The premise of TFA is wrong. GNU/LInux does sell on the desktop and soon Android/Linux will as well. Many OEMs sell GNU/Linux and many retailers do as well, just not all of them. It's different in various parts of the world. In Germany you can go into shops with lots of shelf-space reserved for GNU/Linux and the share, according to NetApplications is 1.84%. In USA there are few shops that sell GNU/Linux and the share is reported to be 2.13%, not significantly different because it's business usage that NetApplications measure. Mountain View, California shows 80% because that's what Google uses.
see this German site and compare it with Dell, who have hundreds of stores in China selling GNU/Linux
http://www.notebooksbilliger.de/notebooks/notebooks+ohne+windows (That's Notebooks without Windows, according to Google Translate)
Most of us are not in such a situation. Where I work we depended on only two apps, IE and Office. We chucked them for OpenOffice.org and FireFox and Chrome Browser. Things are working well. We get greater speed and reliability without having to buy lots of new hardware.
There is so much FUD in this topic. M$ and "partners" try to upsell this technology to make sure they can tax it. If you run GNU/Linux terminal servers and simple X window system clients you get all the benefits of virtual desktops at much lower costs: cheaper servers (more processes per gigabyte and no licensing fees), cheaper thin clients (no need for gB of RAM or hard drive) and better performance (files are cached in RAM on the server or retrieved by a hot RAID). I use this technology a lot. I get 5s logins and 2s opening of windows to huge apps even using old PCs as thin clients. The usual VDI solution involves one virtual machine per client, a huge waste of resources although flexible. If you want low cost and reliability keep it simple and stick with GNU/Linux. It costs about $30 per client to have a good server on-line. New thin clients can be bought for less than $50 and used ones cost nothing (old XP machines are $0).
Don't listen to the FUD. Go all-in for thin clients and forget the VDI bloat. Use GNU/Linux.
One of the differences between working in a GNU/Linux shop and that other OS is stress. Last year, I worked in a shop using that other OS. It was always frightening on zero-day-malware-day because I always had to work late making sure updates were done in spite of having automatic updates enabled. With GNU/Linux, I type a few commands and it gets done for the whole system in a few minutes and I can go home to sleep. Last year I had trouble sleeping more than four hours. This year, I sleep as long as I want knowing things are safe. Next year, we will be 95% M$-free. I look forward to that.
So I can "invent" the shovel today? By moving manure, dirt, sand, rice, potatoes,....
I hope the USPTO has an infinite staff... Wait a minute! They CHARGE for applications! That's their new business plan: generate an infinite number of patent applications and pay off the national debt with the fees. I understand.
If Apple has to spend a bundle on marketing to get the meagre share they have, why has GNU/Linux twice the share? Could it be that its not about marketing only but price also matters?
Parent leans on facts not in evidence. 100 million users of GNU/Linux on the desktop cannot be GNU/Linux freaks/geeks/sociopaths. GNU/Linux works well. If someone with too much cash wants to spend money on CPU power to round the corners on rectangular areas of the screen, so be it. Others want to get on with what they are doing with less regard to M$ and pals bottom lines.
I use GNU/Linux because it works and that other OS does not. That other OS phones home, sniffs files for DRM, BSODs, has a very EXCLUSIVE EULA, invites malware in, needs re-re-reboots, and messes up memory, storage and everything else it touches. I do not give a damn about how beautiful the UI is. I use PCs to get things done.
One of the richest companies in the world produces a crappy user interface. So much for the thesis that GNU/Linux must have professional UI designers.
Take, for example, GIMP. Many say it has a lousy user-interface. I can give GIMP to folks who have never used PS and they have no trouble at all making the fish larger, eliminating fly-away hair, whitening teeth and eliminating ex-boyfriends. In what way is the UI not good? Only in that GIMP is not identical to PS, apparently. Folks who take the trouble to learn how to use GIMP have no problems with its user-interface.
That does not look like a fair comparison, especially with bombs that can fly in a window from miles away, killer drones flying day and night and taxpayers tolerating thousands of casualties and billion dollar budgets for occupation.
