The entire article is about the fact that MS effectively coded requirement to reserve a certain amount of GPU into the XB1 OS. This amount is not released to the game running unless MS recodes the OS as it appears to be doing
Right, so what's this about consoles not being able to utilise the entire GPU? It seems by 'consoles' you meant the xbox one, and even then that - according to the article - won't be the case for long correct?
It's also fairly interesting that you're suggesting offloading tasks from GPU to CPU, when in general the offloading goes in the opposite direction, both on consoles and in PC world.
Most games these days are GPU bound so why would you offload to a resource that is being so heavily utilised?
Consoles are talked about in the article, and PC games often offload things like physics calculations to GPU as much as possible (PhysX, various CUDA implementations and so on).
Of course PCs offload to the GPU, because the tasks are often more suited to that architecture but the problem is as far as PCs are concerned you end up with a resource (CPU and main memory) that is under-utilised, you cannot effectively use that to support the GPU calculations because unlike consoles you don't know what the cost is to transfer data back and forth, you don't know what capabilities the CPU has (could be many low-clocked cores, few high-clocked cores, etc...), how fast the RAM is, how much CPU cache you have and how much there is so there are too many variables to work with. Console architecture is constant so you don't have to code to abstractions to support wide hardware variances which obviously results in more efficient implementations.
Which doesn't change the fact that modern PCs in game usage can utilize almost entire GPU, while consoles in fact cannot.
Why can't they? Sure in this case of the xbox the game can only use 98% (which I would certainly characterize as "almost entire GPU") and that additional 2% is used for voice processing.
PCs have almost none, and the classic overhead bottleneck of the PC, CPU/RAM does not constrain a significant amount of games any more, as most of them hit the GPU ceiling first.
Of course they hit a GPU ceiling first, an architecture that makes buffer copies between resources so costly naturally results in avoiding buffer copies by overutilizing the GPU and underutilizing the CPU and main memory which is why in PC games the usage of resources is so unbalanced. It is not even worth considering offloading tasks from the GPU to CPU because you have no idea what your latency might be. It is much easier to effectively utilize all of your resources when you know what their capabilities are.
But he is right that console developers target a GPU where they know its architecture, instruction set, memory amount, memory bandwidth, number of ALUs, ALU clocks, etc... (in addition to knowing that about all the other elements in the system that the GPU is connected to) and knowing this means you can develop much more efficient software and optimize for that hardware. On the PC side you don't know any of those things and even if you set a baseline for one of them you cant guarantee that it isnt offset by something else.
For example you are going to take a very different approach to developing software for a high-clocked dual core CPU than you are for a low-clocked 8 core CPU and that's before you start to consider the architectural differences between them and all the other system components.
Count on having to upgrade your gaming console ver the years to keep games running at a decent level too.
Not over its lifetime though, over its lifetime console games get better as devs get more proficient with optimizations and extracting the most out of hardware. On PCs you can't do that because the hardware range is so large so instead they are coded to hardware abstractions and the solution is just to 'throw more hardware at it'.
Meanwhile that same computer that players last year's games will still play last decade's games just fine, and do your taxes on the side, be your development workstation, etc.
I could also get a 3G Nexus 7, use Skype or Viber or whatever to make phone calls, connect a PS3 controller (and possibly HDMI-out) to play games and chroot debian to run all my applications with a connected bluetooth keyboard and mouse so I don't need a tablet, phone, console and laptop, I could just have tablet but it would be a penny-pinching kludge. Now while I could do that I prefer to have dedicated devices for that stuff, it means I don't have to compromise but if a few games here and there work better on my PC than my console I'll play them on my PC.
If what you are interested in is value for money then yes a PC is the way to go because it can do many of the things a console can with a bit of fucking around (connecting it to the TV, messing with keyboard/mouse and then switching to controller or having a dual-boot setup) and much more.
In many cases (as far as gaming systems are concerned) PCs are more powerful than consoles. The issue is that games cannot fully exploit that power because developers have to code to higher level abstractions to deal with the differences between systems, you can't rely on a specific speed of CPU, GPU, RAM, video RAM, bus, etc... and you can't rely on a specific amount of CPU cache, CPU cores, RAM, video RAM, GPU ALUs, etc... you can't rely on a specific architecture (exact supported instruction set) of CPU or GPU or the capabilities of that GPU. Then there is the operating system overhead and the graphics drivers that have to provide schedulers to manage multiple processes accessing GPU at once.
