Slashdot Mirror
Microsoft's CLR - Providing a Break from HW Vendors?
"It seems to me that once the CLR has matured enough, there won't be a need for Microsoft to wait for others to innovate on the hardware front and start offering its own hardware (and charge whatever it wants for it) to go with future versions of Windows.Net. Worst still, 99.99% of the population will not be able to say no to this strategy since they'll have no choice but continue using the Windows monopoly in order to run their favorite apps."
Jamie comments: I don't think it's about hardware innovation, or beating Java. It's about absolute control.
The big money over the next decade will be in transforming the computer into an entertainment device. AOL Time-Warner sees a computer as a revenue producer, with the unfortunate ability to copy digital works. They and the other five media giants want to put a stop to it; Microsoft and Intel will find it very profitable to help them.
One good step along the way is to give the computer a common interpreted language to run everything. We're there already. And when developers have to code to a virtual machine, not the actual bare iron, then whoever writes the virtual machine holds all the cards. And since the authors of the virtual machine will make a lot of money by enforcing intellectual property rights, the arms races are all over: copy protection is absolute, DeCSS won't compile, unauthorized MP3s won't play.
Of course developers rarely write on the bare metal anyway: we write to APIs, we write scripts, we write code that doesn't (need to) run in the CPU's supervisor mode. We're used to surrendering the ultimate control over the machine to the operating system, or to be more precise, to the BIOS that decides how and which operating system to run.
If we surrender this control, though, we'll find ourselves with a monopoly operating system that makes it impossible freely to write code for. (And it's not hard to cut off Linux and every other rogue free OS at the knees. The day that every motherboard's BIOS uses strong crypto to demand the master boot record be signed with a secret key known only to Microsoft is the day that Linux becomes a thing of the past.)
Naturally, to prevent you from firing up GCC and doing a rogue compilation of DeCSS or Lame or other unauthorized code, the operating system will have to stop you from running anything that isn't written in its language for its virtual machine. Requiring code to be signed by a central authority will make its first appearance as virus-prevention but its real purpose too will be control. Universities will be able to buy special licensed exemptions, at least until corporations decide universities are hotbeds of piracy and theft. At which point your alma mater begins teaching Computer Science 101 (and 201, and 301, and 401) in C#.
My prediction is that, unless antitrust legislation in the U.S. gets some teeth between now and then, the PC will become a Gameboy within fifteen years. Enjoy computers while they last.
While I absolutely despise the approach Microsoft takes toward software development, I can safely say they won't ever get "absolute control" over it. Yeah, they're big, they're rich, they're formidable, but they're also bumbling and very error-prone, as we all know from leaked e-mails, virus reports, etc.
The worst thing I see happening is a sort of class society, with Microsoft developing code for its circle of businesses, and everybody else in a sort of underground. Black market code, if you will. I very seriously doubt that things will come to, say, Microsoft getting the USGovt to pass a law forbidding software development by unlicensed, uncertified developers, and then fixing the game so only Microsoft developers can be easily certified.
Lately democracy seems to be based on the skybox, the Happy Meal box, the X-box, and the idiot box.
Why is control of the only layer that makes a difference evil when done by M$oft but innovative when done by Sun and Netscape?
(Java, the browser as a platform (see Judge Jackson's findings of fact) I have to admit that M$ is not being so obvious of their intentions, if that is what they are.
1. Who spends more money? Businesses or consumers? Businesses. Why the hell would MS want to transform a device for doing work into an entertainment machine? It just doesn't make sense. Think of it this way: Businesses buy pens, not crayons. I bet you see a lot more pens sold.
2. The CLR is just a collection of library code that developers can use or choose not to use. Think STL for many different languages. Already the CLR has support for many languages.
3. An evil empire built by Microsoft does not really benefit them in the long run. Microsoft is in the business of making money, not taking over the world.
I would expect to see a story with FUD like this in the Weekly World News next to Bat Boy's latest adventure, not in a respectable technical publication.
