This isn't theoretical: numerous operating systems have been implemented in safe, garbage collected languages, including Modula-3, Lisp, and Smalltalk.
Safety doesn't mean that you can't do unsafe things, it just means that you won't do unsafe things accidentally. C#, for example, has the same pointer manipulation primitives built-in as C.
And garbage collection doesn't have to mean that your program gets interrupted at inconvenient times, while C memory managers make no guarantees and are actually quite unpredictable in practice.
(Sun Java and Microsoft.NET are unsuitable implementations for writing a kernel, but with a good native code compiler, Java and C# are perfectly fine languages for implementing a kernel.)
Put virtual machines on top, like Java and.NET. Claim that they're more secure than the OS.
It's not the virtual machines that make Java and C# applications more secure, it's the fact that the languages are better designed than C and C++.
It's not working.
Sure it is. Java applications have no buffer overflow problems, no low-level memory management bugs, and no undetected type errors.
If the Hurd guys had a clue, and could write something as good as QNX, there might be some hope from that direction. But after ten years of screwing up, there's not much hope there.
Microkernels are the wrong solution: they attempt to compensate for limitations in the programming language used to write the kernel by introducing unnecessary complexity and overhead in other places.
Until kernel designers kick the C/C++ habit, we are not going to get a decent kernel.
It's unfortunate that Java and C# (pretty decent language designs) happen to be implemented with such bloated runtimes by their main proponents. But that's not necessary: Java and C#-like languages can be implemented natively with pretty much the same efficiency as C and C++.
I don't see a "rush" to produce Java-capable products. Java has become a niche language: some server side apps (for people who don't know any better), introductory CS courses (hello, Pascal), and small cell phone gimmicks (that don't work well across phones).
It's a shame, but.NET probably has a bigger installed based than Java. And while Microsoft's monopoly gave them an unfair advantage, this is still Sun's mistake: they had a half dozen year lead until Microsoft finally got their act together, and Sun still screwed it up.
There is a whole range of enrichment methods; see here for an overview. Note the article on EMIS and use by the Iraqis. Furthermore, the chemical exchange processes sound like something competent chemical engineers could implement. And I suspect that with improvements in materials, computers, and chemical engineering, such processes become cheaper and easier to implement.
I think we'll just have to face the fact that nuclear weapons will become accessible to many more nations over the coming decades.
Well, I think this has been clear for a long time. The Windows XP kernel and many of its key user mode libraries will continue to be written in C/C++. I suspect that Explorer, Internet Explorer, and the control panel will likely continue to remain C/C++ based.
However, it looks like they are going to ship a full.NET runtime with Longhorn, as well as lots of.NET libraries. I suspect that if you removed the.NET runtime, some applications and system utilities would break, although the system overall would probably still boot. (Didn't SP2 come with a.NET runtime anyway?)
All of that is pretty reasonable. Why break working code? Why alienate thousands of developers? The inclusion of.NET isn't a revolution, it's an evolution. Think of it as a Visual Basic replacement--a better designed runtime with a choice of better designed libraries. And, unlike Visual Basic,.NET may actually be good enough for Microsoft to start writing small applications and system utilities in.
It'll run them exactly as fast as any other PPC 970 core. As far as I can see from the information that's been released so far, to use the coprocessors at all you'll need to redesign your application around an asymmetric coarse-grained parallel processing model
Applications are already written for that model: they use APIs like BLAS and LINPACK that hardware vendors can reimplement to take maximum advantage of their parallel capabilities, and they use intrinsically parallelizable language primitives (like Fortran's parallel array primitives). Furthermore, many applications have been parallelized for either multithreaded execution or MPI. In different words, today's high-end numerical applications already contain the hooks and declarations for extensive SIMD and MIMD parallelism.
If the Cell architecture can't provide high performance for those kinds of existing parallel codes, then there is something wrong with Cell and/or its compilers.