In a console you know all of this - or at least have a baseline - so it is obviously much easier to optimize software for a system when you know exactly what that system is.
That's dependent on this being an actual thing. Because when five people with Apple products have the same problem, it's officially an international incident. Whereas other companies can have crap like the Red Ring of Death and never get a fraction of the hatorade.
What the hell are you talking about? Where have only 5 people had this issue? And the Red Ring of Death is constantly brought up whenever Microsoft products are discussed and it resulted in additional replacement warranties being extended.
Like last year, one of my friends had his Razr Maxx replaced four times for a cracking screen. At first he thought it was because he was being careless, but it turns out there was a design flaw that would crack the screen while charging the phone. Zero articles on Slashdot with anyone raging at how Motorola was a shoddy company.
So what are you suggesting? All Motorola Razr Maxxs had this issue? Maybe only your friend had the issue, or maybe it was only "5 people with Motorola Razr Maxxs" that had this problem.
What auto manufacturer, anywhere, worldwide, at any price point, offers a warranty that long?
Hint: None of them.
Then why did you say you got warranty work done on your car when it was 13 years old? Was it under warranty or wasn't it? Why can't you keep it straight?
In any case what you bought was defective so that's why they fixed it, Windows XP will not be supported/maintained past April 2014 at which point it will still be less than 13 years old. So what you have is a nice anecdote but doesn't really relate to this situation at all.
if the reports are not just accurate but symptomatic of the parts in question, Apple should offer to extend coverage for said parts, then take it out of Nvidia's hide.
Of course they should, so why aren't they? They didn't in 2008 either. These things should be system recalls but ultimately they should be better tested to begin with, and that goes for any manufacturer.
Yes you are correct, the concern with Canonical's CLA is that while they ship GPL software the CLA gives Canonical permission to relicense and distribute your contributions under a proprietary license. The question is why would they need that?
Actually it isn't, we all know the reason malware continues on any platform is because of people that think they know what they are doing but really don't. Even if you aren't installing new software that you've downloaded from unsavory sources your unpatched OS is vulnerable to a variety of drive-by malware, you may think you can "maintain" Windows XP just by defragging and other user tasks but that is not OS maintenance, which is your problem, you don't actually know what maintenance is. The answer for Windows XP-era hardware is to upgrade to a maintained Linux distribution and you can even keep a Windows XP VM if it is absolutely necessary for a particular application.
I'm done regurgitating the same points that you will not read anyway.
No you're not, you keep saying that but you keep posting.
It is nothing to do with executing the code and that is precisely why Tivoization is a "problem" with the GPL v3.
Tivoization was specifically addressed in the GPL with v3, it is versions < v3 that don't address tivoization. But yes as you say installation and execution are the key elements added to v3 that prevent tivoization.
First sentence in the linked post "Contributor License Agreements ("CLAs") are a mechanism for an upstream software developer to insist that contributors grant the upstream developer some additional set of rights." Contrary to your assumption I in fact did not know what CLA stood for in this context, so simply clicked the link to find out.
Yes they could have put that in the summary but sadly had they done that it seems there would have been next to no comments here. I'm sorry if you found some part - feel free to point it out - of the post you found "hateful".
I think it's more to do with adding additional layers of legal crap to free software making contributions to and use of such products overly convoluted. That said I do agree with you that to keep it free the community needs to be united and come to terms with the reality of the society they exist in: it would be nice if everybody worked together and nobody ever acted purely in self-interest but that's not going to happen.
Interesting that on a website that claims to be 'news for nerds' you need to scroll almost to the bottom of the comments to find something that isn't a "I don't know what CLA is, is it (any number of irrelevant or crude possible expansions of the acronym)?", that's pretty pathetic.
I can maintain XP just fine. I have maintained XP just fine for years with no updates.
How exactly? Are you running a stock Windows XP from 2001 with no updates at all? Updates are somehow pointless? It's retards like you who are the reason there is so much malware and so many viruses out there because you don't know the first thing about maintaining a system, leaving it unpatched is completely stupid.