Why does everyone have to see the evil in whatever MS does?? True, they have done some evil things in the past, but can't people see that the CLR is part of an effort to accelerate interoperability of software that we developers will be creating in the future, regardless of what language it's written in?
I'm tired of reading about how everything M$ does is evil...they are a corporation, and they have their best interests in mind, just like other corporations(i.e. Sun). Let's stop focusing on the negatives and start focusing on the positives, like the fact that MS and Sun have done alot to work together to further the standardizion of the SOAP protocol!
Attention all planets of the Solar Federation! We have assumed control! - Neil Peart
When I was in high school, we were spec'ing out Alphas for the school file servers. The problem with the Alphas? No software. You could only buy them as File and Print servers.
J++ looked like it was going to change things. If you wrote Java code, it would in theory, run anywhere. If you wrote J++ code, it would run on any Windows. Given the Windows Everywhere initiatives (the separate NT, Windows, and WinCE lines), J++ would have given Microsoft that platform independance.
MS wanted to split from Intel years ago. Everyone thought that Intel was dead after the Pentium. RISC processors were blowing them away, and Intel's CISC ISA was holding them back.
Well, Intel figured out how to build a RISC processor with a hardware decoder, Windows NT took off faster than expected, the 64-bit Alpha version never shipped, and now MS/Intel split a HUGE monopoly.
This gives their Windows Everywhere initiative some teeth. They are pushing Win32 APIs everywhere, but you need to code differently for the Xbox, Win32, or WinCE. Sure the APIs are the same, but not the compiled version.
The CLR means that Windows is Windows, and Windows code will run there.
Look at UNIX, there has been decent source compatibility, but no binary compatibility (until the recent Linux emulation everywhere). Outside of software distributed in source form, nobody supports every Unix, just the 1-3 that are profitable for them.
Source compatibility helps, but isn't enough. The CLR gives a form of binary compatibility.
Sun could have had this market with Java, but they fucked up. We'll see what happens.
MS won't be able to do that without a major compatibility break, and they've just barely thrown off (or started to throw off) DOS. Their grip on the upgrade cycle has already started to loosen with Windows XP. You're acting as if MS will break into your house and force you, at gunpoint, to install XP 3.0. Don't be stupid.
And you can always get a Mac or something.
Eventually the CLR will replace Microsoft's platform revenue stream.
Right now they get a nice chunk of money every time somebody buys a PC. Windows is one of the most expensive components of a desktop computer.
If you look far enough down the road, Linux on the desktop is a reality. So they know that the OS monopoly is coming to an end. It is time to start getting a new monopoly ready to take its place.
They will ride this gravy train as long as they can, and then they will concede the OS market and start charging the same per-computer tax for the CLR. They won't care what OS is running underneath it. The OS will become a low-margin commodity, and they may even just starting giving Windows away for free. The profit margins will simply be relocated upward to a higher layer of this new and thicker notion of a platform.
BTW, don't even think about suggesting that Java will win because it was here first. Java is to the CLR as Lotus-1-2-3 was to Excel. Some people innovate. Other people specialize in refinement and broad market penetration.
Eric Sink
Software Craftsman
People are always assuming MS is going to pull some crazy crap using its monopoly power.
It does of course, but nothing like what Jamie is thinking. Whenever it does try something bizarre, like making MSN only work with IE, people call them on it. And they stop.
And if they pulled something like this, they'd have to. The DOJ isn't going to sign off its case without some sort of oversight.
And I think the oversight committee might have a problem with
"Proposal 1A: Drop support for any PC that's capable of booting a non-MS OS."
These stupid ideas only serve to make the real ones look silly.
Why should Jamie get to post moderation free, Katzian garbage like this? Put it in a comment like everyone else.
.
Let's not stir that bag of worms...
Choose? M$ doesn't give you a choice.
3. An evil empire built by Microsoft does not really benefit them in the long run. Microsoft is in the business of making money, not taking over the world.
