31333) [Link]
It's very regional.
I have yet to see a _SINGLE_ Mac OS laptop in any meetings or with anybody
A few people developers have Linux laptops, but by and far the vast majority
I know that there is a single Mac floating around somewhere, due to me
Mac users Posted Nov 19, 2009 15:22 UTC (Thu) by sumanah (guest, #59891) [Link] Yes, I imagine the extra cost of shipping Apple hardware to Europe and Australia is high enough that it affects what proportion of users choose a Mac.The first few rows of a lecture at a university in Missouri (middle of the USA). Zoom out. Nearly every single one is a Mac.
Mac users Posted Nov 19, 2009 15:57 UTC (Thu) by boudewijn (subscriber, #14185) [Link]
Macs cost in euros what they cost in US dollars in Europe generally
speaking -- even though the the US dollar is worth less than a sesterce after Obelix is done producing menhirs for the Roman market. But I doubt the reason for that is transport.
Mac users Posted Nov 19, 2009 16:18 UTC (Thu) by nye (guest, #51576) [Link]
Wow. Regardless of the type of machine, I'm amazed at how many there are. Is that typical? I graduated with a CS degree from a fairly highly-rated UK university in 2007, and there were never more than a handful of laptops taken into lectures. I wonder if things have changed so much already...
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 5:23 UTC (Wed) by Per_Bothner (subscriber, #7375) [Link] Er - hardware manufacturers do develop and supply Linux drivers. Intel chipsets and video, AMD/ATI likewise, HP printers are some examples just off the top of my head. Last week I bought a HDHomeRun tuner which advertises on the box "Compatible with DVR/PVR software for Windows, Macintosh, and Linux" (and they apparently did or at least contributed to MythTV support - which works very nicely, thank you).True, not all hardware manufacturers support Free OSes, but many do.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 18:33 UTC (Wed) by mikov (subscriber, #33179) [Link]
Perhaps you mean most hardware manufacturer don't support free OSes, but some do.
Put yourself in the position of a hw manufacturer who wants to support Linux. You have just designed a new device. How are you going to support Linux?
- Obviously you cannot rely on your driver being incorporated in the upstream kernel. One, it may take an unpredictable amount of time, and two most people are not likely to start using the upstream kernel for years. Exactly the same problem applies to driver updates.
- So, you need to develop and maintain an out-of-tree module. For which kernel version and which distribution? In practice you need constant maintenance and testing for many kernel versions at the same time. What is even worse, there is no easy standard way for people to install your out-of-of-tree module.
- There are additional complications. Since you are a good Linux citizen, you have submitted your module upstream, while at the same time continuing to maintain the out of tree module. Eventually some of your customers start using the upstream kernel and you end up in an even worse position. You have to support many versions of the out-of-tree module and versions of the upstream kernel. What is worse, you have no direct control over bugfixes and updates in the upstream driver anymore. If a customers are having problems you cannot just tell them to go complain on LKML.
To be honest, I really understand why most manufacturers don't bother. You need to have a team of full time paid kernel developers for this. Perhaps it makes sense for huge manufacturers (Intel, Nvidia) and for extremely popular products, but not for anything else.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 18:43 UTC (Wed) by drag (subscriber, #31333) [Link]
Well your wrong.
OEMs only support the stuff they ship. If your using something other then
That is the way for any pre-installed OS from any manufacturer.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 19:23 UTC (Wed) by mikov (subscriber, #33179) [Link] If your using something other then what they have installed on the machine or otherwise officially support then your on your own. I think this sums it up nicely. People probably don't want to be "on their own" if they upgrade their PC, or if they happen to plug a new USB device in it, so they don't use Linux. One of the fundamental reasons for success of the PC platform has always been the openness and the ability to mix and match components. If you remove that...
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 19:29 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]
if you use windows you are in the same boat.
if you install a driver that comes with hardware and your system no longer works do you think microsoft, HP, or dell are going to take care of it for you?
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 21:41 UTC (Wed) by drag (subscriber, #31333) [Link]
That is one of the things that Linux has problems with.
The best you can expect from OEMs is if you nuke your system and install from scratch and it still does not work then they may take back your system and work on the software if it is in warrenty.