So yes, YOU should upgrade whenever microsoft tells you to. Always.
Wrong, you should upgrade when your OS goes out of support. Or, if you're the sort of person who would rather maintain it yourself, you wouldn't be such a complete mental defective as to have chosen an operating system like Windows that you cannot maintain, you would choose a Linux distribution that is free and open.
Is it really an anecdote when you have owned 10+ Macs and 10+ PCs in your lifetime, and every Mac lived a long time or died suddenly due to a hardware failure (happened to me three times, one being from an externality where an idiot spilled a coke on my laptop), while every PC suffered from slowdown to the point of being unusable after 3-4 years?
If every PC were unusable after 3-4 years then I doubt there would be so many people on Windows XP. But that's beside the point, Macs don't last any longer than PCs because it's the same hardware and if anything they fall out of support much sooner than other PCs. That said it hasn't deterred me from buying Macs, when Apple stops supporting them I'll just put Linux on them.
Of further note is that when the PCs were switched over to Linux, the slowdown issues disappeared.
So it's nothing to do with Macs being any more reliable at all, your complaint is related to Windows.
Question you ask in the bold is clearly answered in the post you posted the previous message as an answer. In its own paragraph no less.
But it's wrong, very obviously wrong. You can continue to maintain your car as long as you want, you cannot maintain Windows XP and Microsoft is not going to maintain Windows XP. What part of that is so hard for you to understand?
Answer the question: What parts are analogous? It's a simple question and if you understand what an analogy is you will be able to answer it, but you failed to do it several times already and you will fail again.
And you made a point to avoid the actual point of analogy - that old computers and old cars that serve their purpose are perfectly fine now, and will likely be perfectly fine five years from now
The obvious reason old Windows XP computers are fine now is because they have been constantly updated, where the car has not. If Windows XP had not been updated it would not be anywhere near suitable these days.
All in all, you behaved fully like a bought and paid for microsoft astroturfer, spreading fear, uncertainty and doubt as to argue that upgrade is mandatory and should happen sooner rather than later. And it's well known that when one's income is tied to denying facts in their face, it's almost impossible to convince them to face the facts.
Wrong again, I never said you should upgrade, not once, much less to another Microsoft OS, so it is clear that you aren't even comprehending the posts and are projecting your failings. The fact is that a Windows XP computer should not be upgraded to a later version of Windows because in most cases it won't be powerful enough since Windows is not written to be particularly efficient, such hardware should be upgraded to a suitable Linux distribution that offers long term support.
So answer the question and stop your rubbish conjecture, you just look stupider and stupider with every post. What parts are analogous?
It's not hair splitting at all, it's simply that the analogy is completely flawed, proven by the fact can't answer any of those questions. The car never had any updates provided by the manufacturer that increased its security or features whereas Windows XP did. You disproved your own analogy in the first sentence of this post and didn't even realise it.
An analogy has to have some analogous points, you stated that the reason you believed the analogy hold was "Car security becomes noticeably better. Cars become easier to drive and more resilient to accidents even if they do occur", but how is this analogous to a Windows XP PC? The car you are talking about wasn't updated by the manufacturer with noticeably better security, or features that made it easier to drive or increased its resilience in a crash. Whereas the Windows XP system was updated by the manufacturer to have better security and features that increased functionality and made it easier to use so clearly it is you who needs to read that link because you obviously don't understand the concept of an analogy.
Your inability to answer the bolded question will demonstrate to you why you don't understand the concept of an analogy.
Really? I've always been shaped when I hit my data cap on broadband, never charged additional for overages. And I've certainly never paid anywhere near the prices they are charging (unless you count the way I used to pay for 'hours' on my old dialup connections, getting 5GB on dialup would take quite a while).
Slashdot Mirror
The entire article is about the fact that MS effectively coded requirement to reserve a certain amount of GPU into the XB1 OS. This amount is not released to the game running unless MS recodes the OS as it appears to be doing
Right, so what's this about consoles not being able to utilise the entire GPU? It seems by 'consoles' you meant the xbox one, and even then that - according to the article - won't be the case for long correct?
It's also fairly interesting that you're suggesting offloading tasks from GPU to CPU, when in general the offloading goes in the opposite direction, both on consoles and in PC world.
