No, it's not a deliberate torpedo unless it specifically targets that specific CPU.
Let me ask you a question. What if, next week, AMD or VIA or anyone else, released a chip that claimed SSE compatibility but had a flaky implementation. There would be no way for code compiled earlier to know about this flaky implementation and avoid it. Intel is making sure that they only support Intel CPU's with their optmizations because they can't control anyone else.
While i'm not saying it's a great justificaiton, it is one that (not matter how shaky) holds some credence.
All intel has to do is prove a valid technical reason for this, and I can think of a lot of technical reasons that would stand up. The fact that it's beneficial to them as well is the holy grail: A technical AND business reason to limit your competition.
It's not like Intel's compiler is the only compiler out there either. It's not like it's even the dominant compiler in the market. I can't see how any court would compel them to optimize for AMD's chips.
Apart from the fact that the Intel and AMD chips aren't 100% binary compatible (There are features in both that don't exist in the other, though their overlap is in the high 90%), just being binary compatible doesn't make them functionally compatible.
Things like pipeline length and branch prediction are different for both processors, which means that if you order the instructions to optimize performance for one, it's possible that you're creating a pipeline stall that decreases performance in the other.
While Intel *COULD* do extra work to make sure their compiler optimizes code for every processor out there, why should thet?
No, a deliberate torpedo is when you test specifically for an AMD chip, then execute code specifically for the AMD chip to make it run slower. What appears to be the case is that Intel is only optmizing for their chips.
Not really. Different CPU's have different word and dword boundary restrictions. I don't see any way to write code that guarantees it will work on all processors without writing it like they did.
I'm just saying that there's a big difference between intentionally writing code that ONLY the AMD processor runs that's very poor, and writing code for everyone but Intel's processor that is poor because it's the only way to write the default code to be certain it will work today and in the future.
That's still not a deliberate torpedo. That's optimizing only for your CPU and not optimizing for everyone else. It would be different if they specifically checked for the AMD processor flag and executed special (very poor) code than that of generic processors.
While that is what AMD are stating, their complaint seems to be missing some information. It states that if the compiler detects and AMD processor it does something different. It doesn't mention what it does when it detects neither a non-amd processor or an intel processor (Cyrix, Transmeta, Via, etc..) and whether that action is the exact same as when it detects and AMD processor (in other words, the complaint doesn't say if it's merely detecting intel processors and then detecting "everyone else").
I'd really like to see the executed code for three scenarios. Intel, AMD, and VIA and see if the VIA code path is the same as the AMD one.
But is it really? Or is it only using optimized code on Intel chips, and using non-optimized code on everything else (which happens to include AMD chips)?
There's actually a very big distinction. One is a deliberate torpedo, the other is merely optmizing only for their own platform.
Compiler optimization is a tricky business, and optimizing for one platform can actually slow performance on another.
The story is light on details and doesn't say if the compiler is generating code optmized for the P4 or if it's code supposedly optimized for the AMD or if it was one of those "blended" things. If it's optimized for P4, then I can easily see how intel's instruction ordering can be beneficial for them, and slow the AMD.
Things like pipeline length and differing branch predictors can cause wildly different results on different CPU's.
I really don't see how Intel is under any obligation to optimize their compiler for AMD.
I don't really see how Intel can be responsible for optimizing their compiler to their competition. Lack of optimization (and SSE is an optimization) is not the same thing deliberately torpedoing (which they might be doing as well, but I haven't seen any evidence of that).
I don't see how Intel has any compunction to optimize for non-intel processors, though certainly it would make their compiler better if they did.
Theoretically, you're correct. The problem is that the FBI can get a wiretap on ANYONE, just by shopping it around to the right judges. One of the provisions of the the patriot act was that the judge need not even be in the same jurisdiction as the person they want to wiretap. That means, they can go to as many judges as they want until they find one that agrees with them.
Wiretapping laws, in general, are not a bad thing. What's bad is that the enforcers have almost no effective restrictions on their use.
It's not just compatibility, though that's also a big issue. The problem is that the compilers for the Itanium just aren't that mature. It's the same reason the PPC sucks so bad on a lot of benchmarks.
Hand optimized assembly will give you screaming fast results. Unfortunately, you can't build modern applications that way and you end up having to rely on the compiler to optimize for you. On the x86, the compilers are amazingly efficient these days by contrast.