No, they don't want to 'take over the world' they want to take over the OS, computer, consumer device, media and content, media and content delivery, media protection, and Internet business[es]. Bill Gates' dream is to have you buying everything from them except groceries.
See the above point I made. They are in the business of making money by taking your choices away.
Get your Unix fortune now!
Java would have been godsend for Microsoft, addressing all these problems, but they didn't control it and it would have given people not only hardware independence but also Microsoft independence.
Technically, there are no significant differences between the CLR and the JVM. The CLR isn't any more or less powerful than the JVM, it won't run much faster or slower, and it won't be any easier or harder to implement. You already have Java compilers for the CLR, and you will see C# compilers for the JVM soon. But Microsoft controls the evolution of the CLR, and that is what matters to them. While Microsoft will probably implement the ECMA standard, they will extend the CLR and libraries in numerous proprietary ways, and that will give them exactly the control they want.
This is just wrong. Hardware is hardware and has no idea what seqences of instructions do. They execute an instruction, then another, then another. You put your code in memory and feed the CPU the address of the code. You can always go under the operating system (stick in a boot disk that loads the OS on top of something else). There's no way a machine could block "illegal code".
Now, maybe a chip that only executes signed bytecode could do something like this. But then development would be essentially impossible and there would be no programs for that achitecture (and if you give developers the private key, it will be public in seconds; hell I'd do it!!).
My other car is first.
I'm a disgrunted MCSE that didn't like the early retirement of the NT 4.0 certification. We were under the impression that it would last until the NT 6.0 certification process, but that didn't happen. I don't like Microsoft.
HOWEVER.
They do play games (Windows isn't done until 1-2-3 won't run, the DR-DOS Win3.1 beta fake error, etc.), but less often then you think. Half the games that they play stem from the fact that their employees don't look outside the Microsoft bubble.
Though I can't find it now, on MSN's Canadian Xbox page, they claimed that it was the first console to support 4 players. This is a company that is SO huge that adventuring to the rest of the tech world involves looking at other divisions. When they break standards, half the time I doubt they realize it. When they do things based upon their bastardized standard in another program, they may not realize it.
It's a large company, they can't act as a single mind despite what Slashdot thinks.
Alex
I think a lot of folks here have been focused on the Linux/desktop issue for so long that they're not understanding what really drives Microsoft: $$$$. (Well, I assume everyone *knows* that... but not sure how many understand it.)
The CLR has two implications.
The first many have commented on... hardware abstraction. Applications compiled for the CLR will be able to run on a wide-variety of different (but similar) platforms... but is this really of long-term value? Are there a lot of applications begging to run unmodified on your enterprise server AND your Palm? Doubtful. Hardware abstraction makes good engineering sense in the sense that it saves future development, but I don't see it as much of a market-stealing development.
Will Microsoft have an advantage over Intel? The ability to move away in the future? Newsflash, it already has that advantage. x86 is, for all intents and purposes, an open standard implemented by a variety of hardware manufacturers (down to AMD and Intel at the top-end.. for now). How will CLR give it more of a death-grip? As someone else said, this aspect of the CLR is equivalent to the HAL.
No, I believe it's the second implication that Microsoft really cares about: multiple language interoperability.
The market Microsoft is going after with CLR is really the enterprise computing market. There is an awful lot of existing business logic written in a wide range of language offerings, and the value in capturing that market is huge. Microsoft is making this move on the basis of a prediction on where enterprise software is headed over the next 5-10 years.
Different pieces of logic (within different systems) are begging (so M$ believes) to interoperate within a single application server, within a single runtime. XML/SOAP/Web services is a basic solution for cross-process interoperability... but what's going to run on the *back* end? Within the same process, with shared rules for security/type-safety, object/thread pools, garbage collection, and shared state?
Java threatened to be the default language to which business logic/applications/"Web services" were about to be built with... which obviously would represent a threat to Microsoft's position. Microsoft made a valiant effort to head this off with COM/COM+, but quickly realized that the fundamentally C++ nature of COM+ was making it not attractive enough for business developers.