NetApps stats for GNU/Linux share are about 20% of W3Schools. W3Schools clearly has a bias for that other OS because large parts of the site are M$-only stuff like.asp/.NET, so the numbers for GNU/Linux should be much higher. One thing that is missed for sure in those stats is global coverage. GNU/Linux is hot in Asia. What proportion of the hits on NetApps and W3Schools stats are from Asia and other regions where GNU/Linux is hot? We do not know, so take those numbers with a grain of salt.
IDC sells reports with comments like the following for thousands of dollars:
"Despite the dominance of the Windows platform, Linux adoption continues to grow in the region in both the COE and server operating environment (SOE) spaces," says Antony Lee, market analyst, Software Research, IDC Asia/Pacific.
...
"On the desktop side, IDC sees Linux share more than doubling, from 3% today to 6% in 2007, while Windows loses a bit of ground."
So, people who scientifically design and implement surveys reported that GNU/Linux was the size of Mac on the desktop a few years ago and it is still growing rapidly.
see this excerpt. That was from 2005. If the share was 3% then and growing rapidly, how can the NetApps share of less than 1% possibly be true unless NetApps' universe is unrepresentative?
That was before Dell and ASUS jumped in.
So. There are no signs of GNU/Linux on the desktop slowing down any time soon and Chinese Linux Market
We know there are millions of GNU/Linux desktops there, because Sun made a deal to supply millions of them.
see Sun story (2003)
Turbolinux is also in China in a big way.
"According to the International Data Corp. (IDC), Turbolinux's market share in servers in China was 62 percent in 2004. On the desktop, it holds a 25 percent share. "
GNU/Linux is huge in China, a country several times the population of the USA with a huge growth in GDP. Hundreds of millions there will be first time computer buyers within a few years and they are not locked-in to M$
Numbers are not too much different in the BRIC (Brazil Russia India and China). There, governments are activly promoting GNU/Linux by using it themselves, putting it into schools or insisting on open file formats.
"Sun executives were meeting with Brazilian government executives and were told in no uncertain terms that the government would not consider any technology that wasn't open source. "
see www.internetnews.com/bus-news/article.php/3697166
So Brazil was the straw that broke the camel's back and caused Sun to open Java.
BRIC is 2.65 thousand million people. see http://www.xminc.com/mt/archives/000177.html
Many are poor but rapidly industrializing and hungry for IT. Are they going to want a bloated OS or a lean, mean, computing machine? Do not be misled by NetApps. Unless their clients are audited and deemed to be representative of the world somehow, they must be considered way off base.
China is huge. If you look at http://google.com/trends and enter linux,windows you will see that other OS has a steady lead over time with Google. Now, zero in on China. Interesting, eh? Now, zero in on Beijing. Whoops! Where did the lead go? Beijing is a huge city and the seat of government. Stories about that other OS taking over there are overstated, even at $3 a licence.
People! Do the maths. You get most of the functionality of a $400 PC for $200. That is an instant return on investment ot 200%. Not bad. The rating in TFA should have been 4 - 4.5 stars.
Besides value for money, the product has important value to society: many people who cannot afford a more expensive PC will be able to buy this product. Of course, it might be better to recycle an older PC for such purposes, but that would take more tech-savvy. Look at it this way. If schools used this machine, they could have machines in the lab, library, home-ec/industrial arts and classrooms for the same price as machines in the lab only using the more expensive machines. That really opens opportunities to use IT in a school and get more done faster. This machine deserves more respect just for the price. Running GNU/Linux is a bonus.
I predict that 2008 will see a large increase in production and sale of low-end boxes and laptops to satisfy a need in emerging IT markets, BRIC (Brazil, Russia, India and China), for instance. If OEMs can ramp up production, this technology could rapidly displace that other OS, globally. Vista, with its obscene bloat and MIPS-eating ways has no future in that market.
