They either really are getting close to ARM in power consumption, or they think x86/Wintel compatibility on a phone is more important than battery life....... If even Microsoft think power consumption is more important, then you have to wonder. Who wants to run a desktop app on a phone anyway? Be interesting to see, maybe they have something after their bluster about "bring it on". Personally, I don't think it matters even if they do, Wintel compatibility doesn't really matter, certainly not on a phone (or to some of us, at all), and there is no way it will be as cheap as ARM. Still, probably will help them (and Microsoft, because they can't escape x86) fight the ARM laptops coming after the tablets. If might also help them both against the ARM data centres. Anyway, nice to see a processor architecture fight, we all win with real competition.:-)
Uses the armel Debian userland in a disk image and chroot. http://pavelmachek.livejournal.com/73828.html
Been something actually making me thinking of finally getting a smart phone.
No, he's not. (Which is why he has 5 and you 0). Just googled "William Edward Hickman" "Ayn Rand."
William Edward Hickman was a real life monster. Anyone who admires such a man is broken. Ayn Rand did seam to admire him. She either was a psychopath herself, or admired them as being "free".
The idea that psychopathy helps in the world of business is not new. First I saw it was in a NewScientist article years ago about, I think it was called "Snakes in suits" which probably related to a book on the same subject under the same name.
This happens with every generation of console. When they come out they are great (a fixed standard of decent hardware), but by the end of their life span, they are looking pretty rubbish against even a moderate PC. Not long after that really starts to be a problem, a new console is announced and not long after that, released. Then the cycle starts again. No nothing new here at all.
What you say is true, and I think it's very sad. This is how MS will control us for another decade or two. When.NET has enough weight, MS will make Mono's life difficult. All those *nix server that where running Java will be in the end swapped for Windows servers running.NET. In our company *nix boxes have been replace with Windows boxes to run.NET because Mono doesn't quite do the job. That is right now and this is only going to get worse. I'll be damned if I'm going to help this, so I fight in what little way I can. But it's not just the politics, C# is replacing C++ for desktop apps, least internally at my company, and the results are quite a bit slower and hungrier. This has caused an increase in user dissatisfaction, at very least from users able to use task manager. Though to be fair, other measure of user satisfaction may have gone up, for reasons the C++ app was replaced (though those reasons where nothing to do with language!). It might be a pointless fight, but while I can, I will fight and hope others see the clouds ahead and join, or I'm wrong.
No, no and no. Boost and STL are good, but they aren't right for every job. It's good for high level C++, I'm not convinced it's good for low level C++. If you are doing low level C++, best make your own lib that does only what you want in the way you want. For instance, you are using relative pointers everywhere so you can use the same memory in multiple address spaces, or so copying and read/writing to/from disk doesn't require any extra work. Then C++ is like C with classes and templates. It's this breadth that is C++ greatest strength, but also the cause of it's criticized complexity.
"
But Bill starts to wonder By 1996 Gates is wondering if "by creating cross-platform solutions we risk weakening the Windows franchise - hurting ourselves rather than let Netscape hurt us." Neault's staff are writing that they want "the WIN32 layer to be fairly mediocre in performance and feature coverage. We want it to be just good/cheap/timely enough to get a lot of people to use it," and that "we don't want it to work too well. A non-objective is total redeployment of Windows on Unix."
"
How does that not ring Mono before Mono?
The tuxdeluxe article is that MS make sure they can check the box for interoperability, with out acturally do it. This is surspected to be deliberate policy.
Just swap Win32 for.NET, WISE for Mono, Java for POSIX (though not mentioned in this article by name)
It's just a little bit of history repeating.......
Especially if they pick up a load of Java business. Then there is no way they are going to just let Mono/Linux have any more of the pie then they can't prevent. If Java goes down in flames and C# picks up the pieces we are stuffed, even if Windows is your platform of choice (think IE after they owned the browser). Maybe worse than ever because it's not just the API they control, but the language. I know many think this is crazy tin hat stuff, but to me it echoes past behaviour and it seems like some would go for an easy life today for a nightmare tomorrow. At the end of the day, it's what makes money not provides service, you only get service when the two line up. Yer, mark me down, I hope I'm wrong too.