Microsoft will give you a shit.. I think a troubleshooting phone call is a minimum of 75 dollars, after that they charge a large fee per minute or bill by the hour or something like that. It's been a while since I last looked at it, but for a consumer Microsoft's support costs are attrocious.
There are a couple issues here:
1. There is a large turn around for accepting kernel drivers into the kernel. The process is long and drawn out and can take 6 months to a year to get new hardware supported properly in the kernel.
This is a serious issue for OEMs, but it can be dealt with.
The much bigger one is #2
2. There is no Geek Squad for Linux. Anywere in any part of the country you can walk into computer stores and get help with Windows. You pay them 300 bucks to remove viruses or some other thing, but it's still avialable.
When people want to learn how to use Windows and Office they don't go and search mailing lists, they go and pay for class time in a community college. They do this because it's worth paying 300-700 dollars to learn Office and Windows because it'll help them get a job.
With Linux there is almost no places were people can take it in and get help. You go into Best Buy with a Linux netbook they will happily sell you a 'student' license for Windows and OFfice and charge you a hundred bucks to install it on your netbook, but they are going to refuse to help you get a printer or VPN working.
This is a societial issue. The negative costs of running Windows is absorbed by the society at large as just a cost of doing business or running a PC... Linux has no similar infrastructure.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 18:56 UTC (Wed) by dlang (✭ supporter ✭, #313) [Link]
if you wait until the device is complete before starting to get the driver upstream you will have the problems you describe.
however, if you involve the kernel developer earlier things are much easier
there is the staging tree and the help from Greg to get things into the kernel.
most distros backport drivers that are added in later releases into their supported distros.
however, the biggest question to start with is if your device really needs a completely new device driver to start with?
if you are designing a new chipset, you are probably in a similar category as the 'huge manufacturers' you mention. but 99%+ of the time new hardware is not built using newly designed chips, it's built using existing chips with slight tweaks and in new ways. drivers for this category of devices are _much_ simpler to develop, and usually result in tweaks to existing drivers and not completely new drivers.
The Future of Linux is Google (PCWorld) Posted Nov 20, 2009 3:52 UTC (Fri) by smokeing (guest, #53685) [Link]
What does it take to support a piece of hardware? It takes to abide by whatever underlying industry standard is.
Back in 2007, we bought a Motorola Z6 ROKR, and it wasn't supported as a storage device at the time. After googling the matter, it appeared it misbehaved during handshake with the host, violating USB protocol, and needed a "quirk" (specifically, its product id needed to be listed in appropriate header; this was already done in the next kernel version).
If Motorola had run the phone through proper QA procedure, it wouldn't have broken the standard, and no drivers would be necessary *at all*.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 5:31 UTC (Wed) by jmm82 (guest, #59425) [Link]
"a) There should be commercial desktop software development for it."
I don't want to pay for software.
"b) Hardware manufacturers should develop and supply Linux drivers."
I see *a lot* of companies with Linux driver support. Intel, Sierra Wireless, Atheros are three companies I have used drivers for and all seem to be working with the community to improve support for their hardware.
"On the other hand I believe the situation for fixed hardware devices, embedded and servers is much more positive, and in the context of this article Linux doesn't need help there (from Google)."
I just bought an Android and it is the first phone I have seen to give Apple a run for their money. It may run Linux, but if I did not follow Linux I would not know this and that is a good this. Most Linux systems are TOO flexible and complex for the average user to understand. GNU/{Gnome,Kde}/Linux ,may never win the desktop, but Google/Linux a simple desktop with a slick interface and good Google app integration hiding all the complexities of Linux, but still gaining the stability in the kernel, MAY. WHEN Linux does win the desktop I have a feeling most people will not even know it is Linux.
I personally could care less if only non-existent LWN users are the only Linux desktop users, I like my red-headed stepchild desktop just the way it is, warts and all.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 6:44 UTC (Thu) by lysse (guest, #3190) [Link]
> All of the LWN readers probably use Linux on their desktops
Some of us don't use desktops on our Linux, though.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 9:29 UTC (Thu) by Klavs (subscriber, #10563) [Link]
I must say that, it looks to me as if Linux does indeed have a future on the desktop.
