Is Linux At the End of Its Life Cycle? 676
Glyn Moody writes "That's what Nikolai Pryanishnikov, president of Microsoft Russia, seems to think. Quoted in the context of continuing questions about Russia's plans to create its own national operating system based on GNU/Linux, Pryanishnikov said [via Google Translate]: 'We must bear in mind that Linux is not a Russian OS and, moreover, is at the end of its life cycle.' An off-the-cuff comment, or something more?"
Google Translate (Score:5, Informative)
"We must bear in mind that Linux is not a Russian OS and, moreover, is at the end of its life cycle."
could also be:
"We must bear in mind that Linux is not a Russian OS and, moreover, is deprecated"
"We must bear in mind that Linux is not a Russian OS and, moreover, is obsolete"
"We must bear in mind that Linux is not a Russian OS and, moreover, is old fashioned"
Does anyone have the exact translation for what the guy really meant or just a Google translation.
Also, of course it's off-the-cuff. A Microsoft guy saying nothing more than "Linux is [i]x[/i]" with nothing more to back up the statement or shed more light on it.
This is news?
Just Days After.... (Score:5, Informative)
...Steve Ballmer said that (paraphrasing) Linux is what all our competitors use
This was in response to a question by their stockholders about the possibility of breaking the company up
http://www.techflash.com/seattle/2010/11/ballmer-and-gates-heres-why-were.html [techflash.com]
Russian OS? (Score:3, Informative)
Linux is not a Russian OS
Neither is Windows! I don't see the relevance of that statement.
Re:And Windows is? (Score:3, Informative)
Windows retains a lot of very insecure backwards compatibility cruft (eg lanman hash types to cite just one example)... Linux is far better in that regard..
It was NT which was the ground up rewrite, but although NT provided a new kernel they bolted a lot of the existing legacy cruft on top of it, many of the security holes in windows are a result of weaknesses in (or as a direct result of) this cruft rather than the core NT kernel.
Re:It's Hindsight (Score:1, Informative)
1. cmake sucks ass
2. cmake generates makefiles for gnu make, amongst other things
Re:It's Hindsight (Score:3, Informative)
cmake is built on OS tools, including GNU make in the case of Linux. What you're thinking of is the use of cmake over GNU autotools. And we're all happy for it.
The GNU Compiler Collection isn't going anywhere (though competition from LLVM is good and welcome), but the sooner Autotools dies, the better.
Re:And Windows is? (Score:3, Informative)
Re:And Windows is? (Score:2, Informative)
Re:And Windows is? (Score:5, Informative)
It's pretty brazen of you to imply that Windows is less secure than Linux. Put a desktop distro on linux and connect it to the internet, give it Window's marketshare and watch hackers make swiss cheese of it.
LOL. Where do all these 'there is no difference in security between operating system' trolls come from?
Wasn't Ubuntu pulled from OS cracking contests recently because it was too hard to crack when compared to Windows and MacOS?
Re:In Soviet Russia... (Score:4, Informative)
Did you miss the part where it says the dude is a Microsoft employee? He's a capitalist, not a socialist. Not that the facts have ever stopped you from making cracks about your imaginary version of socialism.
Re:And Windows is? (Score:3, Informative)
Sorry, but thats nonsense. It's the old argument o proportions. The Linux kernel is not as perfect as the BSD kernel :P but its by far more modular, open source (so those hackers could start right away) and witten with a different methodology in mind. Also there were no time constraints but lots of hackers and gifted coders around the globe that took their time to write something they would like. That's in sharp contrast to the thing a company has to do to pay its employees. In case you still don't believe me check the security records and forget about comparing the numbers of bugs/flaws but their actual impact/severity on the OS. Finally check out what most ISPs do. Yes, they run GNU/Linux, not even because it's cheaper and more stable, but also because its more secure.
Re:And Windows is? (Score:3, Informative)
That's the same old argument that has been variously applied to Linux, BSD, and Macs. At least in the Linux world, it doesn't hold water, since the code is subject to so many more eyes. And in the case of an exploitable bug, a patch is usually available within a few hours, not weeks, as on a "patch Tuesday" system.
Additionally, Linux (via distributions) updates all of the software on a system from one location. Windows is getting there (Windows Update), but still has a way to go (Flash, Reader, and Java updates). As a result, Linux application bugs get fixed just as fast as kernel bugs. In the Windows world, even if your Windows install is up to date, your version of Office, Internet Explorer, or TurboTax might still have a critical bug.