Most games these days are GPU bound so why would you offload to a resource that is being so heavily utilised?
Consoles are talked about in the article, and PC games often offload things like physics calculations to GPU as much as possible (PhysX, various CUDA implementations and so on).
Of course PCs offload to the GPU, because the tasks are often more suited to that architecture but the problem is as far as PCs are concerned you end up with a resource (CPU and main memory) that is under-utilised, you cannot effectively use that to support the GPU calculations because unlike consoles you don't know what the cost is to transfer data back and forth, you don't know what capabilities the CPU has (could be many low-clocked cores, few high-clocked cores, etc...), how fast the RAM is, how much CPU cache you have and how much there is so there are too many variables to work with. Console architecture is constant so you don't have to code to abstractions to support wide hardware variances which obviously results in more efficient implementations.
Which doesn't change the fact that modern PCs in game usage can utilize almost entire GPU, while consoles in fact cannot.
Why can't they? Sure in this case of the xbox the game can only use 98% (which I would certainly characterize as "almost entire GPU") and that additional 2% is used for voice processing.
PCs have almost none, and the classic overhead bottleneck of the PC, CPU/RAM does not constrain a significant amount of games any more, as most of them hit the GPU ceiling first.
Of course they hit a GPU ceiling first, an architecture that makes buffer copies between resources so costly naturally results in avoiding buffer copies by overutilizing the GPU and underutilizing the CPU and main memory which is why in PC games the usage of resources is so unbalanced. It is not even worth considering offloading tasks from the GPU to CPU because you have no idea what your latency might be. It is much easier to effectively utilize all of your resources when you know what their capabilities are.
But he is right that console developers target a GPU where they know its architecture, instruction set, memory amount, memory bandwidth, number of ALUs, ALU clocks, etc... (in addition to knowing that about all the other elements in the system that the GPU is connected to) and knowing this means you can develop much more efficient software and optimize for that hardware. On the PC side you don't know any of those things and even if you set a baseline for one of them you cant guarantee that it isnt offset by something else.
For example you are going to take a very different approach to developing software for a high-clocked dual core CPU than you are for a low-clocked 8 core CPU and that's before you start to consider the architectural differences between them and all the other system components.
Count on having to upgrade your gaming console ver the years to keep games running at a decent level too.
Not over its lifetime though, over its lifetime console games get better as devs get more proficient with optimizations and extracting the most out of hardware. On PCs you can't do that because the hardware range is so large so instead they are coded to hardware abstractions and the solution is just to 'throw more hardware at it'.
Meanwhile that same computer that players last year's games will still play last decade's games just fine, and do your taxes on the side, be your development workstation, etc.
I could also get a 3G Nexus 7, use Skype or Viber or whatever to make phone calls, connect a PS3 controller (and possibly HDMI-out) to play games and chroot debian to run all my applications with a connected bluetooth keyboard and mouse so I don't need a tablet, phone, console and laptop, I could just have tablet but it would be a penny-pinching kludge. Now while I could do that I prefer to have dedicated devices for that stuff, it means I don't have to compromise but if a few games here and there work better on my PC than my console I'll play them on my PC.
If what you are interested in is value for money then yes a PC is the way to go because it can do many of the things a console can with a bit of fucking around (connecting it to the TV, messing with keyboard/mouse and then switching to controller or having a dual-boot setup) and much more.
In many cases (as far as gaming systems are concerned) PCs are more powerful than consoles. The issue is that games cannot fully exploit that power because developers have to code to higher level abstractions to deal with the differences between systems, you can't rely on a specific speed of CPU, GPU, RAM, video RAM, bus, etc... and you can't rely on a specific amount of CPU cache, CPU cores, RAM, video RAM, GPU ALUs, etc... you can't rely on a specific architecture (exact supported instruction set) of CPU or GPU or the capabilities of that GPU. Then there is the operating system overhead and the graphics drivers that have to provide schedulers to manage multiple processes accessing GPU at once.
In a console you know all of this - or at least have a baseline - so it is obviously much easier to optimize software for a system when you know exactly what that system is.
Teach us how to punish ourselves to make these big corps feel bad, Obi-Wan.