2. How do I get put in charge of the 'Hot Chick' section
You need to be a woman, or at least be able to pass for one. In order to find the "chicks" attractive under those circumstances, you need to be lesbian or heterosexual but transgendered, respectively. Oh, and you still need to keep quiet about it because if you are found out, you'll be transfered.
To the author of the article (a "professional" writer?): the proper gerund of "backscatter" is "backscattering"; "backscatting" means something entirely different (musical or scatological, take your pick).
To all the readers, if you want to see thousands of nude bodies parading around (in color), go to a nude beach. I think you'll soon appreciate the custom of wearing clothes, and you'll see that we don't wear clothes to prevent lust and desire, but to create and enhance them. Clothes allow even people with fairly unattractive bodies to look decent; as geeks, we should all appreciate that.
I pity the screeners that have to look at thousands of average human bodies without the protection of clothing.
With Cell, IBM keeps talking about "theoretical GFLOPS". I don't care about theoretical numbers. What I care about is how fast the thing runs when I run normal code compiled with a normal compiler and (possibly hand-optimized) numerical libraries.
So, what kind of SPECfp numbers does the thing get? What kind of BLAS performance does it get?
They have 2.6.11 running on it, so compiling the benchmarks should be trivial. If they haven't published anything yet (I haven't seen it), we have to believe that the numbers are less than impressive.
(Another company used to make inflated claims about the performance of their processors by computing theoretical maximums for a few SIMD instructions, unachievable in most real code. When people actually did some real benchmarks and published them against the wishes of the company, they found that their processor was no faster MHz for MHz than Pentium on real code with real compilers.)
The industry is not taking Barella's statements lightly. In a letter last month addressed to the French Minister of Justice Dominique Perben, more than 20 representatives of France's entertainment, music and film association bodies and advocacy groups expressed their outrage that French content was not pirated to anywhere near the same degree as content in English and even German.
"It is an insult to French Culture that American and even German language music, rap, and books are being shared widely over file sharing networks, while nobody else in the world bothers with French content; even our own teenagers prefer American content, despite decade-long indoctrination in French government facilities." wrote Perben.
Perben proposed criminalizing the distribution of copyrighted content containing lyrics that were not easily recognizable as French, creating a high-speed French government Internet that would facilitate rapid piracy of certified French content, and scanning user's harddisks for foreign content, whether copyrighted or not. He also said the French government was considering financial rewards for particularly active filesharers of the least desirable French content, often considered the most culturally valuable by the French government.
regardless of how much energy we do or do not need or use, this would (assuming the theory is correct) provide a clean, renewable energy source, which we can substitute for currently-used unclean or non-renewable energy sources.
Well, heat transfer into the deep ocean is not a clean or a renewable energy source.
Unless you're planning to argue that we should reduce our energy use to what present clean/renewable sources can provide
Yes, that's exactly what I'm arguing.
an argument I don't think you'll find much support for
Not in the US, that's sure. Other nations are actively working towards that with genuinely sustainable and environmentally friendly technologies.
I assume you can see the reason for developing such an alternative.
That may be the reason, but it is neither reasonable nor rational. But humanity has the freedom to choose to go to hell in whatever way it likes.
who is actually paying for innovation
on
McVoy Strikes Back
·
· Score: 5, Insightful
But McVoy says open source advocates fail to recognize that building new software requires lots of trial and error, which means investing lots of money. Software companies won't make those investments unless they can earn a return by selling programs rather than giving them away.
Software companies don't make those investments at all. The institutions that make those investments are the government and a few large private research labs. Almost all the software and almost all the innovation you see around you ultimately comes from those sources.
People like McVoy and other self-proclaimed innovators are adding little gimmicks and tweaks on top of that massive, publicly funded innovation. The question we should be asking is why we should let people like McVoy continue to leech off the investments that taxpayers and a few private labs are making.
LOL and though it was written well before any of this became relevant, the Emperor is Shaddam! On the note of plagiarism:
Reusing themes, ideas, concepts, characters, etc. may be copyright infringement in our new corporate-copyright-overlord-dominated media culture, but it isn't plagiarism.
