Opinion Column by PC Magazine: How to Kill Linux
Personally I think this is an interesting column I just finished for PC magazine although the readers seem nonplussed. And the Linux folks think I’m both stupid and crazy (pick one and stick to it, please!) Sometimes you never know.
John, you’re entire theory is that people will suddenly drop Windows for a Microsoft supported Linux is simply ludicrous. No one is going to dump Microsoft’s draconian business practices only to get MORE of the same.
erm… some of windows’s lack of stability and problems come from the drivers architucture, and some of teh security problems come from the fact that drivers can run with elevated prrivileges in the OS.
So your solution would cause a loss of security and stability, that’s why I don’t see it ever hitting the market.
As usual, interesting analysis, John. Certainly, possible that your projection could/would exist as a 3rd/4th force in the range of OS options available to us. Something a changing marketplace can always use.
There have been 3 of us in my immediate family experimenting with alternatives to Windows. I concluded that the question of drivers was going to keep me from any timely commitment to Linux.
What to do next? Well, the introduction of the Mac Mini solved that question. An affordable experiment has become a solution for me. I’ve already moved all my personal stuff over to OS X. Some of the software I’ve been using turns out to have OS X versions on the original install CD [Photoshop Elements, for example]. The 3 new pieces I’ve added — were chosen because they were less expensive than the OS X versions offered by the software firms who developed the pieces I’d used on my WXP machines. A pleasant surprise — contradicting my preconception about Mac software being more expensive than M$oft equivalents.
Looks to be an interesting couple of years coming up – if the software solutions to personal/business computing continue to spiral out as dramatically as do the hardware solutions.
For me — I’m just waiting to see what Apple comes up with for an HD Home Media Center. Oh, I’ll be giving my WXP machine away to some unsuspecting young member of my extended family. As soon as the OpenOffice crew get their OS X version sorted and bug-free. Then, I’ll just add a second standalone HD jukebox to my Mini — for the business side of my home/office.
do you know about this website?? http://www.mslinux.org/
your assumptions that the linux community will just die are ridiculous.
Someone posted the article on http://www.osnews.com
The forum went ape $h!%. They had to moderate half the comments. But when you filter all the attacks on John, there are quite a few comments that make sense.
Mr. Dvorak,
When, if ever have you used GNU/Linux?
Are you aware that most recent GNU/Linux distros have far better hardware support than ANY MS OS?
I run 4 boxes. 1 SuSE 9.2, 1 Fedora Core 3, 1 Mandrake 10.1, and 1 Win XP. All of my GNU Linux boxes properly installed and configured ALL of the hardware on their respective boxes including and old TV card and even my Firewire DV camcorder. If I install a new peripheral it is automatically recognized and a screen comes up which allows me to configure the device if it’s not configured automagically as is often the case. Need an app to use with that new device? Start APT, select the app, it is then downloaded along with all of the necesary dependencies and installed with two mouse clicks.
MS has nothing to offer GNU/Linux. Why would anyone waste any time with a Windows driver base which is already inferior to that of GNU/Linux? Are you not aware that most hardware vendors are now committed to supporting Linux? Including Intel, AMD, hp, Dell, IBM, Nvidia, ATI, etc….
Are you not aware that most major software and Harware comapanies are now contributing members of the OSDL? In fact virtually every company that’s not Microsoft is actively working to promote GNU/Linux. Why is this Mr. Dvorak?
It would serve you well to learn something, anything, about GNU/Linux before you write about it. It would be far less embarassing for you.
The bottom line is that lower price will take down any giant. Microsoft may bite off the head of Linux, but they’ll never kill its soul.
The secret project bit I found amusing. It’s neither secret no cancelled. See http://www.colinux.org. In a sense it is irrelevant as next generation x86 processor roadmaps all include hardware virtualisation support. PC owners will finally get the mainframe toys.
If you ask mainframe folks they not only run multiple instances but they run multiple OS’s and often multiple versions and upgrade by moving services one at a time carefully between the old and new release of the OS while running both.
Virtualisation breaks the “OShardware’ tie. An OS becomes a thing that runs applications not a hardware manager. Already on the mainframe Linux is middleware rather than the base product.
Mr Dvorak,
For most users, the decision to use GNU/Linux is based on the
freedoms offered by the GPL. Obtaining Linux-friendly hardware has not been a problem for me in the last 8 years.
Anyway, seems to me Microsoft is spending most of it’s time attacking Linux by attempting to push so called ‘standards’ through organizations such as W3C, IETF and now OASIS.
If Microsoft’s Martin Taylor is an honest man, Microsoft has no interest in implementing anything Linux related.
Believe it or not, I actually predicted something very similar to that a couple of years ago on the Extremetech forums. I’m trying to find my original post (the search there really sucks), but it was along the same vein.
The point that most people miss is that MS doesn’t need to actually kill Linux. If everybody (meaning the average consumer) thinks that MS’s brand of linux is better, then the real linux will simply stay small and not be a growing threat. They don’t need to kill it, they just need to stop the growth before it gets too late.
Interesting article and thought provoking. From the comments, it looks like you’ve touched a nerve among some of my fellow Linux users. I’ve had mixed success in the past with drivers for both Windows and Linux. Actually the drivers for Windows have only improved to their current state since XP.
The big question, that you yourself asked, is could MS pull it off? I’m thinking no, they don’t have the skill to create the code, that would actually work…at least not within a 5 to 7 year period (if Longhorn is any indication).
Again, interesting idea though and a good discussion topic.