Yes punish yourself by not buying a new game console!
That's dependent on this being an actual thing. Because when five people with Apple products have the same problem, it's officially an international incident. Whereas other companies can have crap like the Red Ring of Death and never get a fraction of the hatorade.
What the hell are you talking about? Where have only 5 people had this issue? And the Red Ring of Death is constantly brought up whenever Microsoft products are discussed and it resulted in additional replacement warranties being extended.
Like last year, one of my friends had his Razr Maxx replaced four times for a cracking screen. At first he thought it was because he was being careless, but it turns out there was a design flaw that would crack the screen while charging the phone. Zero articles on Slashdot with anyone raging at how Motorola was a shoddy company.
So what are you suggesting? All Motorola Razr Maxxs had this issue? Maybe only your friend had the issue, or maybe it was only "5 people with Motorola Razr Maxxs" that had this problem.
open source != free software
Precisely! Nobody willingly "signs up" to get a virus and the entire time that you have a virus you are trying to get rid of it.
What auto manufacturer, anywhere, worldwide, at any price point, offers a warranty that long?
Hint: None of them.
Then why did you say you got warranty work done on your car when it was 13 years old? Was it under warranty or wasn't it? Why can't you keep it straight?
In any case what you bought was defective so that's why they fixed it, Windows XP will not be supported/maintained past April 2014 at which point it will still be less than 13 years old. So what you have is a nice anecdote but doesn't really relate to this situation at all.
I get your point but I doubt grandma who is installing malware is going to be aware of operating system patches anyway whether they happen or not.
if the reports are not just accurate but symptomatic of the parts in question, Apple should offer to extend coverage for said parts, then take it out of Nvidia's hide.
Of course they should, so why aren't they? They didn't in 2008 either. These things should be system recalls but ultimately they should be better tested to begin with, and that goes for any manufacturer.
Yes you are correct, the concern with Canonical's CLA is that while they ship GPL software the CLA gives Canonical permission to relicense and distribute your contributions under a proprietary license. The question is why would they need that?
Read the thread. It's all in here.
Actually it isn't, we all know the reason malware continues on any platform is because of people that think they know what they are doing but really don't. Even if you aren't installing new software that you've downloaded from unsavory sources your unpatched OS is vulnerable to a variety of drive-by malware, you may think you can "maintain" Windows XP just by defragging and other user tasks but that is not OS maintenance, which is your problem, you don't actually know what maintenance is. The answer for Windows XP-era hardware is to upgrade to a maintained Linux distribution and you can even keep a Windows XP VM if it is absolutely necessary for a particular application.
I'm done regurgitating the same points that you will not read anyway.
No you're not, you keep saying that but you keep posting.
It is nothing to do with executing the code and that is precisely why Tivoization is a "problem" with the GPL v3.
Tivoization was specifically addressed in the GPL with v3, it is versions < v3 that don't address tivoization. But yes as you say installation and execution are the key elements added to v3 that prevent tivoization.
First sentence in the linked post "Contributor License Agreements ("CLAs") are a mechanism for an upstream software developer to insist that contributors grant the upstream developer some additional set of rights." Contrary to your assumption I in fact did not know what CLA stood for in this context, so simply clicked the link to find out.
Yes they could have put that in the summary but sadly had they done that it seems there would have been next to no comments here. I'm sorry if you found some part - feel free to point it out - of the post you found "hateful".
I think it's more to do with adding additional layers of legal crap to free software making contributions to and use of such products overly convoluted. That said I do agree with you that to keep it free the community needs to be united and come to terms with the reality of the society they exist in: it would be nice if everybody worked together and nobody ever acted purely in self-interest but that's not going to happen.
Interesting that on a website that claims to be 'news for nerds' you need to scroll almost to the bottom of the comments to find something that isn't a "I don't know what CLA is, is it (any number of irrelevant or crude possible expansions of the acronym)?", that's pretty pathetic.
I can maintain XP just fine. I have maintained XP just fine for years with no updates.
How exactly? Are you running a stock Windows XP from 2001 with no updates at all? Updates are somehow pointless? It's retards like you who are the reason there is so much malware and so many viruses out there because you don't know the first thing about maintaining a system, leaving it unpatched is completely stupid.