As a native Russian speaker... (Score:5, Informative)
Yes, he did literally say "end of life cycle". Most probably because in modern Russian corporate-speak expressions and terms like this are direct translations from English (in the same way 300 to 100 years ago they were borrowed from French :) ).
Paul B.
Re:And Windows is? (Score:5, Informative)
By implying that Windows has undergone a ground up rewrite wile Linux has not, you imply that you already know the answer to your own question, which means your question is not at all earnest.
One might even say he was begging the question...
Re:And Windows is? (Score:5, Informative)
Nope, WOW was part of WIndows 95, and allowed 16-bit Windows 3.1 apps to run on the 32-bit OS. Sure, Microsoft applied the desktop shell improvements to both product lines in parallel, but the desktop shell has never really been the problem with Windows. The real backwards compatibility comes from Win32 - allowing user-mode code to use the same systems library with either kernel. Basically everything you can call the OS, from the kernel up through the implementation of the systems libraries, was different between NT and Win9x.
If you just want to argue that the new OS was backwards compatible for apps with the old one, sure - that's true. It's also why Windows won. People like thier apps.
Re:And Windows is? (Score:3, Informative)
Finally check out what most ISPs do. Yes, they run GNU/Linux, not even because it's cheaper and more stable, but also because its more secure.
Comparing linux on the server to linux on the desktop, in terms of security, is comparing apples to oranges.
Re:And Windows is? (Score:3, Informative)
The common theme in the replies here is that linux on the server side is secure. No shit. Go read my original response and you'll notice I made it a point to say "Desktop."
Re:It's Hindsight (Score:4, Informative)
the sooner Autotools dies, the better.
I quite like autotools, actually. If you actually think about what you're doing when writing your configure.ac [gpleda.org] and M4 macros [gpleda.org], it's an elegant, clean and easy to understand solution.
Unfortunately, at the moment it seems fashionable to throw all the configuration macros into a single, poorly commented file, with all the code copied and pasted from other projects with little understanding demonstrated of what it does or why it does it, with the predictable poor performance and low maintainability.
Re:It's Hindsight (Score:4, Informative)
I hear this idiocy all the time on IRC. When it happens, I ask if the person if they actually know how to use autotools. The answer is *always* no. Usually with a justification like "why would I want to learn a system that sucks?"
You gave zero reasons for bashing autotools, so I put you in this same camp. Back up your assertions or GTFO.
It is actually a handy piece of software. When used properly, most projects need just one or two macros - AC_CHECK_LIB and AC_CHECK_HEADERS and then just list out your sources and flags in a Makefile.am.
There's really very little to complain about. It does it's job, does it fairly well. The only catch is that you have to RTFM.
Re:And Windows is? (Score:5, Informative)
Umm.. no.
Windows NT was released in 1993, 2 years before Windows 9x. So it's not really possible for it to have been a port of 9x. In reality, the guys that wrote the 9x UI were NT guys who were on loan to the 9x team. They had intended to write the UI for NT (then code named Cairo) but 9x got higher priority due to the need to bridge the dos/nt barrier with app and driver compatibility.
Re:Google Translate (Score:2, Informative)
Google translated that sentence perfectly. He actually used the phrase "life cycle". The rest of the translation has minor errors and obfuscations, but overall is not too bad. Automatic translators do much better when translating formal writing since the grammar is more regular.
Since you asked, here is a human translation which you can compare to the Google translation:
"Microsoft Corporation defends technological neutrality and is of the opinion that the choice of OS should be predicated solely on the characteristics of the OS itself, on bang-for-the-buck, on the actual real-world problems to be solved, on security, and not on ideological considerations."
From our point of view, the most effective way to develop the country's innovation economy is not to create an equivalent of an existing OS, the cost of which would be enormous in time and resources, but to start with the most widely OS, which has been tested by the Russian special services, in order to create our own applications and solutions, while at the same time investing resources in Russian scientific projects which have a future. One should remember that Linux is not a Russian OS, and moreover it is at the end of its life cycle.
(You will note that he is contradicting himself: To use Linux is to create a new OS (at great trouble and expense). But we should not use it because it is at the end of its life cycle.
Re:And Windows is? (Score:5, Informative)
What kind of reasoning is that? Sounds like a very elite hacking contest...
If the 'elite hackers' can break into Windows and MacOS but not Ubuntu, that should tell you something.
Besides, there have been countless amount of Linux hacks and exploits.
No there haven't:
a) the number is clearly countable.
b) the number is far, far less than the number of Windows hacks and exploits.
c) the Linux exploits are generally fixed much faster: my Ubuntu machines are normally patched automatically before the exploit hits the media.
d) Windows has staggering amounts of insecure backwards compatibility crud which guarantees security holes. For example, including the current directory on the DLL search path by default... that is quite simply insane, but Microsoft won't change it in case they break WhizzbangSoft-95 and those users complain about it.