If such similarities were plagiarism, most great Western literature and art would have to be considered "plagiarized".
Warming deep ocean water may have all sorts of consequences, like changes in ocean currents, releases of nutrients and toxins, and harm to unique ecologies. Doing this on a massive scale seems in advisable.
I also fail to see the point. We just don't need more energy, we need to be more energy efficient and halt population growth.
Of course, they are doing this for business reasons and to hurt Microsoft. But that doesn't diminish the value of their actions: they are still helping an open source platform used by many people, including their own competitors. That is the spirit of open source.
Contrast that with Sun's recent patent grant. Sun is shipping both Solaris and Linux systems, but they grant their patents only for Solaris, not Linux. That's presumably because they are the primary user of Solaris, but their competitors would benefit from anything they do to make Linux more attractive.
A binding, blanket statement that they will not enforce their patents against arbitrary new functionality in future Linux kernel releases could invite abuse: companies could try to circumvent arbitrary Nokia patents by simply putting implementations into the Linux kernel, even if the patent has nothing to do with kernel functionality.
Software patents are an unfortunate fact of life right now. Nokia has done pretty much the best one can under the circumstances. I think one should applaud them for that.
Because this history of the platform completely disregards Cobalt, it is quite wrong about when moving to Linux means.
It disregards Cobalt because it isn't relevant to my statement. Given the history of PalmOS, you had a simple, obvious path towards a highly backwards compatible PalmOS on a Linux kernel. Are you saying that what you are actually creating is going to be less backwards compatible than the obvious and simple solution?
all you would end up with is PalmOS applications sitting in a big, isolated box, running in a single-threaded single-process world.
Some Palm guys I talked to when the story broke seemed to imply that you were considering that option. If you had followed it, you would already be shipping fully backwards compatible devices based on a Linux kernel (which has several advantages in itself), and you could be putting a lean Cobalt-based set of APIs next to that without the burden of backwards compatibility for the new APIs.
That's probably what you meant, but your posting was unclear. It's important to be clear about such things.
(You also made several factual errors. For example. you claimed that Cobalt is a combination of CMS/Linux, PalmOS, and BeOS. But the first version of Cobalt was completed before the acquisition of CMS, so that can't be true. We don't know yet what PalmOS/Linux will be called.)
I have tried SciPy. I could not get it to install (SuSE 9 on x86). [...] What commercial vendors understand is that their software must install easily.
And Matlab is a pain to install on Debian, while installing SciPy on Debian is a single command. Apparently, Mathworks does not understand that their software must install easily.
In any case, for most things, Numeric is sufficient, and last I tried, Numeric installed trivially on SuSE (as well as on Debian, MacOSX, and Windows).
I presume your seat belt analogy refers to type safety,
I stated what the analogy referred to and it was not type safety.
As for your assertion that computer vision can only be realistically done in C++, this is absurd.
I did not make such an assertion (you are having trouble with quantifiers). What is absurd that you generalize from your narrow area to all of computer vision and all of numerical computation.
As my username suggests, I do computer vision research and my dept. uses Matlab, C++ and C. Most new members are encouraged to use Matlab as it has proven to be more productive.
I get the picture just what kind of department you are in. You may do well academically in the short term, in the sense of maximizing papers published and having a department with a good reputation; I don't think you are getting/providing a good education.
Well, yes, it should: since it is X11 based and since KDE and Gnome have worked on standardizing common desktop features between them, Qt and KDE applications should work pretty well on it.
Furthermore, deriving a KDE handheld environment from the KDE desktop software should not be much work.
This is the way handheld Linux software should be done. Qt/Embedded and Qtopia's approach is unnecessarily exclusionary: with Qt/Embedded, you cannot realistically use any non-Qt GUI applications on the handheld.
Slashdot Mirror
This isn't theoretical: numerous operating systems have been implemented in safe, garbage collected languages, including Modula-3, Lisp, and Smalltalk.
.NET are unsuitable implementations for writing a kernel, but with a good native code compiler, Java and C# are perfectly fine languages for implementing a kernel.)
