Did you read that article? Huge chunks of it are nonsense. For example that flaws aren't as bad... what hackers need is write access to the file system if they get that they get everything else. And those are still being found. He gives an example of FreeBSD as a place where these security flaws don't happen, which is of course nonsense as HeartBeat shows quite well.
The power to sue comes from congress. Congress has not granted the people broad tort powers over the government. Rather they have offered specific torts for specific types of relief. There is no special rule here. The special rule is when you can sue the federal government.
There are quite a few rumors / gossip she is gay or in a lesbian relationship herself. She couldn't break with Bush's anti-gay agenda but she advocated respect and has come out in favor of civil unions. So consider he mildly supportive of gay rights and not bad at all for a Republican.
So, Microsoft, live with the consequences of your greed and offer free upgrades, it's not going to hurt the bottom line - hell, it might even prop up those dismal sales figures for Win8.
They included Windows XP mode in Windows 7. Windows 7 upgrades are still for sale. Windows 8 includes a Hypervisor which will run licensed XP just fine.
I often imagine the goodwill that MS would generate if they offered all Win8 users a free "downgrade" to win7.
They have good strategic reasons to move their conservative customer base to Windows 8. I wish if anything they were more aggressive.
Think about that for a second. The kinds of people who have been on the same OS for well over a decade and are cheap enough not to upgrade their machines for years.... how are they going to feel about Apple's higher hardware costs, culture of upgrades every year and applications at best supporting OSes from 2 years back?
Microsoft never promised forever support. This infrastructure was sold with an assumption of regular required upgrades. More like a waterheater than waterpipes.
10 years ago when cross platform libraries (like, say, Qt) were virtually unheard of.
10 year ago Qt was being used for major projects. I was using it in 2001 for security related projects it was so solid. KDE an entire GUI based on Qt was founded in 1996, and on KDE 3 by 2004. GTK+ was 1997. wxWidgets is even older, 1992. Open Interface dates to 1991. MKS which was quite good is from 1994.
As for being dependent on XP I don't see how that happened because of libraries. I can understand the TCO argument for Windows but that doesn't say anything about XP.
What are you talking about? iPhone users are more likely to be Republican than Democrat. They are more likely to be social liberals but that has nothing to do with economic policy. There is nothing hypocritical from decrying the economic divide regardless of whether you are benefiting from it. The hypocrisy would be to claim to decry it in general and then to rebel against most all specific policies to correct it.
I read your post. I guess in some truly vague sense I could locate a pool of memory at both points and including a synchronization mechanism in them. But if everything expects latencies in the nanosecond range with gigabytes of bandwidth per second and I move to latency in the millisecond range with bandwidth of megabytes per second things will slow down a lot. 5-10ms of jitter is highly noticeable and disturbing. An application going from 2ms to 2seconds of jitter is going to go from subconsciously uncomfortable to completely uninterpretable by the human brain. X11 does a lot to compensate for differences in performance. None of that is there with just a naive buffer.
As for the rest, about what you are or are not going to use ultimately unless you are networking on a LAN I think the networking in Wayland will be so superior there will be no holdups. So say by 2020 you will be using Wayland. But Wayland's first use case is completely local getting performance for things like video and games up when both the server and client are on the same machine.
More generally though. End users aren't going to really vote on X11 vs. Wayland. X11 on Wayland works fine so there is going to be no good reason for the clients not to be on Wayland. I'm sure there is going to be some stubborn group of people who insist on X11, but ultimately the hardware aren't going to support X11, nor are the distributions going to focus on it so they'll mostly get moved in a few years after the main body of Linux users move regardless. The choice is at the application writer side. The applications are going to have to decide if they want to implement X11 or Wayland. If they pick Wayland then there is no X11 networking: you either use the Wayland networking, you don't network or you pick a different application. And frankly the advantages for local (and WAN networking) are going to be so huge I can't see application designers making the X11 choice. Plus as I mentioned in the other thread, its possible the toolkits will move forcing the applications. So I don't really see it as optional.