I use Linux terminal servers which magnifies the above effect. Imagine a room full of XP machines that have just been re-imaged to clean out the cruft. Useless. It would take days to re-install all the stuff I need to teach my courses. I plugged in my Linux terminal server, adjusted the BIOS of each client to boot PXE and voila! A working lab! If I installed the server from scratch it would take a while but I get a bunch of computer seats for the effort. There is no way M$ makes sense competitively against the numbers one can get with Linux. As the price/performance ratio for hardware keeps getting better, GNU/Linux keeps looking better for me because a single server can handle more clients. Last year, I built a system with six servers to run thin clients all over a school. This year, with the quad-core Opterons coming on-stream, a medium sized school could run from a single terminal server, perhaps 200 thin clients and one server costing $4000. If you have the package lists and a local repository on a gigabit/s link, you could install the software for such a system in an hour. Imagine 200 installations of XP! Even re-imaging 10 gB images all over the building would take way more time and trouble than one installation of Linux. Tell me about cleaning malware off 200 machines! M$ is malware.
The problem isn't that linux is intimidating for the average/. poster. It is that Linux is pretty freaking intimidating for the average computer user.
Nonsense! I teach grade 1 to 10 students to use PCs running Debian GNU/Linux and the only real problem they have had was logging in and pointing and clicking a mouse. I use scripts to log the grade 1s and 2s in automatically on bootup as they are the first classes of the day. The others just do it. If these unsophisticated newbies can handle it, almost anyone can. Sure, there are exceptional technophobes, but they are certainly not your typical user. Teachers are your typical user and they wander into my lab looking for a PC and they are off. Few even realize they are using GNU/Linux. People change their desktop backgrounds all the time, right? There are icons and folders and menus all over. Where is the problem? No one needs to RTFM to use a GUI. M$ and Apple copied the GUI from Xerox Parc, so it is a pretty standard interface these days.
Re:Fair??? Language, please...
on
SCO Loses
·
· Score: 1
LWATCDR wrote:Heck he might have actually believed that Linux was ripping off SCO's IP.
Judge Kimball:
On January 4, 2003, McBride received an email from Michael Anderer, a consultant for SCO retained to examine its intellectual property. Supp. Brakebill Decl. Ex. 12. Anderer stated that the APA "transferred substantially less" of Novell's intellectual property than Novell owned. Anderer noted that Santa Cruz's "asset purchase" from Novell "excludes all patents, copyrights, and just about everything else." Id. Anderer cautioned that "[w]e really need to be clear on what we can license. It may be a lot less than we think."
On February 4, 2003, McBride contacted Christopher Stone, Vice Chairman of Novell, and stated that he wanted Novell to "amend" the APA to give SCO "the copyrights to UNIX." Supp. Brakebill Decl. Ex. 17; id. Ex. 18 ("Stone Dep." at 108-09). Then, on February 25, 2003, McBride twice called a Novell employee in business development, David Wright, and said, "SCO needs the copyrights." Wright passed on McBride's request to Novell's in-house legal department. Supp. Brakebill Decl. Ex. 13. McBride's request was memorialized in an email written that day by a Novell in-house attorney, Greg Jones. Id.
So, SCOG knew when they started the litigation that they did not have the copyrights. They also had no evidence of literal copying even after getting truck loads of source code from IBM in discovery. They could present no evidence of copying after that discovery so they must not have had any evidence before they started litigating.
Another thing one must address is that the heat flow can only be used where permeable strata exists in the ground making it possible to circulate water to extract the heat. In places with crystalline bedrock, the heat flow can not be used.
In cold regions, near the ocean, like Canada's arctic, the sea is much warmer than the ambient temperatures which go as low as -50C. A heat pump from the ocean to buildings is quite a feasible way of exploiting geo-thermal power. One lays a pipeline instead of drilling downward. Even if they just surrounded the buildings with a blanket at sea temperatures, they would cut heating costs greatly. This would actually help global warming a bit by cooling the sea slightly.
thegnu wrote:
There is NO reason for the average home user to install a completely new OS they've never seen. The hurdle for Linux is to get on enough work PCs that people are relatively comfortable enough with it, so that next virus they get, or next Norton Death Knell, they leap off their burning Windows install onto something stable.