This is based on several indicators.
1) the danish "public tv" channel (also the largest channel provider AFAIK) had a story about how a regular family switched their computers to using Linux - and how much money they saved. And after the family tried it for a few weeks, they followed up and the family decided to stay with linux on all computers but one - for gaming.
2) I own linuxpusher.com - which was mentioned in the aforementioned show - and got a LOT of orders for Linux (Ubuntu in this case) - and I often talk to normal people who want to try Linux - because their windows has gotten too slow etc. or they are just curious.
3) I recently installed Linux on my brothers computer, because his Vista didn't work properly with his 3G phone modem - Linux (Ubuntu 9.04) worked flawlessly - so he's still running Linux - without any help from me (the install was the only thing - and he could have done that himself easily).
So it seems to me as Linux is making inroads in to the desktop market - atleast in demark.. slowly, but still.
The Future of Linux is Google (PCWorld) Posted Nov 22, 2009 18:29 UTC (Sun) by nixternal (guest, #41048) [Link]
"There are at least a couple of things that would need to happen in theory
before Linux can have a meaningful chance on the desktop:" "a) There should be commercial desktop software development for it."
I think you must have meant "There should be MORE..." as there is
"b) Hardware manufacturers should develop and supply Linux drivers"
Once again, I think you must have meant "There should be more hardware
I feel there is a chance for an eventual mainstream adoption of Linux. For
Now if we could only get some real marketing folks behind the Linux desktop
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 10:10 UTC (Wed) by edmundo (guest, #616) [Link]
My 4-year-old daughter uses desktop Linux. It's an antique version of Debian running on an antique ThinkPad. She can log in and run Tux Paint and draw fairly recognisable pictures, which is quite impressive if you take into account how much fluff there is in the 15-year-old mouse and how sticky she keeps her mouse pad, which sometimes also serves as a mat for her hot chocolate mug when she gets back from school.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 23:56 UTC (Wed) by drag (subscriber, #31333) [Link]
Open up the bottom of the mouse.
Let the ball fall out. Take the ball and clean it up with some cool soapy
Inside the bottom of the mouse you'll notice some rollers. Usually there
If you take your finger nail and scrape down the roller you can usually
With that crust scraped off the mouse will run like new. :)
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 6:50 UTC (Thu) by lysse (guest, #3190) [Link]
...for approximately three weeks, before it gets covered in just as much fur as was there before. I don't know what it is about mice, but once they get thoroughly linted up they seem to prefer it as their default state and seek to return to it as quickly as possible; what took several months to arrive takes just days to return.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 12:33 UTC (Thu) by NRArnot (subscriber, #3033) [Link]
Please, PLEASE buy the kid an optical mouse! Just because she's only four doesn't at all mean she can't experience extreme frustration when the mouse cursor doesn't follow what she's doing with the mouse.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 10:00 UTC (Thu) by forthy (guest, #1525) [Link] What's the difference to a Windows user? My father has Linux, installed by me through the standard OpenSuSE install (i.e. select language and contry, desktop environment, root password and user account, and nothing more), and his major problems with it just come from age, and would be the same as with Windows. His PC never needs hand-holding, it just works. Compare that to my sister, who uses Windows, because she wants to run some (pirated) proprietary software like AutoCAD. This PC needs far more handholding from me than the Linux box. If she wouldn't need AutoCAD for her job, I'd just give her a Linux box, too. Sh's quite afraid of Linux, since she only knows it from my usage pattern with four xterms and an emacs window ;-). I told here that my Windows desktop (when I port something to Windows) looks exactly the same.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 18:48 UTC (Wed) by leoc (subscriber, #39773) [Link] IMHO Linux doesn't need to have majority desktop market share, it only needs enough to become self sustaining (that is to say that Linux users can do everything they want to do without having to use something else). IMHO Linux is just about there, as there are very few things you simply cannot do with Linux on the desktop at this point.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 1:30 UTC (Thu) by jmm82 (guest, #59425) [Link]
All I am missing is Silverlight in my browser so I can watch Netflix instant movies, but I have a Virtual Box windows 7 install which acts as a over glorified internet explorer. ITunes would be nice, but I get by without.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 15:52 UTC (Thu) by Trelane (subscriber, #56877) [Link]
You can get silverlight in your browser (well, moonlight anyway). What you can't get is Microsoft's DRM.