Re:And Windows is? (Score:5, Informative)
Kids these days.
Windows 95 was an updated GUI running on DOS. You must be thinking of something else.
Lessee... Windows NT 3.1 was the Windows 3.1 GUI running on a new (NT) kernel.
("New" is relative, as NT was created by a bunch of VMS coders, from which it gets message passing and other features. One could argue somewhat whimsically that Dos-based Windows up to ME was based on 1981 technology, and every Windows version since then was based on 1975 technology.)
NT 3.51 would be called a service pack today. It was pretty solid for the time.
Windows NT 4.0 was the NT 3.5 core with a GUI that looked more like Windows 95.
Windows 2000 (still my favorite Windows desktop for business use) was basically a huge service pack on NT 4.
Windows XP was a substantial update of 2000, but by no means a "ground up" rewrite.
Vista started as a "ground up" rewrite (Longhorn) but was plagued by project delays and restarts. I'm not certain, but I wouldn't be surprised at all that what actually made it to GA had a substantial amount of XP code.
Then there's discussions on thunking and code reuse and backwards compatibility...
I'm by no means an expert, but I don't think that Windows has ever had a complete bare-metal ground-up rewrite.
But if it did, it was not Windows 95.
Re:And Windows is? (Score:3, Informative)
Linux gets attacked and compromised a lot.
Attacked a lot? Sure. compromised a lot? No. Barely ever. The only common Linux attacks are just scripts that check the system for default passwords. There are other attacks but they are extremely rare.
I mean, it happens, but it is so damn rare to come across an infected Linux (or any Unix or Unix-like OS) machine that you might as well also claim that women give birth to sextuplets "a lot" as well.
Re:And Windows is? (Score:5, Informative)
Re:And Windows is? (Score:3, Informative)
Telnet is also included in ever version of Windows. The telnet client is not a security risk if used properly, and there are devices out there that do not speak SSH.
Sure, running a telnet server could be a security risk. But if you are using it on an intranet, maybe not.
NFS was not "highly insecure" although there have been vulnerabilities in certain implementations.
One key difference between Unix and Unix-like systems, and Windows, is that Windows attracts hordes of clueless morons who use the computer like an expensive typewriter, and do not know about security, nor do they wish to know. If their machine is compromised they bring the machine to the computer man. Linux and Unix users tend to be a bit more savvy.
Re:And Windows is? (Score:3, Informative)
Ubuntu, by default, has one open port [ubuntu.com], for Avahi.
Windows 7 has seven, including Netbios, RPC and SSDP.
A decent distro should have none, but I would still trust more Ubuntu than Windows 7, with its 700% more open ports, including the famously secure netbios /sarcasm.
Re:It's Hindsight (Score:3, Informative)
The use of Dash as the default shell over Bash
That's just a matter of choosing the right tool for the job.
Bash is (as POSIX shells go) very full-featured with a bunch of enhancements (nonstandard extensions) for use in scripting and various niceties for interactive use. It's one of several shells these days that are commonly accepted as good choices for login shells.
However, quite a lot of shell scripts on a typical system don't need or use bash extensions. (And Debian policy is that shell scripts that are interpreted with /bin/sh should not use any non-POSIX extensions) A shell that's not so rich in features, but rather written with the goal of providing a quick and compact implementation of the standard is a better choice for these shell scripts. If switching /bin/sh to point at dash produces a noticeable improvement in boot-time, that justifies dash's existence on the system. Interactive command history (via readline), tab completion, associative arrays, built in regex, and so on justify bash.
Re:In Soviet Russia... (Score:3, Informative)
Wrong, as I explained, labor doesn't get you a portable, tradable, universal, and everlasting unit of wealth exchange, it gets you products.
Here's capitalism: A man knocks on your door with an offer, you get to continue top breath his air and eat his food if you suck his cock.
It is not stealing if you get something in return, and you freely accept the exchange. If you don't like what is offered by society, don't live in society. You have the choice not to play, you can live as a hermit and no one will bother you. But as long as you would like to be a part of our mutual benefit society, you will pay the membership dues.
Without society, those Amish have no ability to trade in money. They can barter, using whatever of their perishable crop survives, directly for items from others. They have every right to keep the direct product of their labor, the food and goods they produce. If they want to use currency, they need to be part of a society. And if they want to be part of a society, they need to pay he membership dues.