For the 80% of "easy" cases where browsing/e-mail/word-processing are the important functions, there are several reasons to migrate:
they can run 2007 software on machines that run stuff released in 2001 or earlier and cannot run Vista
they can pay what it costs to install the software instead of what the monopoly in the desktop market demands
often, installing Linux is easier than installing/delousing that other OS
even if they have never seen/heard "Linux" they can learn about it while researching problems with malware, the IT industry, headlines about anti-competion cases against M$, even the 10-Q or Analyst Meetings of M$ or just watching TV. Linux has been in the news one way or another heavily since about 2000. In 2000 when I did that, I read that Linux was hard to install but Caldera's installer was one of the easiest to use.
before 2004 I had only met one person who was aware of Linux. Recently a much larger rate of familiarity exists even in persons who have never seen Linux. e.g. a high percentage of businesses use Linux on servers at work and a smaller fraction have examined FLOSS on the desktop and even Linux on the desktop see IDC survey presented at LinuxWolrd Summit 2006
Linux has been ready for the desktop for years. I was using that other OS 8 years ago and had no end of trouble with it freezing, crashing and losing our data. I switched to Linux on the desktop and it has been relatively trouble-free ever since. I have introduced hundreds of students and teachers to Linux and very few had any trouble as newbies. They liked the fact that for no cost I could provide them systems with greatly improved performance. The idea that Linux is not ready must stem from propaganda or low market share. Reality is much different. Free as in free beer installations are not counted in market shares. Web statistics are very unreliable. Surveys of OEMs and business show rapid adoption of Linux on the desktop. That is why Dell, HP, and many other firms are providing PCs with Linux pre-installed. That is why the global market for Linux servers, service and applications amounts to billions of dollars and is growing rapidly.
see IDC report 2007
Worldwide revenue from standalone open source software reached $1.8 billion in 2006.
This revenue will reach $5.8 billion in 2011, representing a compound annual growth rate (CAGR) of 26% from 2006 to 2011. see IDC report on Linux
in China, 2007
On the other hand, 2006 was a good year for the Linux desktop. The Ministry of Information Industry, State Copyright Bureau and Ministry of Commerce first issued a joint decree that required all new PCs to be installed with a legal-version OS. This was followed by a directive that forbade the installation of FreeDos in new PCs. Given such a regulatory setting, the price advantage of a Linux desktop became more attractive and a number of PC vendors, who previously did not install any OS, quickly adopted Linux desktop products. This led to a sharp increase in OEM revenue for Linux desktop. At the same time, Linux vendors launched and heavily promoted new desktop products, which contributed to the robust development of the Linux desktop market that year. Bolstered by favorable government policies, Linux desktops shipments grew apace, rapidly reaching new users via OEMs. The value of the Linux OS also became more widely recognized, offering greater opportunity for active development and deployment of Linux desktop products, said Vivian.
So, the reality is that Linux on the desktop is growing at 20% per annum in the commercial market which lags the personal/free market by a large margin. M$ had to cut its price to $3 just to stay competitive. That is all Linux needs, to be allowed to compete on price and performance. For years, M$ has had a free ride. That is soon stopping. Get used to it. It is doubtful that Linux will KO M$ because some will always want to pay too much or be swayed by sales campaigns , but M$ will fall into the pack with realistic prices and market shares. Remember the glory days of the Soviet Union, when every election resulted in the landslide for a single candidate of the party's choosing? Those days are gone forever in Russia and they will soon be gone for M$.
There are lots of PCs out there. I have seen a few '95 and '98 machines still in use (Horrors!).
The aspect that eludes M$ is that of the existing PCs, very few can run Vista. M$ plans to kill off XP next year for new licences, and upgrades of XP soon. Will the world trash hundreds of millions of working PCs, PIII and later? Why would they? The world will find Linux ready to run on them with modern software. It takes a salesman to announce a problem with the M$ empire is an advantage.
Whether it likes it or not, the world will not trash that many working PCs and M$ will have to supply a product for them or drop out. This is not like the good old days when M$ had to persuade folks to shift from 8MHz CPUs to 400MHz CPUs where they could see a real benefit. Nowadays, 3000 MHz CPUs are idling and they want folks to run dual core models that can do 200 frames per second in high resolution to read text.