The Future of Linux is Google (PCWorld) Posted Nov 18, 2009 7:27 UTC (Wed) by Requiem (guest, #51519) [Link]
Of course Linux has no presence on the desktop. Because the roughly 10-20 million desktops and notebooks running Linux don't count.
The Future of Linux on the Desktop? Posted Nov 18, 2009 14:58 UTC (Wed) by dowdle (subscriber, #659) [Link]
I assume you were being sarcastic when you said that 10-20 million Linux desktops don't count. I'd argue that the numbers are actually larger than that (probably by as much as 2x) but let's stick with a medium number of 15 million... for my discussion below.
Some people want commercial software on Linux, some don't. I attended the Utah Open Source Conference 2009 in Oct. and attended a presentation by a big wig from Adobe where he talked about FLOSS and Adobe. Of course the usual question came up about when will we get Photoshop and various other Adobe products for Linux and the answer was something like, "when there are enough Linux users to guarantee sales of at least 50 million copies". That is a rather high hurdle. Seriously, you have to sell 50 million copies of something before it becomes profitable? What a poorly run company you must have.
But seriously, there are enormous opportunitities in a number of software genres for FLOSS shareware... or FLOSS support funded development. Educational software for K-12, games, small business apps.
The truth is that there is a huge software catalog currently available for Linux any anyone who has been using it knows that. There are some bald spots but those are where many of the opportunities lie.
We believe in the Open Source development model and I also believe in Free Software... and yes there are a number of viable business models available for those groups who can find good business-minded leaders.
Mac hardware sales (especially laptops) have supposedly fluctuated up and down over the last few years. Yes, Mac OS X has poached some Linux users. Anyone who has been to a Linux or FLOSS oriented conference can tell you a significant number of presenters seem to be on Mac laptops... but the good thing about that is how stubborn Apple is. They will only allow Mac OS X to run on their hardware, which is very limited in the variety of hardware components they ship. Mac OS X's growth potential is limited to the number of computers Apple can make and sell. While for a single hardware company they do very well compared to folks like HP and Apple (not talking servers or netbooks of course), a significant portion of their income comes from iPods, iPhones, and the iTunes Music Store. Their concentration on their computers has waned and I don't see that changing unless they finally decide to offer Mac OS X on non-Mac hardware. I don't see that happening... mainly because of the barriers that exist making it work with a larger variety of hardware that exists in the generic PC market. Apple seems to have basically thrown in the towel with regards to servers.
Apple computers are really just another hardware platform targeted by Linux.
I don't really care if Linux ever goes "mainstream". I've been using it as my primary desktop since 1995 and it has gotten so good these days... and as long as it continues to progress is all I care about. I have fears that if Linux ever does go mainstream, I'll have to suffer as it is dumbed down for computer novices... but luckily with the number and variety of Linux distros that are out there, that probably won't be a problem.
I really don't want to have to deal with masses of computer newbies asking me Linux questions and expecting free support... because I'm a Linux guy... and giving free support to friends is what I've been doing for years anyway.
The reality is that 1-5% of the desktop market share actually is a significant number of computers and a large userbase. If Nintendo can make a fortune on the Wii... with about the same userbase as Desktop Linux, why can't software vendors? I realise that with console, hardware diversity doesn't exist and that the hardware diversity that exists in the generic PC market is a barrier for some software (particularly games and multimedia apps) but come on. Many FLOSS projects have accepted the challenge as well as a handful of companies... but it would be great if more commercial FLOSS could fill in the cracks.
Of course the other major barrier that exists is getting more OEMs to offer Linux as an option on more of their hardware... and not to charge more for it.