Diagrammatic/Node-base programming is an idea that's been pushed as the future for a long time. I don't buy it. If it was going to work, it would be common place now, and it's not. Input method has nothing to do with it. The reason I think it never became common is that it only works in special cases. Believe it or not, I was once a nodal believer, about six years ago, and wrote a plug/node system (game middleware engine/editor thing) as a home project (with optional relative pointers so encapsulated node systems (as a node) could be copied to and from disk/memory easily). It's damn hard to do without adding lots of cost. It's harder to keep things clear. I would write a node system again in the right places, but it's not something that is ever going to replace text based programming. Try doing something large with an existing node/diagrammatic system and compare it with one done in a normal text based language. Your argument mirrors that GUIs are better than CLIs for everything, which I'm sure you don't believe either. Text is the simplest place computers and humans meet. Quite a good light read on this is http://artlung.com/smorgasborg/C_R_Y_P_T_O_N_O_M_I_C_O_N.shtml (In the Beginning There Was the Command Line)
I don't want what you are smoking, I don't think it's good for you..... All the magic 2D or 3D programming done with arms waving in the air, will be done by children or for simple tasks/"programmers", much as it has always been. The real work, like writing the magic 2D/3D programming interface for children/simple-task/simple-programmers, will be done as now, in text. In the same way the GUI never replaced the CLI for advanced use. Text is the best common language for computers and humans to talk to each other in, it's simple and clear. An example of the node programming world right now is art packages like Maya. They have nodes you wire together to create behaviour. Advanced artists/animators do some very clever things with these nodes, but not everything can be done through nodes. So there is a script language the advanced learn. But they still can't make new nodes types. The new node types are provided by programmers. Also somethings are just not quick enough done through node wiring or scripting, and have to be programmed properly. There is actually a path here for artists/animators to become programmers, and I've seen it happen a few times. Node base programming has it's place, but it will never replace good old text based programming for the real work.
Oh come on. Keyboards aren't going anywhere. You are always going to have a keyboard if you are going to enter text, touch thingys will never be as fast, not until there is some tactile feedback. Not only are keyboards not going anywhere, neither are text interfaces.
Touch screen are great but for speed and prolonged use, do you want to move you whole arm (covering you view of the output device), or twitch you wrist of your rested arm?
For personally browsing Facebook, sure something that is basically just a screen is fine, but let's not pretend that's the future for everything.
Have you never watched some SciFi where they are waving their arms about and talking to use the computer and thought "that's taking ages and must be exhausting".
In reality it would be "Computer, I'm tired, my arms hurt. Can I have a keyboard and mouse? No, an ASCII keyboard. No, one connected to you. No, not that kind of mouse. Oh sod it, can I have my old PC back? What do you mean you can't do that and why are you calling me Dave?"
This is something I was wondering. As I understand it splice/sendpage just marks a page belonging to a block of one filesystem as a page of a block of another filesystem. Thus no memory copying, the same RAM the data was read from one disc is used to write that data to another disc. So I think markhahn has a point, the problem is userspace, not the kernel. The cp code could be faster if it used splice/sendpage zero copy stuff (when dealing with filesystems that it is possible to do so (i.e non-FUSE)).
Have to say I've been wondering the same thing. Microsoft is pushing towards.NET for the whole platform while the Linux world stays C,C++ and only something else, normally python, where speed doesn't matter. Java and Mono aren't very often used for popular apps and tend to be replaced for smaller, faster, C/C++ alternatives. This move is only going to make the Windows platform seam even slower compared with Linux (or Mac and everything else for that matter).
Yes, I believe so. I thought they had bought some C++ stock exchange development software house, but I can't find anything about it now. In which case it's not so much Linux faster as C++ faster, which we already knew. Not to take anything away from Linux, but I'm not sure it's what is making the big speed difference.
Yes, it's a marketing problem. But it's a big one for them. They can't use the weight of the Windows brand. Now true, they pulled that off, just about, with XBox. But man did it cost them, few companies would have had the money to keep pumping in at that scale for so long. If they do use the Window brand, it needs to run Windows software, this isn't like a phone, people expect it to be more of a computer. What caught people by surprise with the iPhone was that it almost was a computer, but because it wasn't, they didn't expect the same software. It was successful because of that, and iPad is piggy backing on the iPhone success by using the same software. When this MS-tablet comes out, all it will have against the other tablets is that it is thin (and the others are quite thin already). Being able to run the software people know, would be a win, but that means x86, and even then, it must be powerful enough to run the day's Windows software. That means hotter and shorter battery life compared with the ARM tablets that people will be use to by then. If they do a ARM thing, it will have much less software then the other tablets, and I bet they use the Windows brand because it will increase sales, until everyone understands it's not really Windows in that it can't run Windows software. Either way, I see this as a coming flop. But time will tell.
They don't advertise WinCE as Windows. It's not a visible brand. Anything they called Windows, must run Windows software, and Windows software is x86. Windows is its software, few comes to Windows for Windows. Oh and I'm not complaining, just commenting.