OM was intended to be something on a private netowrk like a lab, so, if web access is needed, a sturdy firewall is in order. The auto-discovery stuff just is stupid on the WWW, but makes sense in a local cluster. Performance would be sad over the WWW, anyway, because of latency.
Architecture
The 2.6 kernel version of OM was supposed to be very rational with most of the fancy stuff in userland. I hope this gets working this summer as planned. OM is useful in 2.4 but who has distros for 2.4 coming out these days?
OM may be of limited interest to most except number crunchers/searchers but folks who use labs love it because you can set up a cluster in minutes for a project or run a whole lab as a cluster to serve something to the rest of the building or the world. Many labs are populated with 3 gHz processors (90 gHz total) and 512MB RAM (16 gB RAM in total) so they can really do something using OM. The load balancing and auto-discovery are lovely features. Migration of shared memory was available in 2.4 so it could even be used with Apache or OpenOffice! How versatile can you get?
Amen, Brother!
Most of us stay home or drive a car to work. At work, does a laptop make sense? I don't think so. The keyboards are flat. We live in a 3D world and a big, klunky Fujitsu keyboard works for me. If I need to carry work home, I can use a USB drive. When I go to the garden or the beach, I want to get away from computing for a break. Why spoil it with a portable thingy prone to droppage, breakage, leakage, theft, fire and explosion?
If the space-saving feature of laptops is desired, look into thin client technology. The boxes are the size of chocolate/cigar boxes and can sometimes be bolted to the back of an LCD monitor or included inside the monitor. A thin client is way cheaper than a laptop, too. I do not see the value of a laptop being twice what a desktop device is. Why can I get a good desktop for a few hundred dollars but a laptop is starting at $600? If we were all salesmen or writing novels on the bus, I could see a laptop, but most of us are moving from one flat spot to another. We can afford a PC at both places for the cost of a laptop.
Here is an image of a thin client setup. Monitor and keyboard are way larger than a laptop but it takes up very little space/material/money/maintenance compared to a laptop. Combined with a network and a Linux terminal server, setups like this enable a whole department to be maintained by keeping one server going so we can compute instead of labour.
The one situation where I personally would use a laptop is if I were moving around and giving presentations and needed an application like OpenOffice running where I went. I am looking at one of those Dell laptops with Linux and in bright yellow... They do not sell them in Canada yet.
With 8-core computers on the market and 16-core computers coming soon, there's just not as much need for openMosix.
For small clusters, bunratty is correct, but there can be huge OpenMosix clusters with hundreds or thousands of machines. SMP will not work for that.
I noticed last year that 2.6 development in OpenMosix was dragging but was assured that they had the manpower. I do not accept that SMP/multi-core is the cause of this stagnation. It sounds like the leader(s) of the group lost interest.
OpenMosix was a great idea. I wanted to use if for LTSP so I would not need to obtain powerful servers. Others use it for number-crunching or rendering. I cannot live with 2.4 because most distros have dropped 2.4 . Let us hope that 2.6 OpenMosix is in a usable state by 2008. It is in alpha still.
Slashdot Mirror
The premise of TFA is wrong. GNU/LInux does sell on the desktop and soon Android/Linux will as well. Many OEMs sell GNU/Linux and many retailers do as well, just not all of them. It's different in various parts of the world. In Germany you can go into shops with lots of shelf-space reserved for GNU/Linux and the share, according to NetApplications is 1.84%. In USA there are few shops that sell GNU/Linux and the share is reported to be 2.13%, not significantly different because it's business usage that NetApplications measure. Mountain View, California shows 80% because that's what Google uses. see this German site and compare it with Dell, who have hundreds of stores in China selling GNU/Linux http://www.notebooksbilliger.de/notebooks/notebooks+ohne+windows (That's Notebooks without Windows, according to Google Translate)
Change apps or run that other OS in a virtual machine or with WINE. WINE is much cheaper because you don't need a licence fee for that other OS.
Most of us are not in such a situation. Where I work we depended on only two apps, IE and Office. We chucked them for OpenOffice.org and FireFox and Chrome Browser. Things are working well. We get greater speed and reliability without having to buy lots of new hardware.