John, such a product already exists:
VMware Workstation 4.5
That column would have made sense 5 years ago. These days almost all hardware sellers provide Linux drivers for their products. Although people with old computers sometimes can’t use their machines with Linux, those people are not the ones who would pay for new software either.
MS can hurt Linux if they come up with some kind of “software layer” that would allow user to run Windows software in Linux and charged a hefty fee for that. Let’s face it, although Linux is slowing gaining market share, there are still a lack for good software for Linux. Most users want to run Photoshop, MS Office, games, etc.
People buy new computers to run the latest game hit, not to run their spreadsheets faster.
Hey John I think you should have targeted the application layer. Imagine a MS-Linux offering exclusively the APIs from Windows, with tunned versions of Microsoft Office et al with a nice (and familiar) Windows Shell.
I know there are a few projects which (attempt to) run Windows native applications…!?
On the user’s side: they would remain using their current (May I say favourite) applications (Even those mammoth DX 9 games).
The downside would be lots of (angry?) programmers out of job since the porting of applications would be reduced to almost nil
And forget the additional training/testing/debugging/whatever Linux applications need.
Application developers will reduce their cost because their application (Windows version) will run on Linux too and those who didn’t have a Linux version will have the added benefit of automatically getting a “cross-platform” product.
Of course some folks would say ‘why bother’ in shifting from Windows to MS-Linux while others will say Windows is dead/will die then…
Whatever they say, we will have the nice big monopoly back.
“almost all hardware sellers provide Linux drivers for their products”
Uh, no. That’s why the 1st answer from most Linux pros to questions about wi-fi is “works fine as long as you’re not using Broadcom silicon”. Sadly, that’s a lot of brands.
The original routers from Linksys — which had openings to Linux — the openings were closed, last year.
[ RANT & RAMBLE WARNING ]
Where do you all come up with this stuff? Grid computing is the future, static memory modules (or their equivalent) are the future, robotics is the future, advanced AI is the future, etc, etc, etc… Who cares how we get there. All I know is that I just want to get there as soon as possible. Microsoft spends about a billion dollars a year on research. I’m sure IBM, SUN, SONY, etc. all spend large sums of money too. Who cares who wins over technology based in on the past. These companies put forth unimaginable resources in an attempt to solve some of the more interesting problems in computer science (and make an extra buck). I do not believe that the vast resources of the open source community are able to be sharply focused enough on developing these new technologies (though I wish). It appears that they spend most of their time re implementing open versions of things that already exist. Hell, the major draw of Linux in the beginning was that it was a free (meaning both free and FREE) version of a POSIX compliant OS. UNIX had existed for a long while before Linux came along. That being said, Linux and my favorite distribution (Gentoo) is great fun to tweak and play with. When you think of advanced OSs, think of large distributed systems solving real problems. Clustering boxes is a great start, but virtual platform independent computing environments is my prediction (not solely, I just buy into the concept). This type of grid computing allows for massive generational genetic algorithms to solve/closely approximate the answers to interesting worthwhile questions. These are the things that our computers should be taking part in while I take a break from my word document to go ‘drain the tank’ (think seti and http://www.grid.org, but bigger).
Windows NT/ XP stem from research started 20 years ago with IBM and Linux is based on POSIX which is an ancient standard (Not only that, but it is a monolithic architecture). Who freak’n cares about either anymore? Anyway, as far as the current OS go, it’s not like it is rocket science anymore. The problems that current OSs solve have become uninteresting and are hardly cutting edge.
The future needs open standards. And not one company’s bastardized version of a standard, but a true open standard where we as computing professionals can give input into its creation. We need independent bodies of review, to be critical of the software and hardware produced. We need better security and the removal of software patents. The whole idea of a software patent is ludicrous. Seriously, it’s like putting a patent on a book intended to teach calculus to students. So, we can only have one crappy book to teach (unless a proper license is purchased) calculus from. You’ve got to be kidding me.
Sorry for the rant. Carry on.
The problem with your column, Mr. Dvorak, is that you describe an exceedingly obvious strategy, one discussed at length by luminaries in the Linux community for the first time more than seven years ago!
You look ridiculous discussing this as some sort of original, contemporary insight. Ditto your suggestion that Microsoft needed to get its feet wet over the years in open source litigation to figure this out.
People might be more interested in this column than you think…someone just sent it to me this morning. Personally, I run SuSE/KDE and also XP with dual boot (GRUB). Device support on SuSE 9.2 is stunning. The wireless on my thinkpad worked straight away with no fine tuning, etc. My only remaining tie to windows is a small amount of Windows programs that I have not yet been able to successfully run under Crossover or Wine. I see this as temporary though.
John,
I’m not sure that you know what are you writing about. Linux is not about little kids OS for playing games without drivers for brand new super dooper graphic accelerators. If Windows’s kernel will be mature enough and most of stable – people won’t look for alternatives. And whatsmore, Windows is really not ready for big iron and real world scenarios in big computer centers. I’m not sure if someone will be so brave to run Linux on windows kernel, even it will be possible. I’m sorry for people that they have to read this tabloid press…
I am one of those “spoiled” users who likes the _idea_ of Linux but I get so fed up with it. I am on my third new computer that won’t install Linux… each time it’s some piddly little problem that stops the whole installation.
For this new computer it’s my SATA disk. I wasted hours on the install and about $40 on a Dummies Linux book. After much web surfing I learned that there is no easy fix for this.
With Linux it’s always something!
Hank