Safety doesn't mean that you can't do unsafe things, it just means that you won't do unsafe things accidentally. C#, for example, has the same pointer manipulation primitives built-in as C.
And garbage collection doesn't have to mean that your program gets interrupted at inconvenient times, while C memory managers make no guarantees and are actually quite unpredictable in practice.
(Sun Java and Microsoft
Put virtual machines on top, like Java and .NET. Claim that they're more secure than the OS.
It's not the virtual machines that make Java and C# applications more secure, it's the fact that the languages are better designed than C and C++.
It's not working.
Sure it is. Java applications have no buffer overflow problems, no low-level memory management bugs, and no undetected type errors.
If the Hurd guys had a clue, and could write something as good as QNX, there might be some hope from that direction. But after ten years of screwing up, there's not much hope there.
Microkernels are the wrong solution: they attempt to compensate for limitations in the programming language used to write the kernel by introducing unnecessary complexity and overhead in other places.
Until kernel designers kick the C/C++ habit, we are not going to get a decent kernel.
It's unfortunate that Java and C# (pretty decent language designs) happen to be implemented with such bloated runtimes by their main proponents. But that's not necessary: Java and C#-like languages can be implemented natively with pretty much the same efficiency as C and C++.
I don't see a "rush" to produce Java-capable products. Java has become a niche language: some server side apps (for people who don't know any better), introductory CS courses (hello, Pascal), and small cell phone gimmicks (that don't work well across phones).
.NET probably has a bigger installed based than Java. And while Microsoft's monopoly gave them an unfair advantage, this is still Sun's mistake: they had a half dozen year lead until Microsoft finally got their act together, and Sun still screwed it up.
It's a shame, but
There is a whole range of enrichment methods; see here for an overview. Note the article on EMIS and use by the Iraqis. Furthermore, the chemical exchange processes sound like something competent chemical engineers could implement. And I suspect that with improvements in materials, computers, and chemical engineering, such processes become cheaper and easier to implement.
I think we'll just have to face the fact that nuclear weapons will become accessible to many more nations over the coming decades.
Well, I think this has been clear for a long time. The Windows XP kernel and many of its key user mode libraries will continue to be written in C/C++. I suspect that Explorer, Internet Explorer, and the control panel will likely continue to remain C/C++ based.
.NET runtime with Longhorn, as well as lots of .NET libraries. I suspect that if you removed the .NET runtime, some applications and system utilities would break, although the system overall would probably still boot. (Didn't SP2 come with a .NET runtime anyway?)
.NET isn't a revolution, it's an evolution. Think of it as a Visual Basic replacement--a better designed runtime with a choice of better designed libraries. And, unlike Visual Basic, .NET may actually be good enough for Microsoft to start writing small applications and system utilities in.
However, it looks like they are going to ship a full
All of that is pretty reasonable. Why break working code? Why alienate thousands of developers? The inclusion of
Supercomputing ? Here you are.
Well, and we have benchmarks for that. That's why I'm asking what its actual performance on real numerical code is, not some theoretical maximums.
It'll run them exactly as fast as any other PPC 970 core. As far as I can see from the information that's been released so far, to use the coprocessors at all you'll need to redesign your application around an asymmetric coarse-grained parallel processing model
Applications are already written for that model: they use APIs like BLAS and LINPACK that hardware vendors can reimplement to take maximum advantage of their parallel capabilities, and they use intrinsically parallelizable language primitives (like Fortran's parallel array primitives). Furthermore, many applications have been parallelized for either multithreaded execution or MPI. In different words, today's high-end numerical applications already contain the hooks and declarations for extensive SIMD and MIMD parallelism.
If the Cell architecture can't provide high performance for those kinds of existing parallel codes, then there is something wrong with Cell and/or its compilers.
If a male has nothing to hide, that's exactly the reason why he is concerned to be seen naked.