There is so much FUD in this topic. M$ and "partners" try to upsell this technology to make sure they can tax it. If you run GNU/Linux terminal servers and simple X window system clients you get all the benefits of virtual desktops at much lower costs: cheaper servers (more processes per gigabyte and no licensing fees), cheaper thin clients (no need for gB of RAM or hard drive) and better performance (files are cached in RAM on the server or retrieved by a hot RAID). I use this technology a lot. I get 5s logins and 2s opening of windows to huge apps even using old PCs as thin clients. The usual VDI solution involves one virtual machine per client, a huge waste of resources although flexible. If you want low cost and reliability keep it simple and stick with GNU/Linux. It costs about $30 per client to have a good server on-line. New thin clients can be bought for less than $50 and used ones cost nothing (old XP machines are $0). Don't listen to the FUD. Go all-in for thin clients and forget the VDI bloat. Use GNU/Linux.
Amen
One of the differences between working in a GNU/Linux shop and that other OS is stress. Last year, I worked in a shop using that other OS. It was always frightening on zero-day-malware-day because I always had to work late making sure updates were done in spite of having automatic updates enabled. With GNU/Linux, I type a few commands and it gets done for the whole system in a few minutes and I can go home to sleep. Last year I had trouble sleeping more than four hours. This year, I sleep as long as I want knowing things are safe. Next year, we will be 95% M$-free. I look forward to that.
So I can "invent" the shovel today? By moving manure, dirt, sand, rice, potatoes,.... I hope the USPTO has an infinite staff... Wait a minute! They CHARGE for applications! That's their new business plan: generate an infinite number of patent applications and pay off the national debt with the fees. I understand.
If Apple has to spend a bundle on marketing to get the meagre share they have, why has GNU/Linux twice the share? Could it be that its not about marketing only but price also matters?
Parent leans on facts not in evidence. 100 million users of GNU/Linux on the desktop cannot be GNU/Linux freaks/geeks/sociopaths. GNU/Linux works well. If someone with too much cash wants to spend money on CPU power to round the corners on rectangular areas of the screen, so be it. Others want to get on with what they are doing with less regard to M$ and pals bottom lines. I use GNU/Linux because it works and that other OS does not. That other OS phones home, sniffs files for DRM, BSODs, has a very EXCLUSIVE EULA, invites malware in, needs re-re-reboots, and messes up memory, storage and everything else it touches. I do not give a damn about how beautiful the UI is. I use PCs to get things done.
One of the richest companies in the world produces a crappy user interface. So much for the thesis that GNU/Linux must have professional UI designers. Take, for example, GIMP. Many say it has a lousy user-interface. I can give GIMP to folks who have never used PS and they have no trouble at all making the fish larger, eliminating fly-away hair, whitening teeth and eliminating ex-boyfriends. In what way is the UI not good? Only in that GIMP is not identical to PS, apparently. Folks who take the trouble to learn how to use GIMP have no problems with its user-interface.
"Stable" has two meanings: 1)features set 2)failures are rare
Iraq: suicide bombers, IEDs, AK47s and RPGs
DC: handguns...
That does not look like a fair comparison, especially with bombs that can fly in a window from miles away, killer drones flying day and night and taxpayers tolerating thousands of casualties and billion dollar budgets for occupation.
IDC sells reports with comments like the following for thousands of dollars:
"Despite the dominance of the Windows platform, Linux adoption continues to grow in the region in both the COE and server operating environment (SOE) spaces," says Antony Lee, market analyst, Software Research, IDC Asia/Pacific.
"On the desktop side, IDC sees Linux share more than doubling, from 3% today to 6% in 2007, while Windows loses a bit of ground."
So, people who scientifically design and implement surveys reported that GNU/Linux was the size of Mac on the desktop a few years ago and it is still growing rapidly.
see this excerpt. That was from 2005. If the share was 3% then and growing rapidly, how can the NetApps share of less than 1% possibly be true unless NetApps' universe is unrepresentative? That was before Dell and ASUS jumped in.