If you've got a 64 bit database, and a 64 bit OS and a 64 bit middleware, what more do you need? You don't need to run photoshop on it. Compatibility is only marginally an issue on servers.
I'm kind of curious. What exactly did you expect from a screenshot anyways? A screenshot won't show how the system behaves. It won't show non-GUI improvements. It won't show a communication layer.
While I 100% encourage people to make their opinion on this known, I think there really isn't enough information on just what MS plans to do with claria.
There are any number of scenario's, some good, some bad, as to what MS will do. They could be buying them to shut them down (and don't want to tip their cards before the purchase, thus we have their silence on the matter). Or they could be buying them for some other technology they hold, and don't plan to continue their business (or sell it off without said technology to someone else).
I think it's a bit premature to be claiming the sky is falling, especially in light of the fact that other spyware vendors seem to reclassify vendors all the time:
"However, despite persistent rumors to the contrary, Mosaic was never released as open source software during its brief reign as a major browser; there were always constraints on permissible uses without payment."
OS/2 didn't have that feature at all. It didn't even HAVE a menuing system for program selection.
Also, you can disable drag and drop if you're really that concerned about it. I've never seen anyone do what you suggest, so that leads me to believe that it's not as common as you would suggest.
Also, personalized menus are not the default configuration for XP.
You're just trying to confuse the issue. How the menus are laid out doesn't effect how they FUNCTION. You can change the layout. You can't (easily) change the way they work. Arguing that one layout is better than the other is PREFERENCE, not fact. Nor is it innovation.
However, my claiming that the Start menu is more FUNCTIONAL than the KDE menu *IS* fact. It does a whole lot more. It's object oriented, and allows easy and simple reconfiguration with drag and drop. Neither KDE, Gnome, or any other Window Manager/Desktop Environment do this. Nor does MacOS (any version). That makes it innovative. Just because you may not like that feature doesn't mean it's not.
Hardware doesn't count because we're talking about innovation in the SOFTWARE industry, and who does and doesn't innovate in it. Software is now a mature market, while nano-technology is an infant one. Infant markets provide monstrous opportunity for innovation while mature ones do not, largely because all the easy ideas (not that carbon nanotubes are easy, but relative to the ease of the invention itself) have been done already, and it's harder and harder to find ideas that haven't already been done. Stop trying to confuse the issue by bringin in non-software examples.
If Microsoft is such a crappy company because it doesn't innovate in SOFTWARE, then explain why nobody else is either. Including FOSS.
Btw, you do realize that Mozilla has it's own set of security zones, right? For example, the extension whitelist is a security zone. So if security zones are only for ActiveX, why does Mozilla implement them?
How precisely do you plan to install all that other software that doesn't ship with your OS?
The original poster said that the *ONLY* difference between the system was a regsistry key. That's patently false in many ways that anyone with 2 brain cells to rub together could see.
Is Fedora Core and Red Hat Advanced Server the same thing with only a configuration file difference? No, it's the software that comes with it that makes the difference.
You're not quite getting my point. True paradigm changing innovation almost never happens today. You are having a hell of a time trying to come up with even miniscule things "countless things" indeed, considering you can't even come up with one that IBM has come up with (hardware doesn't count, we're talking about software here).
Companies like 3M are constantly talking about innovation. It's because it's a culture thing, it's not a bragging thing. You keep telling your employees about innovation and they start doing it.
As for the start menu, you're not quite getting it. KDE has the same problem many other GUI's have. They only allow you to perform Menu operations in the menu. The Start menu is object oriented. You put an object on the menu, and you can right click and do any operation that object supports (such as My Computer, or Network Places, etc..). You also can't drag program onto the KDE menu or reposition them via drag and drop. Just because you can do KDE operations on objects in the menu is not anywhere near the same thing.
Yes, a menu with programs in it is no big deal. It's what that menu allows you to do that makes it interesting.
Not even close. 1 & 2 were examples of IBM being the first to commercialize technology developed largely in academic realms. Virtual Memory was invented (depending on who you talk to) at either The University of Manchestor or MIT. And the term "virtual machine" is pretty overloaded. If you're referring to VM Hypervizors, like IBM's VM/XXX architecture, then perhaps, but if you mean Virtual Machine such as the JVM, then no way.
Microsoft, like many other companies "innovate" every day. It's just things that aren't sexy, or even visible to most people.