This is like any other upgrade. Some people resist for a while but eventually everyone goes. I remember people holding on to XyWrite on dot matrix printers well into the 1990s, but there are probably 0 doing it still today. Sure there will be some group of people who try and maintain X11 after Wayland becomes normative. Sure there will be distributions focused on X11. But honestly how passionate do you see those people 5 years later? Do you really see people sticking with making those distributions for years out of nothing but spite?
They are not following the X11 model. From a networking perspective there is not going to be a "native Wayland app" but rather a "native KDE app" or a "native Gnome app". The connection is going to be KDE application to client's KDE library with Wayland providing hooks on both sides. Wayland cannot possible know enough about the applications to make intelligent choices about things like caching and prioritization, while the client's GUI libraries can do that. Which is exactly how RDP works.
Now you are absolutely right that the only one existence is the Wayland model. Last year that one wasn't existent. Both KDE and Gnome have announced they are going to start working the Wayland model into their systems. My guess is this takes several years and probably a really good implementation will require breaking changes and so will be done for the next major release (Gnome 4, KDE 5). As they start seriously moving in this direction Gnome 4 / KDE 5 will have to make a choice about whether they are Wayland primarily with some level of X11 support or whether they are X11 with some Wayland advanced features. I suspect Gnome having already decided touch (where getting down to 1ms latency is incredibly important) is key will definitely make the change.
So no. This is not going to be naive networking. This is going to be intelligent networking where the vast majority of applications / GUIs / toolkits will know they are talking over a network and will be expected to compensate and respond intelligently to the additional complexity. This is not "network transparency" because it is not transparent to the application / toolkit. Now there may be some base Wayland way of doing things, using the built in RDP structures entirely but that's going to suck. Possibly it will suck a little less than X11 over WAN sucks because there will be less roundtrips and more happening locally, but that's not the final desired result. Possibly it will suck a little more because unlike the X11 case no one is going to care too much about making it work well, so that while in theory it will be better that may only be in theory. The goal is to implement RDP for Linux and that requires the cooperation of parts higher than Wayland on the stack.
Application designers are quite likely going to spend time coming up with application, data and window caching and prioritization strategies that make sense for their application. GUI designers are quite likely going to spend time coming up with application, data and window caching and prioritization strategies that make sense for their application. And those strategies will change as use cases change. It is entirely possible for example that a local wired connection with a latency under 20ms will do things entirely different than an international or cellular connection with latencies around 150ms. And if you think about it makes sense to push the intelligence as far up the chain as possible. The specific runtime can make better choices than the application and compile time can make better choices than the GUI can make better choices than Wayland.
The compositor is just going to provide the most basic level the hooks for the GUI implementation. This isn't like X11 where X11 is going to do most of the heavy lifting for networking. The RDP is mostly going to be at the GUI level with the Wayland having hooks.
That's not true. The RDP can be individual windows or even less, it is up the the widget set. How KDE or Gnome implement it will be up to them not the Wayland team.
Networks are getting faster and lower latency every day.
What? They are getting higher latency every day. We are increasing hops by doing things like wireless or bouncing off bluetooth. Moreover fundamentally we can't make the planet smaller of the speed of light faster. So there is nothing in the near tern that's likely to bring X11 latencies down to acceptable levels for WAN.
And in the other direction, screens are getting bigger (though OpenGL offers a workaround here) faster than memory is getting faster. So pushing windows buffers around of multiple megabytes is stressing performance for local.
You can't add network transparency later. It is a fundamental design choice. If application and display buffers are shared then lots of things simplify and you get performance improvements but transparency is out the window. That's why you can't have transparency under GDI, Aqua or Aero for example.
Networking on Wayland is mostly like Windows RDP. Which works fantastically over WANs. So local is an upgrade over X11, WAN is an upgrade over X11 and LAN is a downgrade. That's a reasonable tradeoff given today and likely future usage.
An original work that happened to be released under GPLv2. There is some question whether any code was ever accepted into Waze outside of that owned by the group of people that signed all rights over to Google (i.e. if Google owned every line of Wave version 2). According to Google they unquestionably own ever line of Wave v3.
So yes you are understanding this properly and IMHO Roey Gorodish is going to lose fast.
They are claiming the particular data was licensed under GPLv2. However the copyright entity Freemaps have never claimed that Waze's proprietary version was a violation. Waze claimed to have a separate license and they've never disputed this claim.