Libertarians want to be members of the club, they just don't want to have to pay to be members of the club. Which makes libertarians no good freeloading leaches.
Re:It's Hindsight (Score:3, Informative)
Here's what is good about autotools: 'make dist', ./configure --enable-XXX, and easy integration with debhelper and rpmbuild.
I've got a small ~60kloc project [sf.net] out there that I started out using a simple makefile and C code, and then later migrated to gettext and autotools. I really wish in hindsight that I had just started with GNU Hello World and gettext from the get-go and then built out my project. As it was, I spent days re-factoring strings for gettext and more days getting my configure.ac and Makefile.am doing what they should have done rather than what I wanted them to do.
But now I've got a build system that works very cleanly on Linux, OS X, FreeBSD, Cygwin, and Win32 (using the mingw32 cross compiler), and can also be used with dpkg-buildpackage to make .debs. All with the same configure.ac and Makefile.am, and automatically compiling the right things when a header is changed. I have conditionals for three different user interfaces and two optional supporting libraries. All told it was a LOT easier than make by itself or even boost.build.
Re:In Soviet Russia... (Score:3, Informative)
It means different things in America than in the rest of the word, where it is just another term for 'anarchist.' Here in America, it means "Individualist free market anarchist," and implies a belief in absolute property rights.
I would wager that I know quite a bit more about libertarians and the history and philosophy of anarchism than you do, in fact, it is because I know what anarchism is that I hate libertarians. American libertarians are basically nothing more than closet fascists.
Re:In Soviet Russia... (Score:3, Informative)
If you're truly wanting to live as a hermit, why would you care about social constructs such as laws?
Because trespassing on someone else's land (as in trying to live there) will get you either shot, or arrested and thrown in jail. You can choose to not care about laws as much as you want, but the police don't care about your choice.
It's not like there's tons of lush, fertile, unoccupied land out there that you can live on and get all your food from. If the land is the kind that you can live off of, there's either someone already there, or it's a government park (and park rangers aren't very accommodating of squatters). If the land is wide-open with no one around for hundreds of miles, then it's also barren and not a place where you can find food.
A couple decades ago, some guy named McCandless tried to be a hermit in Alaska just for about 6 months, and he died of starvation because he couldn't find enough food to sustain himself, even though he was in a remote place where no one cared that he was staying.
Re:In Soviet Russia... (Score:2, Informative)
> how would you call making $20000 worth of goods in a given month and being paid a $1000 salary?
I'd call it delusional or very stupid. If your sole labor is creating $20K of value and you are only getting $1K of it you are an idiot for not striking out on your own. But you are more likely just delusional and have an overly inflated sense of your worth and dismiss the other inputs in the value equation.
Your mythical worker is only adding a little more than $1K to the equation. On a typical product there is a lot of inputs, not just labor. There is raw materials, the plant & equipment, R&D costs, marketing, overhead (payroll/hr dept, health benefits, unemployment insurance, liability insurance, warranty & support costs) corporate taxes, financing costs (dem crafty joo bankers gotta get their cut) and a thousand other things the typical socialist never dreams of. Then there has to be profits left over to pay the owner/shareholders to grease the wheels of capitialism to launch the next round of companies, and ya might have to buy hookers and blow for the venture capitalists, etc. And that is equally important, no profits for the capitalists and no new jobs. See the current economic situation for an example.
Re:In Soviet Russia... (Score:3, Informative)
This is my point exactly: the places left in the world where you are able to get lost and not be noticed are places where it is extremely difficult or impossible to survive on your own. Those places are empty precisely because they're not nice places to live: the land isn't very fertile, the weather too extreme, etc. Humans require a fair amount of land per person to grow the food they need to survive. Back in the days before agriculture, this wasn't the case, because humans just took food from the land: either wild-growing plants or wild animals. However, we overpopulated the world several thousand years ago to the point where it was impossible to enjoy a hunter-gatherer lifestyle, and we had to invent agriculture to survive in our increased numbers. In the transition, humans' health got much worse, and they lost a full foot of height due to the much poorer nutrition that an agricultural lifestyle provides, until they figured out how to vary their diet enough to compensate. Jared Diamond wrote a whole book on this.
If there were still nice places to live where you could live as a hunter-gatherer, people would be moving there. The places that are empty are places where no sane person would want to move to. Some of them are actually quite beautiful in their wild state (like Alaska), but even then they are not very habitable (very cold and not suitable for agriculture). Sure, Ray Mears might be able to do it, but he's an exception. Plus, it would only take a handful of people like him in one area to completely exhaust its resources: there's only so many huntable wild animals to go around.