So. There are no signs of GNU/Linux on the desktop slowing down any time soon and Chinese Linux Market
We know there are millions of GNU/Linux desktops there, because Sun made a deal to supply millions of them. see Sun story (2003)
Turbolinux is also in China in a big way. "According to the International Data Corp. (IDC), Turbolinux's market share in servers in China was 62 percent in 2004. On the desktop, it holds a 25 percent share. "
see http://www.informationweek.com/story/showArticle.jhtml?articleID=170700943 (2005)
GNU/Linux is huge in China, a country several times the population of the USA with a huge growth in GDP. Hundreds of millions there will be first time computer buyers within a few years and they are not locked-in to M$
Numbers are not too much different in the BRIC (Brazil Russia India and China). There, governments are activly promoting GNU/Linux by using it themselves, putting it into schools or insisting on open file formats.
"Sun executives were meeting with Brazilian government executives and were told in no uncertain terms that the government would not consider any technology that wasn't open source. " see www.internetnews.com/bus-news/article.php/3697166
So Brazil was the straw that broke the camel's back and caused Sun to open Java.
BRIC is 2.65 thousand million people. see http://www.xminc.com/mt/archives/000177.html Many are poor but rapidly industrializing and hungry for IT. Are they going to want a bloated OS or a lean, mean, computing machine? Do not be misled by NetApps. Unless their clients are audited and deemed to be representative of the world somehow, they must be considered way off base.
China is huge. If you look at http://google.com/trends and enter linux,windows you will see that other OS has a steady lead over time with Google. Now, zero in on China. Interesting, eh? Now, zero in on Beijing. Whoops! Where did the lead go? Beijing is a huge city and the seat of government. Stories about that other OS taking over there are overstated, even at $3 a licence.
Besides value for money, the product has important value to society: many people who cannot afford a more expensive PC will be able to buy this product. Of course, it might be better to recycle an older PC for such purposes, but that would take more tech-savvy. Look at it this way. If schools used this machine, they could have machines in the lab, library, home-ec/industrial arts and classrooms for the same price as machines in the lab only using the more expensive machines. That really opens opportunities to use IT in a school and get more done faster. This machine deserves more respect just for the price. Running GNU/Linux is a bonus.
I predict that 2008 will see a large increase in production and sale of low-end boxes and laptops to satisfy a need in emerging IT markets, BRIC (Brazil, Russia, India and China), for instance. If OEMs can ramp up production, this technology could rapidly displace that other OS, globally. Vista, with its obscene bloat and MIPS-eating ways has no future in that market.
I use Linux terminal servers which magnifies the above effect. Imagine a room full of XP machines that have just been re-imaged to clean out the cruft. Useless. It would take days to re-install all the stuff I need to teach my courses. I plugged in my Linux terminal server, adjusted the BIOS of each client to boot PXE and voila! A working lab! If I installed the server from scratch it would take a while but I get a bunch of computer seats for the effort. There is no way M$ makes sense competitively against the numbers one can get with Linux. As the price/performance ratio for hardware keeps getting better, GNU/Linux keeps looking better for me because a single server can handle more clients. Last year, I built a system with six servers to run thin clients all over a school. This year, with the quad-core Opterons coming on-stream, a medium sized school could run from a single terminal server, perhaps 200 thin clients and one server costing $4000. If you have the package lists and a local repository on a gigabit/s link, you could install the software for such a system in an hour. Imagine 200 installations of XP! Even re-imaging 10 gB images all over the building would take way more time and trouble than one installation of Linux. Tell me about cleaning malware off 200 machines! M$ is malware.
Judge Kimball:
So, SCOG knew when they started the litigation that they did not have the copyrights. They also had no evidence of literal copying even after getting truck loads of source code from IBM in discovery. They could present no evidence of copying after that discovery so they must not have had any evidence before they started litigating.In cold regions, near the ocean, like Canada's arctic, the sea is much warmer than the ambient temperatures which go as low as -50C. A heat pump from the ocean to buildings is quite a feasible way of exploiting geo-thermal power. One lays a pipeline instead of drilling downward. Even if they just surrounded the buildings with a blanket at sea temperatures, they would cut heating costs greatly. This would actually help global warming a bit by cooling the sea slightly.