The introduction of CLR is trying to change that. Multiple languages, multiple types, multiple run-time semantics... standardized in to one run-time. C++ objects making calls on Java objects making calls on COBOL logic...
.... that's the vision of CLR, and why the focus of the CLR paper is about the language features of the CLR, *not* the 'generalized hardware' nature of the hardware.
And this would happen without anyone batting an eyelash, I'm sure. DOJ would be happy with it.
I don't think so. And even if the situation came to pass:
A: It would be easy to remedy this situation, and it would be remedied via antitrust action (though perhaps some group would need to be formed to validate and sign OS booters from open source vendors).
B: The market would supply a vendor who produced equipment to run other OS's.
This is the problem with the "slippery slope" style of arguing. You don't try to evaluate the problems with some projection, you just view it as some inevitable consequence of something reasonable. Everything gets bent into some crazy, hypothetical world where nothing is as it is now.
Here's a projection: Linux will overcome MS by providing a better product for free. Seems a lot more likely than Jamie's scenario.
Why can't this be the topic of our anti-MS conversation: What can we do to make Linux better?
...
Let's not stir that bag of worms...
The poster's original question seems to be a reasonable, thought-out question about the implications of VMs in software development.
Too bad it's followed by 4 paras of paranoid rant, which is what people are replying to, by and large. Why doesn't Jamie just post in the forum, like the rest of us proles? Even if I'd blocked him from my view of Slashdot (which I haven't, although looking back over the stories...), this would slip through as a rider on Cliff's story.
[anyway, what is the benefit to BIOS makers and motherboard manufacturers of limiting their market? The degree of support for overclocking in existing mobos and BIOSes shows that they don't care what their large partners think (Intel, AMD)]
"don't fall into the fallacy of believing that Perl can solve social problems. Maybe Perl 6 can, but that's a ways off"
Microsoft is porting .NET to FreeBSD. How does that help them establish "Windows Everywhere"? They aren't suing or threatening Miguel for his Mono project: in fact, they seem to be encouraging it (or amused by it) judging by
the interview with him on MSDN.
C# is a nice, clean platform for Windows GUI development. And ASP.NET is cool enough to give IIS a feature edge (as opposed to a security edge) over Apache. Not that the Apache group couldn't create something like Tomcat to serve ASP.NET from Apache, mind you. .NET is, after all, an open standard.
Microsoft needs .NET because Microsoft's customers want an easier way to develop. Period. Apple developers love coding for Apple. Linux developers love, well, anything anti-Microsoft. And Windows developers--God help them--should be able to enjoy Windows development. Having actually written an app or two using the CLR I assure you that it is much more enjoyable than MFC.
IANAWT (I am not a Windows Troll). I am a BSD user looking forward to .NET on BSD. I am a Perl coder looking forward to Perl.NET. And yes, I use and code on Windows at work. And for these reasons, it's very cool to have .NET.
I can't believe almost everyone missed this! Everyone thinks this is Microsoft's way of reducing Intel's influence. WRONG! This is the perfect way to MAINTAIN the Wintel hegemony!
The scream you hear every morning is the Intel engineer realizing he/she has to spend another day working on the x86 architecture. Everytime he/she pushes a polygon around on a cad system, he/she curses the baroque design decisions that were made 20 years ago. Intel tried to kill the x86 on multiple occasions (80860, 80960, etc) and failed miserably every time. Their most recent attempt, IA64, shares the same departure in binary compatability.
To wean the market away from x86, Intel has recruited Microsoft to create a language and development environment that generates architecture independent byte-codes. Intel rooted for Java in the early days, but now it is obviously a niche player. Microsoft thinks it gets a leg up on Intel but it is Intel with the most powerful compilers (check out icc) and the fastest hardware (Intel hired away all the Alpha engineers). Wintel lives on.
Microsoft has a multi-prong approach. It is similar to Sun's Java write once run everywhere strategy, but with the benefit of history and money on its side.