Waze owned copyright on most (if not all of the code) they weren't a licensee. Which means they can relicense at will. Now of course any code released under the GPLv2 by them is still under GPLv2 but that's it.
Exactly. And just to add. Even if Waze did accept submission for version 2 that are still in version 3 . Roey Gorodish would only have standing if it were his code. This would be a copyright violation not a license violation and since he doesn't hold copyright....
I'm sure Waze had a proprietary license for the data they clearly said over and over they did.
This seems like a poorly thought out case. . Roey Gorodish loses quick IMHO.
The article linked is really quite good. According to the article are 2 prongs to the suit the code and the data.
Waze version 2 was GPLv2. Wave version 3 was proprietary. Waze claimed it rewrote the code. Roey Gorodish wants to examine the code to prove there was a violation but Israeli law doesn't allow discovery without some evidence. As an aside I also have question in that as as I know if Waze owns the code they can multilicense it. Roey Gorodish IMHO would have to find his code not just any code.
There is also the question of map data. Freemap’s data was released both proprietary and GPL and then only proprietary. I'm not sure what the basis of the claim is.
It's the same question. For a start, you haven't demonstrated that it is even possible to shift the bulk of the worlds cereal crop production from the temperate zones to the polar zones. you haven't accounted for the fact that plants actually need soil
Under those fields up north there is soil. Besides we know how to repair farm territory. Cereals moved rather successfully thousands of years ago we are better now. I can't demonstrate what people 300 years from now will be able to do.
But more pertinently you haven't demonstrated how this plan could possibly be cheaper than replacing the remnant fossil fuel generation equipment, and adopting cleaner technologies for transport. Especially given that we would likely adopt them anyway.
That isn't a point of dispute. I don't disagree with you.
People are going to farm if it makes economic sense to do so. Farms aren't charities. If it becomes, on average, more difficult to farm, prices for food will go up -> more people starve. It's a simple but brutal model.
I don't think that's true. Take the current USA standard of living. If raw commodity food prices were to double there would be little to no change in caloric intake in the USA. Food as a percentage of income has been falling rapidly, which has induced shifts to more meat and even still its falling off. In a world where everyone's income is 380x higher than today (i.e. even 2% growth) food prices wouldn't matter much. The raw price of growing food even for very poor people would be a minor expense. Transportation and distribution might account for close to 100% of the food costs.
Slashdot Mirror
Did you read that article? Huge chunks of it are nonsense. For example that flaws aren't as bad... what hackers need is write access to the file system if they get that they get everything else. And those are still being found. He gives an example of FreeBSD as a place where these security flaws don't happen, which is of course nonsense as HeartBeat shows quite well.
It is not exactly secret. The published dates are:
Windows 7 * January 14, 2020
Windows 8 January 10, 2023
The power to sue comes from congress. Congress has not granted the people broad tort powers over the government. Rather they have offered specific torts for specific types of relief. There is no special rule here. The special rule is when you can sue the federal government.
gay = homosexual inclination
lesbian relationship = currently engaged
One is a question of status the other a question of actions.
There are quite a few rumors / gossip she is gay or in a lesbian relationship herself. She couldn't break with Bush's anti-gay agenda but she advocated respect and has come out in favor of civil unions. So consider he mildly supportive of gay rights and not bad at all for a Republican.
They seem pretty clear to me: http://windows.microsoft.com/e...
How is a free upgrade a sale?
They included Windows XP mode in Windows 7. Windows 7 upgrades are still for sale. Windows 8 includes a Hypervisor which will run licensed XP just fine.
They have good strategic reasons to move their conservative customer base to Windows 8. I wish if anything they were more aggressive.
Think about that for a second. The kinds of people who have been on the same OS for well over a decade and are cheap enough not to upgrade their machines for years.... how are they going to feel about Apple's higher hardware costs, culture of upgrades every year and applications at best supporting OSes from 2 years back?
Microsoft never promised forever support. This infrastructure was sold with an assumption of regular required upgrades. More like a waterheater than waterpipes.
10 year ago Qt was being used for major projects. I was using it in 2001 for security related projects it was so solid. KDE an entire GUI based on Qt was founded in 1996, and on KDE 3 by 2004. GTK+ was 1997. wxWidgets is even older, 1992. Open Interface dates to 1991. MKS which was quite good is from 1994.