If they use a x86 to compete with the ARM tablets it will have shorter battery life and run hot. If they use ARM (or something else giving good mA/mips), then people won't understand why it can't run their Windows software. If it looks and feels like Windows (and actually code wise, is Windows) but can't run Windows software, people won't like it. The platform is Windows software. It's the closed source curse, you are stuck on the hardware and API things are compiled for. Of course their is byte code, but then they will be competing again other tablets of similar spec, but with their apps byte coded while the others (Apple/Linux) are native. If that happens, bet MS's own apps are native for each platform, but they advice developers to use.NET to cover all MS platforms. But even then, are most consumers going to understand the difference between.NET apps and native apps? This to me has all the marks of a money blackhole while they try and complete in the tablet space.
They don't require hacking, it's optional. Unless you count sshing into a Ubuntu box to configure it as hacking...... I moved mine from Ubuntu to Debian, and the guide for doing so is pretty clear: http://www.cyrius.com/debian/kirkwood/sheevaplug/ I had a small issue, probably of my own making, but I solved it and learnt more about uBoot in the process. I don't what problems you had with the mini USB JTAG, again it's clearly documented. It is much much much more powerful then the 8bit AVR butterfly. A much better hack project to point people at is BeagleBoard (which now can run XBMC).
Because if you use a low powered external harddrive (the WD MyPassport, or what ever), it's still less power, less cost, and quieter as there is nothing but the harddrive actually moving (i.e, no fans). But like the laptop and unlike any OpenWRT solutions, it's running a full Linux, ARM Debian/Ubuntu repositories are pretty much the same as the x86 version.
Slashdot Mirror
They either really are getting close to ARM in power consumption, or they think x86/Wintel compatibility on a phone is more important than battery life....... If even Microsoft think power consumption is more important, then you have to wonder. Who wants to run a desktop app on a phone anyway? Be interesting to see, maybe they have something after their bluster about "bring it on". Personally, I don't think it matters even if they do, Wintel compatibility doesn't really matter, certainly not on a phone (or to some of us, at all), and there is no way it will be as cheap as ARM. Still, probably will help them (and Microsoft, because they can't escape x86) fight the ARM laptops coming after the tablets. If might also help them both against the ARM data centres. Anyway, nice to see a processor architecture fight, we all win with real competition. :-)
Uses the armel Debian userland in a disk image and chroot.
http://pavelmachek.livejournal.com/73828.html
Been something actually making me thinking of finally getting a smart phone.
No, he's not. (Which is why he has 5 and you 0). Just googled "William Edward Hickman" "Ayn Rand."
William Edward Hickman was a real life monster. Anyone who admires such a man is broken. Ayn Rand did seam to admire him. She either was a psychopath herself, or admired them as being "free".
The idea that psychopathy helps in the world of business is not new. First I saw it was in a NewScientist article years ago about, I think it was called "Snakes in suits" which probably related to a book on the same subject under the same name.
This happens with every generation of console. When they come out they are great (a fixed standard of decent hardware), but by the end of their life span, they are looking pretty rubbish against even a moderate PC. Not long after that really starts to be a problem, a new console is announced and not long after that, released. Then the cycle starts again. No nothing new here at all.
What you say is true, and I think it's very sad. This is how MS will control us for another decade or two. When .NET has enough weight, MS will make Mono's life difficult. All those *nix server that where running Java will be in the end swapped for Windows servers running .NET. In our company *nix boxes have been replace with Windows boxes to run .NET because Mono doesn't quite do the job. That is right now and this is only going to get worse. I'll be damned if I'm going to help this, so I fight in what little way I can. But it's not just the politics, C# is replacing C++ for desktop apps, least internally at my company, and the results are quite a bit slower and hungrier. This has caused an increase in user dissatisfaction, at very least from users able to use task manager. Though to be fair, other measure of user satisfaction may have gone up, for reasons the C++ app was replaced (though those reasons where nothing to do with language!). It might be a pointless fight, but while I can, I will fight and hope others see the clouds ahead and join, or I'm wrong.
No, no and no. Boost and STL are good, but they aren't right for every job. It's good for high level C++, I'm not convinced it's good for low level C++. If you are doing low level C++, best make your own lib that does only what you want in the way you want. For instance, you are using relative pointers everywhere so you can use the same memory in multiple address spaces, or so copying and read/writing to/from disk doesn't require any extra work. Then C++ is like C with classes and templates. It's this breadth that is C++ greatest strength, but also the cause of it's criticized complexity.