Think about something like the Start menu in XP. No, nothing in it was new, but the way the designed it was rather innovative. You can right click on anything in the menu (something i've not seen in *ANY* other menuing system), drag and drop things from the menu (again, nothing i've ever seen in a menu before, short of NeXT's "tear off menu's).
Why are these simple concepts so hard for even Linux desktop environment developers to accomplish? Obviously it's not something "easy" to do or someone else would have done it.
However, these small and almost insiginficant "innovations" aren't things that people think about when they say "What has so-and-so innovated?"
I used to work for 3M, who use "innovation" as a culture. To them, "innovation" means new ideas. Post it notes were an innovation, but they weren't anything new.. it was just a new way to think about the problem.
Integrating the browser into the Desktop was an innovation. Something that many others, including KDE are following. Yet again, it's not anything that was really new, just a new way of thinking about things.
And while MS does indeed buy into technologies, the end result is seldom anything like what they originally bought. I doubt anyone can look at Spyglass Mosaic and then look at IE6 and say they're anywhere near similar, other than their very basic function.
I think you are seriously underestimating MS if you believe them to be incapable of innovation. It's the little stuff that makes it all happen.
What the hell are you talking about? You seem to forget that NT was released in 1993. It had flat memory addressing, and other OS's before it did as well.
Loadable modules weren't new either. DOS had loadable modules for crying out loud. NT had them as well from the beginning (1993).
Actually, apple didn't invent the iPod. It was invented by Tony Fadell, who had already shopped it to Philips and Real, but were rejected. If either of those companies had jumped on it, Apple would not have been credited with the iPod. How does that make Apple innovative?
Also, hard drive based media players had existed before the iPod, so it wasn't new. It just had some innovative features.
I'm similarly skeptical about your claims on the trackpad, and most of your other bullet points are not examples of innovation.
Slashdot Mirror
No, it's not a deliberate torpedo unless it specifically targets that specific CPU.
Let me ask you a question. What if, next week, AMD or VIA or anyone else, released a chip that claimed SSE compatibility but had a flaky implementation. There would be no way for code compiled earlier to know about this flaky implementation and avoid it. Intel is making sure that they only support Intel CPU's with their optmizations because they can't control anyone else.
While i'm not saying it's a great justificaiton, it is one that (not matter how shaky) holds some credence.
All intel has to do is prove a valid technical reason for this, and I can think of a lot of technical reasons that would stand up. The fact that it's beneficial to them as well is the holy grail: A technical AND business reason to limit your competition.
It's not like Intel's compiler is the only compiler out there either. It's not like it's even the dominant compiler in the market. I can't see how any court would compel them to optimize for AMD's chips.
Apart from the fact that the Intel and AMD chips aren't 100% binary compatible (There are features in both that don't exist in the other, though their overlap is in the high 90%), just being binary compatible doesn't make them functionally compatible.
Things like pipeline length and branch prediction are different for both processors, which means that if you order the instructions to optimize performance for one, it's possible that you're creating a pipeline stall that decreases performance in the other.
While Intel *COULD* do extra work to make sure their compiler optimizes code for every processor out there, why should thet?
No, a deliberate torpedo is when you test specifically for an AMD chip, then execute code specifically for the AMD chip to make it run slower. What appears to be the case is that Intel is only optmizing for their chips.
That's a different thing.
Not really. Different CPU's have different word and dword boundary restrictions. I don't see any way to write code that guarantees it will work on all processors without writing it like they did.
I'm just saying that there's a big difference between intentionally writing code that ONLY the AMD processor runs that's very poor, and writing code for everyone but Intel's processor that is poor because it's the only way to write the default code to be certain it will work today and in the future.
That's still not a deliberate torpedo. That's optimizing only for your CPU and not optimizing for everyone else. It would be different if they specifically checked for the AMD processor flag and executed special (very poor) code than that of generic processors.
While that is what AMD are stating, their complaint seems to be missing some information. It states that if the compiler detects and AMD processor it does something different. It doesn't mention what it does when it detects neither a non-amd processor or an intel processor (Cyrix, Transmeta, Via, etc..) and whether that action is the exact same as when it detects and AMD processor (in other words, the complaint doesn't say if it's merely detecting intel processors and then detecting "everyone else").