2. How do I get put in charge of the 'Hot Chick' section
You need to be a woman, or at least be able to pass for one. In order to find the "chicks" attractive under those circumstances, you need to be lesbian or heterosexual but transgendered, respectively. Oh, and you still need to keep quiet about it because if you are found out, you'll be transfered.
To the author of the article (a "professional" writer?): the proper gerund of "backscatter" is "backscattering"; "backscatting" means something entirely different (musical or scatological, take your pick).
To all the readers, if you want to see thousands of nude bodies parading around (in color), go to a nude beach. I think you'll soon appreciate the custom of wearing clothes, and you'll see that we don't wear clothes to prevent lust and desire, but to create and enhance them. Clothes allow even people with fairly unattractive bodies to look decent; as geeks, we should all appreciate that.
I pity the screeners that have to look at thousands of average human bodies without the protection of clothing.
With Cell, IBM keeps talking about "theoretical GFLOPS". I don't care about theoretical numbers. What I care about is how fast the thing runs when I run normal code compiled with a normal compiler and (possibly hand-optimized) numerical libraries.
So, what kind of SPECfp numbers does the thing get? What kind of BLAS performance does it get?
They have 2.6.11 running on it, so compiling the benchmarks should be trivial. If they haven't published anything yet (I haven't seen it), we have to believe that the numbers are less than impressive.
(Another company used to make inflated claims about the performance of their processors by computing theoretical maximums for a few SIMD instructions, unachievable in most real code. When people actually did some real benchmarks and published them against the wishes of the company, they found that their processor was no faster MHz for MHz than Pentium on real code with real compilers.)
The problem is the registry. UNIX doesn't have it or anything like it, and that's a good thing.
The industry is not taking Barella's statements lightly. In a letter last month addressed to the French Minister of Justice Dominique Perben, more than 20 representatives of France's entertainment, music and film association bodies and advocacy groups expressed their outrage that French content was not pirated to anywhere near the same degree as content in English and even German.
"It is an insult to French Culture that American and even German language music, rap, and books are being shared widely over file sharing networks, while nobody else in the world bothers with French content; even our own teenagers prefer American content, despite decade-long indoctrination in French government facilities." wrote Perben.
Perben proposed criminalizing the distribution of copyrighted content containing lyrics that were not easily recognizable as French, creating a high-speed French government Internet that would facilitate rapid piracy of certified French content, and scanning user's harddisks for foreign content, whether copyrighted or not. He also said the French government was considering financial rewards for particularly active filesharers of the least desirable French content, often considered the most culturally valuable by the French government.
regardless of how much energy we do or do not need or use, this would (assuming the theory is correct) provide a clean, renewable energy source, which we can substitute for currently-used unclean or non-renewable energy sources.
Well, heat transfer into the deep ocean is not a clean or a renewable energy source.
Unless you're planning to argue that we should reduce our energy use to what present clean/renewable sources can provide
Yes, that's exactly what I'm arguing.
an argument I don't think you'll find much support for
Not in the US, that's sure. Other nations are actively working towards that with genuinely sustainable and environmentally friendly technologies.
I assume you can see the reason for developing such an alternative.
That may be the reason, but it is neither reasonable nor rational. But humanity has the freedom to choose to go to hell in whatever way it likes.
But McVoy says open source advocates fail to recognize that building new software requires lots of trial and error, which means investing lots of money. Software companies won't make those investments unless they can earn a return by selling programs rather than giving them away.
Software companies don't make those investments at all. The institutions that make those investments are the government and a few large private research labs. Almost all the software and almost all the innovation you see around you ultimately comes from those sources.
People like McVoy and other self-proclaimed innovators are adding little gimmicks and tweaks on top of that massive, publicly funded innovation. The question we should be asking is why we should let people like McVoy continue to leech off the investments that taxpayers and a few private labs are making.
LOL and though it was written well before any of this became relevant, the Emperor is Shaddam! On the note of plagiarism:
Reusing themes, ideas, concepts, characters, etc. may be copyright infringement in our new corporate-copyright-overlord-dominated media culture, but it isn't plagiarism.