The Future of Linux on the Desktop? Posted Nov 18, 2009 18:09 UTC (Wed) by anselm (subscriber, #2796) [Link] Of course the usual question came up about when will we get Photoshop and various other Adobe products for Linux and the answer was something like, "when there are enough Linux users to guarantee sales of at least 50 million copies". Adobe bought the rights to Photoshop in 1990, when there were barely 50 million Macs around in the whole world (if that many). Presumably not every single one of their owners would have been interested in buying Photoshop. And that was back when Photoshop was a fairly simple program and still needed most of its code written. Nowadays they have a codebase that only requires porting and they need 50 million copies guaranteed sold? I want to be an Adobe engineer, they must be so well paid.
The Future of Linux on the Desktop? Posted Nov 19, 2009 15:15 UTC (Thu) by smokeing (guest, #53685) [Link]
I second the concerns about the dumbification of Linux as an inevitable consequence of desktop Linux advances. Most depressing is the sight of some essentially un-personalised, uncared for desktop with the same Start button in lower-left corner but a different icon on it.
What tends to be lost, on the way thither, is the notion that the end user and the admin ought to be two different virtual, so to speak, persons, even where they are one and the same person in the physical sense of the word.
Current proliferation of PC and non-PC hardware, platforms in general, is far from abating, contrary to what Intel might have wanted. Security issues won't go away either, and countless ways to lose money in a single click exist. I keep a ten-year worth of emails and a few G of other material I want to be preserved indefinitely, but hard drives fail, IDE interfaces disappear, and CDs get scratched and lost. Computing experience is being accumulated at a massive scale and rate, and yet computing (outside, in marketese, "check email and read news") is light years away from, say, end-user experience with cars.
Hence, I believe, computers must be managed by competent people, not "your grandmas", and it's just childish to insist happiness will prevail for all if computer maintenance is eventually made easy for the dull or otherwise uninterested.
From the end-user perspective, it's a totally different story. But here all has been well for quite a few years (since freetype, essentially). I manage my wife's oldish Latitude and IdeaPad, both under Gentoo, by taking the laptops offline for a night every three months, hand it back to her and forget until the next time around. The end-user is happy within her ~, and I make sure everything under / is operating normally, and that's the way it should be.
The Future of Linux is Google (PCWorld) Posted Nov 19, 2009 13:17 UTC (Thu) by NRArnot (subscriber, #3033) [Link]
Maybe it's just me, but I simply can't understand why people get so fussed about "the desktop"! Isn't it just a place to park files and start apps from?
My job involves Linux and Windows systems. On my desk is a monitor connected through a KVM to a Linux desktop and a Windows desktop, which I use fairly much interchangeably.
What I look for are apps that work as near as possible identically in both environments. My top five are Firefox (check), Thunderbird (check), NX Client(check), Gnome term or PuTTY session running ssh (check), OpenOffice (check). I think they'd also be the same on a Mac, if I ever have cause to use one.
The real Windows problem is application lock-in. I can't run Exchange on a Linux system, and Microsoft Exchange Server IMAP support sucks - almost certainly by design. OpenOffice can't yet correctly open every MS-Office created document, and were it not for the EU court, OO.org would have been placed at an ever-increasing disadvantage by Microsoft. Windows contains an unknown number of undocumented WINE-trap system services - ones that are there solely to hurt people running MS apps on Linux systems. Most desktop software vendors don't sell Linux versions, and many further trap you with undocumented proprietary file formats. Turkeys DO vote for Christmas, if you get your propaganda right.
This is why Windows will not be going away in a hurry. Not any deficiency of Gnome or KDE compared to Windows desktop. Not the (diminishing) number of hardware vendors who refuse to co-operate in the development of Linux drivers (perhaps there are now enough companies putting such hardware on a no-buy list? We go for Intel or ATI/AMD wherever possible).
A friend is running a small business with Linux systems only - no Windows unless there is a business-critical requirement for it with no alternative. So far, small-business accounting is the only exception, he'd love to get rid of the only Windows system but there is no sensible tax authority approved small business accounting system that runs on Linux. (Thanks, government, for forcing us to spend our money with a foreign multiply-convicted monopolist!) But there's no problem with staff learning to use Gnome and Openoffice on the job, they might prefer Windows/Office for the first few hours, but that's not what they are being paid to use and they rapidly adapt. It can be done.
|
|||||||||||||