I'd really like to see the executed code for three scenarios. Intel, AMD, and VIA and see if the VIA code path is the same as the AMD one.
But is it really? Or is it only using optimized code on Intel chips, and using non-optimized code on everything else (which happens to include AMD chips)?
There's actually a very big distinction. One is a deliberate torpedo, the other is merely optmizing only for their own platform.
Compiler optimization is a tricky business, and optimizing for one platform can actually slow performance on another.
The story is light on details and doesn't say if the compiler is generating code optmized for the P4 or if it's code supposedly optimized for the AMD or if it was one of those "blended" things. If it's optimized for P4, then I can easily see how intel's instruction ordering can be beneficial for them, and slow the AMD.
Things like pipeline length and differing branch predictors can cause wildly different results on different CPU's.
I really don't see how Intel is under any obligation to optimize their compiler for AMD.
I don't really see how Intel can be responsible for optimizing their compiler to their competition. Lack of optimization (and SSE is an optimization) is not the same thing deliberately torpedoing (which they might be doing as well, but I haven't seen any evidence of that).
I don't see how Intel has any compunction to optimize for non-intel processors, though certainly it would make their compiler better if they did.
Theoretically, you're correct. The problem is that the FBI can get a wiretap on ANYONE, just by shopping it around to the right judges. One of the provisions of the the patriot act was that the judge need not even be in the same jurisdiction as the person they want to wiretap. That means, they can go to as many judges as they want until they find one that agrees with them.
Wiretapping laws, in general, are not a bad thing. What's bad is that the enforcers have almost no effective restrictions on their use.
It's not just compatibility, though that's also a big issue. The problem is that the compilers for the Itanium just aren't that mature. It's the same reason the PPC sucks so bad on a lot of benchmarks.
Hand optimized assembly will give you screaming fast results. Unfortunately, you can't build modern applications that way and you end up having to rely on the compiler to optimize for you. On the x86, the compilers are amazingly efficient these days by contrast.
If you've got a 64 bit database, and a 64 bit OS and a 64 bit middleware, what more do you need? You don't need to run photoshop on it. Compatibility is only marginally an issue on servers.
Funny that:
/ /www.flexbeta.net
http://toolbar.netcraft.com/site_report?url=http:
OS: Linux
Webserver: Apache 1.3.33
I'm kind of curious. What exactly did you expect from a screenshot anyways? A screenshot won't show how the system behaves. It won't show non-GUI improvements. It won't show a communication layer.
While I 100% encourage people to make their opinion on this known, I think there really isn't enough information on just what MS plans to do with claria.
There are any number of scenario's, some good, some bad, as to what MS will do. They could be buying them to shut them down (and don't want to tip their cards before the purchase, thus we have their silence on the matter). Or they could be buying them for some other technology they hold, and don't plan to continue their business (or sell it off without said technology to someone else).
I think it's a bit premature to be claiming the sky is falling, especially in light of the fact that other spyware vendors seem to reclassify vendors all the time:
http://www.benedelman.org/spyware/threats/
No, it wasn't. I saw the posting when there were 0 comments and it clearly said it was a content management system.
Actually, no. I didn't know that. And neither do you.
http://en.wikipedia.org/wiki/Mosaic_(web_browser)
"However, despite persistent rumors to the contrary, Mosaic was never released as open source software during its brief reign as a major browser; there were always constraints on permissible uses without payment."
I gave a ton of examples already. Simply saying I didn't is a bit orwellian, don't you think? Read up a few levels.
OS/2 didn't have that feature at all. It didn't even HAVE a menuing system for program selection.
Also, you can disable drag and drop if you're really that concerned about it. I've never seen anyone do what you suggest, so that leads me to believe that it's not as common as you would suggest.
Also, personalized menus are not the default configuration for XP.
You're just trying to confuse the issue. How the menus are laid out doesn't effect how they FUNCTION. You can change the layout. You can't (easily) change the way they work. Arguing that one layout is better than the other is PREFERENCE, not fact. Nor is it innovation.
However, my claiming that the Start menu is more FUNCTIONAL than the KDE menu *IS* fact. It does a whole lot more. It's object oriented, and allows easy and simple reconfiguration with drag and drop. Neither KDE, Gnome, or any other Window Manager/Desktop Environment do this. Nor does MacOS (any version). That makes it innovative. Just because you may not like that feature doesn't mean it's not.