If such similarities were plagiarism, most great Western literature and art would have to be considered "plagiarized".
I have been put on hold by 911 for longer than by my bank while a gunman was waving a gun outside, or when there was a sofa on the highway at night.
Call me cynical, but if you have to rely on 911 service, you're in trouble.
Warming deep ocean water may have all sorts of consequences, like changes in ocean currents, releases of nutrients and toxins, and harm to unique ecologies. Doing this on a massive scale seems in advisable.
I also fail to see the point. We just don't need more energy, we need to be more energy efficient and halt population growth.
Of course, they are doing this for business reasons and to hurt Microsoft. But that doesn't diminish the value of their actions: they are still helping an open source platform used by many people, including their own competitors. That is the spirit of open source.
Contrast that with Sun's recent patent grant. Sun is shipping both Solaris and Linux systems, but they grant their patents only for Solaris, not Linux. That's presumably because they are the primary user of Solaris, but their competitors would benefit from anything they do to make Linux more attractive.
A binding, blanket statement that they will not enforce their patents against arbitrary new functionality in future Linux kernel releases could invite abuse: companies could try to circumvent arbitrary Nokia patents by simply putting implementations into the Linux kernel, even if the patent has nothing to do with kernel functionality.
Software patents are an unfortunate fact of life right now. Nokia has done pretty much the best one can under the circumstances. I think one should applaud them for that.
Because this history of the platform completely disregards Cobalt, it is quite wrong about when moving to Linux means.
It disregards Cobalt because it isn't relevant to my statement. Given the history of PalmOS, you had a simple, obvious path towards a highly backwards compatible PalmOS on a Linux kernel. Are you saying that what you are actually creating is going to be less backwards compatible than the obvious and simple solution?
all you would end up with is PalmOS applications sitting in a big, isolated box, running in a single-threaded single-process world.
Some Palm guys I talked to when the story broke seemed to imply that you were considering that option. If you had followed it, you would already be shipping fully backwards compatible devices based on a Linux kernel (which has several advantages in itself), and you could be putting a lean Cobalt-based set of APIs next to that without the burden of backwards compatibility for the new APIs.
That's probably what you meant, but your posting was unclear. It's important to be clear about such things.
(You also made several factual errors. For example. you claimed that Cobalt is a combination of CMS/Linux, PalmOS, and BeOS. But the first version of Cobalt was completed before the acquisition of CMS, so that can't be true. We don't know yet what PalmOS/Linux will be called.)
I have tried SciPy. I could not get it to install (SuSE 9 on x86). [...] What commercial vendors understand is that their software must install easily.
And Matlab is a pain to install on Debian, while installing SciPy on Debian is a single command. Apparently, Mathworks does not understand that their software must install easily.
In any case, for most things, Numeric is sufficient, and last I tried, Numeric installed trivially on SuSE (as well as on Debian, MacOSX, and Windows).
I presume your seat belt analogy refers to type safety,
I stated what the analogy referred to and it was not type safety.
As for your assertion that computer vision can only be realistically done in C++, this is absurd.
I did not make such an assertion (you are having trouble with quantifiers). What is absurd that you generalize from your narrow area to all of computer vision and all of numerical computation.
As my username suggests, I do computer vision research and my dept. uses Matlab, C++ and C. Most new members are encouraged to use Matlab as it has proven to be more productive.
I get the picture just what kind of department you are in. You may do well academically in the short term, in the sense of maximizing papers published and having a department with a good reputation; I don't think you are getting/providing a good education.
Well, yes, it should: since it is X11 based and since KDE and Gnome have worked on standardizing common desktop features between them, Qt and KDE applications should work pretty well on it.
Furthermore, deriving a KDE handheld environment from the KDE desktop software should not be much work.
This is the way handheld Linux software should be done. Qt/Embedded and Qtopia's approach is unnecessarily exclusionary: with Qt/Embedded, you cannot realistically use any non-Qt GUI applications on the handheld.
It's open source and it's X11 based. That means that RandR probably works on it and you can just rotate the screen 180 degrees.