As for being dependent on XP I don't see how that happened because of libraries. I can understand the TCO argument for Windows but that doesn't say anything about XP.
What are you talking about? iPhone users are more likely to be Republican than Democrat. They are more likely to be social liberals but that has nothing to do with economic policy. There is nothing hypocritical from decrying the economic divide regardless of whether you are benefiting from it. The hypocrisy would be to claim to decry it in general and then to rebel against most all specific policies to correct it.
I read your post. I guess in some truly vague sense I could locate a pool of memory at both points and including a synchronization mechanism in them. But if everything expects latencies in the nanosecond range with gigabytes of bandwidth per second and I move to latency in the millisecond range with bandwidth of megabytes per second things will slow down a lot. 5-10ms of jitter is highly noticeable and disturbing. An application going from 2ms to 2seconds of jitter is going to go from subconsciously uncomfortable to completely uninterpretable by the human brain. X11 does a lot to compensate for differences in performance. None of that is there with just a naive buffer.
As for the rest, about what you are or are not going to use ultimately unless you are networking on a LAN I think the networking in Wayland will be so superior there will be no holdups. So say by 2020 you will be using Wayland. But Wayland's first use case is completely local getting performance for things like video and games up when both the server and client are on the same machine.
More generally though. End users aren't going to really vote on X11 vs. Wayland. X11 on Wayland works fine so there is going to be no good reason for the clients not to be on Wayland. I'm sure there is going to be some stubborn group of people who insist on X11, but ultimately the hardware aren't going to support X11, nor are the distributions going to focus on it so they'll mostly get moved in a few years after the main body of Linux users move regardless. The choice is at the application writer side. The applications are going to have to decide if they want to implement X11 or Wayland. If they pick Wayland then there is no X11 networking: you either use the Wayland networking, you don't network or you pick a different application. And frankly the advantages for local (and WAN networking) are going to be so huge I can't see application designers making the X11 choice. Plus as I mentioned in the other thread, its possible the toolkits will move forcing the applications. So I don't really see it as optional.
This is like any other upgrade. Some people resist for a while but eventually everyone goes. I remember people holding on to XyWrite on dot matrix printers well into the 1990s, but there are probably 0 doing it still today. Sure there will be some group of people who try and maintain X11 after Wayland becomes normative. Sure there will be distributions focused on X11. But honestly how passionate do you see those people 5 years later? Do you really see people sticking with making those distributions for years out of nothing but spite?
They are not following the X11 model. From a networking perspective there is not going to be a "native Wayland app" but rather a "native KDE app" or a "native Gnome app". The connection is going to be KDE application to client's KDE library with Wayland providing hooks on both sides. Wayland cannot possible know enough about the applications to make intelligent choices about things like caching and prioritization, while the client's GUI libraries can do that. Which is exactly how RDP works.
Now you are absolutely right that the only one existence is the Wayland model. Last year that one wasn't existent. Both KDE and Gnome have announced they are going to start working the Wayland model into their systems. My guess is this takes several years and probably a really good implementation will require breaking changes and so will be done for the next major release (Gnome 4, KDE 5). As they start seriously moving in this direction Gnome 4 / KDE 5 will have to make a choice about whether they are Wayland primarily with some level of X11 support or whether they are X11 with some Wayland advanced features. I suspect Gnome having already decided touch (where getting down to 1ms latency is incredibly important) is key will definitely make the change.
So no. This is not going to be naive networking. This is going to be intelligent networking where the vast majority of applications / GUIs / toolkits will know they are talking over a network and will be expected to compensate and respond intelligently to the additional complexity. This is not "network transparency" because it is not transparent to the application / toolkit. Now there may be some base Wayland way of doing things, using the built in RDP structures entirely but that's going to suck. Possibly it will suck a little less than X11 over WAN sucks because there will be less roundtrips and more happening locally, but that's not the final desired result. Possibly it will suck a little more because unlike the X11 case no one is going to care too much about making it work well, so that while in theory it will be better that may only be in theory. The goal is to implement RDP for Linux and that requires the cooperation of parts higher than Wayland on the stack.