.NET Framework (JFC/Swing) which will eventually replace the Win32 API. Once Microsoft ports the CLR to a new hardware platform or operating system, it is simple to also port the .NET framework. MS really doesn't need to port Windows unless it wants everything from the hardware up (as in JavaOS).
.NET framework in it's upcoming Windows .NET Server (aka Windows XP Server), and will have it included as free upgrades for Windows 2000 and Windows XP before the end of the year. This means that MS could potentially have tens of millions of .NET ready systems on the street before the end of the year. On advantage MS has is that in its first incarnation the .NET Framework just hooks into the Win32 API, giving them time to rewrite the entire Windows codebase (supposedly due with the Blackcomb release).
.NET Framework and eliminates Win32, their will be nothing stopping MS from porting Office to any hardware and/or OS platform on the market. 3rd party developers like Adobe, Macromedia, etc. can port their applications to .NET now with a tryed and true customer base, and once MS is ready, jump with them to other platforms/OS' with an almost minimal risk and expense. Instant application base. The first candidates are MAC OS X and Windows CE (.NET). Adobe for one will probably welcome having less codebases to maintain. Any port that makes economic sense to MS is a candidate, including Unix and Linux.
.NET framework. As new hardware or OS' hit the market, port. Where as Sun could port Java to any enviroment easy enough, it doesn't have the same application base as MS.
MS has the CLR (Common Language Runtime) and MSIL (MS Intermediate Language). This is nothing new, with CLR = JVM, and MSIL being javacode. Additionally MS also has the
So if this is nothing more than MS rehash of Sun's Java approach, what's the difference.
.
First MS has the advantage of learning from Sun's mistakes. For example C#, Visual Basic, & VC++ are not the only languages that can use CLR & MSIL. Any language can compile to MSIL, and MS encourages it, claiming over 20 languages from 3rd party vendors, including PERL and Java. Additionally MS supports both compiled and bytecode, with a built-in native code compiler as part of the framework. These were all possible with the JVM, but not advocated/pushed by Sun.
Second, instant market. MS is including the
Third. Applications. Microsoft has Office. Lets face it. People don't buy Windows for IE and Solitaire. Java never had a killer app.
Fourth. Inertia & Clout. Once MS ports Office to
Five. Future proofing. If the DOJ or anyone else causes problems, MS can easily port Office to Linux just by porting the
It didn't happen then.
I'm not worried now.
It's all fine and dandy to be idealistic about freedom and the like, but a quick examination of society suggests that freedom means very little to your average consumer. What most people care about is convenience , the ease with which they can do their "thing." And Microsoft's CLR, while rough around the edges now, brings "convenience" to developer's lives.
I've developed a lot of code in the last couple of decades, and I gravitated to Java because it was an easy way to write GUIs. Sure, most of my work is the heavy lifting "under the hood" -- but it's the GUI that attracts users who buy product. You and I may love command-line environments, but that isn't what most people want or need. As one of my co-workers puts it, the GUI guys get all the glory -- and the CLR is a superior tool for GUI development.
Why use the CLR? Because unlike Java's Swing, GUI code written in the CLR is reasonably fast (through native widgets) and easy to use. The Visual Studio development environment takes most of the challenge out of GUI development -- and toys like NetBeans/Forte and KDevelop don't even come close to Visual Studio when it comes to easy development. The only advantage Java has now is portability...
The CLR has little or no affect on my engine development; I still write my code as portable C++/Fortran/whatever, and wrap it in a component architecture that can be dropped into a GUI. Microsoft has not made traditional compiled code obsolete -- what they've done is make MFC, ATL, and COM obsolete. In other words, Microsoft is creating a user-interface toolkit that can be used to wrap code that does heavy lifting. They're making it easy and efficient to write GUIs for Windows -- and that, my friends, is what is going to hurt Java and Linux.
The CLR isn't about getting rid of Intel, or platform independence; it's about attracting developers who write code that attracts user who sepnd money on Microsoft operating systems. The Linux developer community would be wise to spend more time on ease of use and less time tilting at windmills.
All about me