For the 80% of "easy" cases where browsing/e-mail/word-processing are the important functions, there are several reasons to migrate:
So, the reality is that Linux on the desktop is growing at 20% per annum in the commercial market which lags the personal/free market by a large margin. M$ had to cut its price to $3 just to stay competitive. That is all Linux needs, to be allowed to compete on price and performance. For years, M$ has had a free ride. That is soon stopping. Get used to it. It is doubtful that Linux will KO M$ because some will always want to pay too much or be swayed by sales campaigns , but M$ will fall into the pack with realistic prices and market shares. Remember the glory days of the Soviet Union, when every election resulted in the landslide for a single candidate of the party's choosing? Those days are gone forever in Russia and they will soon be gone for M$.
There are lots of PCs out there. I have seen a few '95 and '98 machines still in use (Horrors!).
The aspect that eludes M$ is that of the existing PCs, very few can run Vista. M$ plans to kill off XP next year for new licences, and upgrades of XP soon. Will the world trash hundreds of millions of working PCs, PIII and later? Why would they? The world will find Linux ready to run on them with modern software. It takes a salesman to announce a problem with the M$ empire is an advantage.
Whether it likes it or not, the world will not trash that many working PCs and M$ will have to supply a product for them or drop out. This is not like the good old days when M$ had to persuade folks to shift from 8MHz CPUs to 400MHz CPUs where they could see a real benefit. Nowadays, 3000 MHz CPUs are idling and they want folks to run dual core models that can do 200 frames per second in high resolution to read text.
OM was intended to be something on a private netowrk like a lab, so, if web access is needed, a sturdy firewall is in order. The auto-discovery stuff just is stupid on the WWW, but makes sense in a local cluster. Performance would be sad over the WWW, anyway, because of latency.
The 2.6 kernel version of OM was supposed to be very rational with most of the fancy stuff in userland. I hope this gets working this summer as planned. OM is useful in 2.4 but who has distros for 2.4 coming out these days?
OM may be of limited interest to most except number crunchers/searchers but folks who use labs love it because you can set up a cluster in minutes for a project or run a whole lab as a cluster to serve something to the rest of the building or the world. Many labs are populated with 3 gHz processors (90 gHz total) and 512MB RAM (16 gB RAM in total) so they can really do something using OM. The load balancing and auto-discovery are lovely features. Migration of shared memory was available in 2.4 so it could even be used with Apache or OpenOffice! How versatile can you get?
If the space-saving feature of laptops is desired, look into thin client technology. The boxes are the size of chocolate/cigar boxes and can sometimes be bolted to the back of an LCD monitor or included inside the monitor. A thin client is way cheaper than a laptop, too. I do not see the value of a laptop being twice what a desktop device is. Why can I get a good desktop for a few hundred dollars but a laptop is starting at $600? If we were all salesmen or writing novels on the bus, I could see a laptop, but most of us are moving from one flat spot to another. We can afford a PC at both places for the cost of a laptop.
Here is an image of a thin client setup. Monitor and keyboard are way larger than a laptop but it takes up very little space/material/money/maintenance compared to a laptop. Combined with a network and a Linux terminal server, setups like this enable a whole department to be maintained by keeping one server going so we can compute instead of labour.
The one situation where I personally would use a laptop is if I were moving around and giving presentations and needed an application like OpenOffice running where I went. I am looking at one of those Dell laptops with Linux and in bright yellow... They do not sell them in Canada yet.
For small clusters, bunratty is correct, but there can be huge OpenMosix clusters with hundreds or thousands of machines. SMP will not work for that.
I noticed last year that 2.6 development in OpenMosix was dragging but was assured that they had the manpower. I do not accept that SMP/multi-core is the cause of this stagnation. It sounds like the leader(s) of the group lost interest.
OpenMosix was a great idea. I wanted to use if for LTSP so I would not need to obtain powerful servers. Others use it for number-crunching or rendering. I cannot live with 2.4 because most distros have dropped 2.4 . Let us hope that 2.6 OpenMosix is in a usable state by 2008. It is in alpha still.