So yes, YOU should upgrade whenever microsoft tells you to. Always.
Wrong, you should upgrade when your OS goes out of support. Or, if you're the sort of person who would rather maintain it yourself, you wouldn't be such a complete mental defective as to have chosen an operating system like Windows that you cannot maintain, you would choose a Linux distribution that is free and open.
Is it really an anecdote when you have owned 10+ Macs and 10+ PCs in your lifetime, and every Mac lived a long time or died suddenly due to a hardware failure (happened to me three times, one being from an externality where an idiot spilled a coke on my laptop), while every PC suffered from slowdown to the point of being unusable after 3-4 years?
If every PC were unusable after 3-4 years then I doubt there would be so many people on Windows XP. But that's beside the point, Macs don't last any longer than PCs because it's the same hardware and if anything they fall out of support much sooner than other PCs. That said it hasn't deterred me from buying Macs, when Apple stops supporting them I'll just put Linux on them.
Of further note is that when the PCs were switched over to Linux, the slowdown issues disappeared.
So it's nothing to do with Macs being any more reliable at all, your complaint is related to Windows.
In security software, shipping nothing is preferable to shipping bugs, especially when shipping a product does not net you any more revenues.
We're talking about OpenBSD and every software product has bugs. Show me a software product that is 100% guaranteed to be bug free.
Question you ask in the bold is clearly answered in the post you posted the previous message as an answer. In its own paragraph no less.
But it's wrong, very obviously wrong. You can continue to maintain your car as long as you want, you cannot maintain Windows XP and Microsoft is not going to maintain Windows XP. What part of that is so hard for you to understand?
Just taking a look at http://en.wikipedia.org/wiki/List_of_iOS_devices I am seeing that the oldest phone supporting the current IOS version is the 4s.
From that page it looks like the 4 is the oldest, supporting the current iOS 7.0.4.
Answer the question:
What parts are analogous? It's a simple question and if you understand what an analogy is you will be able to answer it, but you failed to do it several times already and you will fail again.
And you made a point to avoid the actual point of analogy - that old computers and old cars that serve their purpose are perfectly fine now, and will likely be perfectly fine five years from now
The obvious reason old Windows XP computers are fine now is because they have been constantly updated, where the car has not. If Windows XP had not been updated it would not be anywhere near suitable these days.
All in all, you behaved fully like a bought and paid for microsoft astroturfer, spreading fear, uncertainty and doubt as to argue that upgrade is mandatory and should happen sooner rather than later. And it's well known that when one's income is tied to denying facts in their face, it's almost impossible to convince them to face the facts.
Wrong again, I never said you should upgrade, not once, much less to another Microsoft OS, so it is clear that you aren't even comprehending the posts and are projecting your failings. The fact is that a Windows XP computer should not be upgraded to a later version of Windows because in most cases it won't be powerful enough since Windows is not written to be particularly efficient, such hardware should be upgraded to a suitable Linux distribution that offers long term support.
So answer the question and stop your rubbish conjecture, you just look stupider and stupider with every post. What parts are analogous?
It's not hair splitting at all, it's simply that the analogy is completely flawed, proven by the fact can't answer any of those questions. The car never had any updates provided by the manufacturer that increased its security or features whereas Windows XP did. You disproved your own analogy in the first sentence of this post and didn't even realise it.
An analogy has to have some analogous points, you stated that the reason you believed the analogy hold was "Car security becomes noticeably better. Cars become easier to drive and more resilient to accidents even if they do occur", but how is this analogous to a Windows XP PC? The car you are talking about wasn't updated by the manufacturer with noticeably better security, or features that made it easier to drive or increased its resilience in a crash. Whereas the Windows XP system was updated by the manufacturer to have better security and features that increased functionality and made it easier to use so clearly it is you who needs to read that link because you obviously don't understand the concept of an analogy.
Your inability to answer the bolded question will demonstrate to you why you don't understand the concept of an analogy.
This is the norm for us...
Really? I've always been shaped when I hit my data cap on broadband, never charged additional for overages. And I've certainly never paid anywhere near the prices they are charging (unless you count the way I used to pay for 'hours' on my old dialup connections, getting 5GB on dialup would take quite a while).