"
But Bill starts to wonder By 1996 Gates is wondering if "by creating cross-platform solutions we risk weakening the Windows franchise - hurting ourselves rather than let Netscape hurt us." Neault's staff are writing that they want "the WIN32 layer to be fairly mediocre in performance and feature coverage. We want it to be just good/cheap/timely enough to get a lot of people to use it," and that "we don't want it to work too well. A non-objective is total redeployment of Windows on Unix."
"
How does that not ring Mono before Mono?
The tuxdeluxe article is that MS make sure they can check the box for interoperability, with out acturally do it. This is surspected to be deliberate policy.
http://www.theregister.co.uk/1999/07/18/analysis_how_ms_used/
.NET, WISE for Mono, Java for POSIX (though not mentioned in this article by name)
It's just a little bit of history repeating.......
Just swap Win32 for
I think Wine and ReactOS have not been targeted because they aren't big enough to be worth the major bad PR. Example, what about WISE ? http://en.wikipedia.org/wiki/Windows_Interface_Source_Environment seams like we might be set to fool for the same trick twice. http://www.theregister.co.uk/1999/07/18/analysis_how_ms_used/
A open standard isn't really enough : http://tuxdeluxe.org/node/296
Especially if they pick up a load of Java business. Then there is no way they are going to just let Mono/Linux have any more of the pie then they can't prevent. If Java goes down in flames and C# picks up the pieces we are stuffed, even if Windows is your platform of choice (think IE after they owned the browser). Maybe worse than ever because it's not just the API they control, but the language. I know many think this is crazy tin hat stuff, but to me it echoes past behaviour and it seems like some would go for an easy life today for a nightmare tomorrow. At the end of the day, it's what makes money not provides service, you only get service when the two line up. Yer, mark me down, I hope I'm wrong too.
Diagrammatic/Node-base programming is an idea that's been pushed as the future for a long time. I don't buy it. If it was going to work, it would be common place now, and it's not. Input method has nothing to do with it. The reason I think it never became common is that it only works in special cases. Believe it or not, I was once a nodal believer, about six years ago, and wrote a plug/node system (game middleware engine/editor thing) as a home project (with optional relative pointers so encapsulated node systems (as a node) could be copied to and from disk/memory easily). It's damn hard to do without adding lots of cost. It's harder to keep things clear. I would write a node system again in the right places, but it's not something that is ever going to replace text based programming. Try doing something large with an existing node/diagrammatic system and compare it with one done in a normal text based language. Your argument mirrors that GUIs are better than CLIs for everything, which I'm sure you don't believe either. Text is the simplest place computers and humans meet. Quite a good light read on this is http://artlung.com/smorgasborg/C_R_Y_P_T_O_N_O_M_I_C_O_N.shtml (In the Beginning There Was the Command Line)
It is early in HCI development, but it's not for nothing things have changed little since: http://www.youtube.com/watch?v=JfIgzSoTMOs
Right, enough, if you still believe this, that is your own problem. You'll have to take my keyboard from my cold dead hands!
I don't want what you are smoking, I don't think it's good for you..... All the magic 2D or 3D programming done with arms waving in the air, will be done by children or for simple tasks/"programmers", much as it has always been. The real work, like writing the magic 2D/3D programming interface for children/simple-task/simple-programmers, will be done as now, in text. In the same way the GUI never replaced the CLI for advanced use. Text is the best common language for computers and humans to talk to each other in, it's simple and clear. An example of the node programming world right now is art packages like Maya. They have nodes you wire together to create behaviour. Advanced artists/animators do some very clever things with these nodes, but not everything can be done through nodes. So there is a script language the advanced learn. But they still can't make new nodes types. The new node types are provided by programmers. Also somethings are just not quick enough done through node wiring or scripting, and have to be programmed properly. There is actually a path here for artists/animators to become programmers, and I've seen it happen a few times. Node base programming has it's place, but it will never replace good old text based programming for the real work.
Oh come on. Keyboards aren't going anywhere. You are always going to have a keyboard if you are going to enter text, touch thingys will never be as fast, not until there is some tactile feedback. Not only are keyboards not going anywhere, neither are text interfaces.
Touch screen are great but for speed and prolonged use, do you want to move you whole arm (covering you view of the output device), or twitch you wrist of your rested arm?
For personally browsing Facebook, sure something that is basically just a screen is fine, but let's not pretend that's the future for everything.
Have you never watched some SciFi where they are waving their arms about and talking to use the computer and thought "that's taking ages and must be exhausting".
In reality it would be "Computer, I'm tired, my arms hurt. Can I have a keyboard and mouse? No, an ASCII keyboard. No, one connected to you. No, not that kind of mouse. Oh sod it, can I have my old PC back? What do you mean you can't do that and why are you calling me Dave?"