Hardware doesn't count because we're talking about innovation in the SOFTWARE industry, and who does and doesn't innovate in it. Software is now a mature market, while nano-technology is an infant one. Infant markets provide monstrous opportunity for innovation while mature ones do not, largely because all the easy ideas (not that carbon nanotubes are easy, but relative to the ease of the invention itself) have been done already, and it's harder and harder to find ideas that haven't already been done. Stop trying to confuse the issue by bringin in non-software examples.
If Microsoft is such a crappy company because it doesn't innovate in SOFTWARE, then explain why nobody else is either. Including FOSS.
Btw, you do realize that Mozilla has it's own set of security zones, right? For example, the extension whitelist is a security zone. So if security zones are only for ActiveX, why does Mozilla implement them?
How precisely do you plan to install all that other software that doesn't ship with your OS?
The original poster said that the *ONLY* difference between the system was a regsistry key. That's patently false in many ways that anyone with 2 brain cells to rub together could see.
Is Fedora Core and Red Hat Advanced Server the same thing with only a configuration file difference? No, it's the software that comes with it that makes the difference.
You're not quite getting my point. True paradigm changing innovation almost never happens today. You are having a hell of a time trying to come up with even miniscule things "countless things" indeed, considering you can't even come up with one that IBM has come up with (hardware doesn't count, we're talking about software here).
Companies like 3M are constantly talking about innovation. It's because it's a culture thing, it's not a bragging thing. You keep telling your employees about innovation and they start doing it.
As for the start menu, you're not quite getting it. KDE has the same problem many other GUI's have. They only allow you to perform Menu operations in the menu. The Start menu is object oriented. You put an object on the menu, and you can right click and do any operation that object supports (such as My Computer, or Network Places, etc..). You also can't drag program onto the KDE menu or reposition them via drag and drop. Just because you can do KDE operations on objects in the menu is not anywhere near the same thing.
Yes, a menu with programs in it is no big deal. It's what that menu allows you to do that makes it interesting.
Not even close. 1 & 2 were examples of IBM being the first to commercialize technology developed largely in academic realms. Virtual Memory was invented (depending on who you talk to) at either The University of Manchestor or MIT. And the term "virtual machine" is pretty overloaded. If you're referring to VM Hypervizors, like IBM's VM/XXX architecture, then perhaps, but if you mean Virtual Machine such as the JVM, then no way.
Microsoft, like many other companies "innovate" every day. It's just things that aren't sexy, or even visible to most people.
Think about something like the Start menu in XP. No, nothing in it was new, but the way the designed it was rather innovative. You can right click on anything in the menu (something i've not seen in *ANY* other menuing system), drag and drop things from the menu (again, nothing i've ever seen in a menu before, short of NeXT's "tear off menu's).
Why are these simple concepts so hard for even Linux desktop environment developers to accomplish? Obviously it's not something "easy" to do or someone else would have done it.
However, these small and almost insiginficant "innovations" aren't things that people think about when they say "What has so-and-so innovated?"
I used to work for 3M, who use "innovation" as a culture. To them, "innovation" means new ideas. Post it notes were an innovation, but they weren't anything new.. it was just a new way to think about the problem.
Integrating the browser into the Desktop was an innovation. Something that many others, including KDE are following. Yet again, it's not anything that was really new, just a new way of thinking about things.
And while MS does indeed buy into technologies, the end result is seldom anything like what they originally bought. I doubt anyone can look at Spyglass Mosaic and then look at IE6 and say they're anywhere near similar, other than their very basic function.
I think you are seriously underestimating MS if you believe them to be incapable of innovation. It's the little stuff that makes it all happen.
What the hell are you talking about? You seem to forget that NT was released in 1993. It had flat memory addressing, and other OS's before it did as well.
Loadable modules weren't new either. DOS had loadable modules for crying out loud. NT had them as well from the beginning (1993).
Emaacs is a great OS, but a horrible text editor.
Actually, apple didn't invent the iPod. It was invented by Tony Fadell, who had already shopped it to Philips and Real, but were rejected. If either of those companies had jumped on it, Apple would not have been credited with the iPod. How does that make Apple innovative?
Also, hard drive based media players had existed before the iPod, so it wasn't new. It just had some innovative features.
I'm similarly skeptical about your claims on the trackpad, and most of your other bullet points are not examples of innovation.