Application designers are quite likely going to spend time coming up with application, data and window caching and prioritization strategies that make sense for their application. GUI designers are quite likely going to spend time coming up with application, data and window caching and prioritization strategies that make sense for their application. And those strategies will change as use cases change. It is entirely possible for example that a local wired connection with a latency under 20ms will do things entirely different than an international or cellular connection with latencies around 150ms. And if you think about it makes sense to push the intelligence as far up the chain as possible. The specific runtime can make better choices than the application and compile time can make better choices than the GUI can make better choices than Wayland.
The compositor is just going to provide the most basic level the hooks for the GUI implementation. This isn't like X11 where X11 is going to do most of the heavy lifting for networking. The RDP is mostly going to be at the GUI level with the Wayland having hooks.
@Sjames
That's not true. The RDP can be individual windows or even less, it is up the the widget set. How KDE or Gnome implement it will be up to them not the Wayland team.
What? They are getting higher latency every day. We are increasing hops by doing things like wireless or bouncing off bluetooth. Moreover fundamentally we can't make the planet smaller of the speed of light faster. So there is nothing in the near tern that's likely to bring X11 latencies down to acceptable levels for WAN.
And in the other direction, screens are getting bigger (though OpenGL offers a workaround here) faster than memory is getting faster. So pushing windows buffers around of multiple megabytes is stressing performance for local.
You can't add network transparency later. It is a fundamental design choice. If application and display buffers are shared then lots of things simplify and you get performance improvements but transparency is out the window. That's why you can't have transparency under GDI, Aqua or Aero for example.
@Sjames
Networking on Wayland is mostly like Windows RDP. Which works fantastically over WANs. So local is an upgrade over X11, WAN is an upgrade over X11 and LAN is a downgrade. That's a reasonable tradeoff given today and likely future usage.
An original work that happened to be released under GPLv2. There is some question whether any code was ever accepted into Waze outside of that owned by the group of people that signed all rights over to Google (i.e. if Google owned every line of Wave version 2). According to Google they unquestionably own ever line of Wave v3.
So yes you are understanding this properly and IMHO Roey Gorodish is going to lose fast.
If the Waze founders own it then Waze may own copyright and not be a license holder which means they can freely relicense.
They are claiming the particular data was licensed under GPLv2. However the copyright entity Freemaps have never claimed that Waze's proprietary version was a violation. Waze claimed to have a separate license and they've never disputed this claim.
Jane --
Waze owned copyright on most (if not all of the code) they weren't a licensee. Which means they can relicense at will. Now of course any code released under the GPLv2 by them is still under GPLv2 but that's it.
Exactly. And just to add. Even if Waze did accept submission for version 2 that are still in version 3 . Roey Gorodish would only have standing if it were his code. This would be a copyright violation not a license violation and since he doesn't hold copyright....
I'm sure Waze had a proprietary license for the data they clearly said over and over they did.
This seems like a poorly thought out case. . Roey Gorodish loses quick IMHO.
The article linked is really quite good. According to the article are 2 prongs to the suit the code and the data.
Waze version 2 was GPLv2. Wave version 3 was proprietary. Waze claimed it rewrote the code. Roey Gorodish wants to examine the code to prove there was a violation but Israeli law doesn't allow discovery without some evidence. As an aside I also have question in that as as I know if Waze owns the code they can multilicense it. Roey Gorodish IMHO would have to find his code not just any code.
There is also the question of map data. Freemap’s data was released both proprietary and GPL and then only proprietary. I'm not sure what the basis of the claim is.
Under those fields up north there is soil. Besides we know how to repair farm territory. Cereals moved rather successfully thousands of years ago we are better now. I can't demonstrate what people 300 years from now will be able to do.
That isn't a point of dispute. I don't disagree with you.
I don't think that's true. Take the current USA standard of living. If raw commodity food prices were to double there would be little to no change in caloric intake in the USA. Food as a percentage of income has been falling rapidly, which has induced shifts to more meat and even still its falling off. In a world where everyone's income is 380x higher than today (i.e. even 2% growth) food prices wouldn't matter much. The raw price of growing food even for very poor people would be a minor expense. Transportation and distribution might account for close to 100% of the food costs.