I think the BSD licence is why the BSD OS is where it is.
The GNU cp doesn't either:
http://git.savannah.gnu.org/cgit/coreutils.git/tree/src/copy.c
Could be an interesting test.
That doesn't seam to be true.
http://stackoverflow.com/questions/1580923/how-can-i-use-linuxs-splice-function-to-copy-a-file-to-another-file
This is something I was wondering. As I understand it splice/sendpage just marks a page belonging to a block of one filesystem as a page of a block of another filesystem. Thus no memory copying, the same RAM the data was read from one disc is used to write that data to another disc. So I think markhahn has a point, the problem is userspace, not the kernel. The cp code could be faster if it used splice/sendpage zero copy stuff (when dealing with filesystems that it is possible to do so (i.e non-FUSE)).
BusyBox certainly doesn't http://git.busybox.net/busybox/tree/libbb/copyfd.c
Have to say I've been wondering the same thing. Microsoft is pushing towards .NET for the whole platform while the Linux world stays C,C++ and only something else, normally python, where speed doesn't matter. Java and Mono aren't very often used for popular apps and tend to be replaced for smaller, faster, C/C++ alternatives. This move is only going to make the Windows platform seam even slower compared with Linux (or Mac and everything else for that matter).
Yes, I believe so. I thought they had bought some C++ stock exchange development software house, but I can't find anything about it now. In which case it's not so much Linux faster as C++ faster, which we already knew. Not to take anything away from Linux, but I'm not sure it's what is making the big speed difference.
Yes, it's a marketing problem. But it's a big one for them. They can't use the weight of the Windows brand. Now true, they pulled that off, just about, with XBox. But man did it cost them, few companies would have had the money to keep pumping in at that scale for so long. If they do use the Window brand, it needs to run Windows software, this isn't like a phone, people expect it to be more of a computer. What caught people by surprise with the iPhone was that it almost was a computer, but because it wasn't, they didn't expect the same software. It was successful because of that, and iPad is piggy backing on the iPhone success by using the same software. When this MS-tablet comes out, all it will have against the other tablets is that it is thin (and the others are quite thin already). Being able to run the software people know, would be a win, but that means x86, and even then, it must be powerful enough to run the day's Windows software. That means hotter and shorter battery life compared with the ARM tablets that people will be use to by then. If they do a ARM thing, it will have much less software then the other tablets, and I bet they use the Windows brand because it will increase sales, until everyone understands it's not really Windows in that it can't run Windows software. Either way, I see this as a coming flop. But time will tell.
They don't advertise WinCE as Windows. It's not a visible brand. Anything they called Windows, must run Windows software, and Windows software is x86. Windows is its software, few comes to Windows for Windows. Oh and I'm not complaining, just commenting.
I'm not so convinced. Long way to go, lots of obstacles inherent to x86. I will believe when I see it.
If they use a x86 to compete with the ARM tablets it will have shorter battery life and run hot. If they use ARM (or something else giving good mA/mips), then people won't understand why it can't run their Windows software. If it looks and feels like Windows (and actually code wise, is Windows) but can't run Windows software, people won't like it. The platform is Windows software. It's the closed source curse, you are stuck on the hardware and API things are compiled for. Of course their is byte code, but then they will be competing again other tablets of similar spec, but with their apps byte coded while the others (Apple/Linux) are native. If that happens, bet MS's own apps are native for each platform, but they advice developers to use .NET to cover all MS platforms. But even then, are most consumers going to understand the difference between .NET apps and native apps? This to me has all the marks of a money blackhole while they try and complete in the tablet space.
http://meetings-archive.debian.net/pub/debian-meetings/2010/debconf10/low/1252_How_We_Can_Be_the_Silver_Lining_of_the_Cloud.ogv
They don't require hacking, it's optional. Unless you count sshing into a Ubuntu box to configure it as hacking...... I moved mine from Ubuntu to Debian, and the guide for doing so is pretty clear: http://www.cyrius.com/debian/kirkwood/sheevaplug/ I had a small issue, probably of my own making, but I solved it and learnt more about uBoot in the process. I don't what problems you had with the mini USB JTAG, again it's clearly documented. It is much much much more powerful then the 8bit AVR butterfly. A much better hack project to point people at is BeagleBoard (which now can run XBMC).
Because if you use a low powered external harddrive (the WD MyPassport, or what ever), it's still less power, less cost, and quieter as there is nothing but the harddrive actually moving (i.e, no fans). But like the laptop and unlike any OpenWRT solutions, it's running a full Linux, ARM Debian/Ubuntu repositories are pretty much the same as the x86 version.