It's not tough to deploy java applications on windows, you just bundle the jre with the app, and use a small tool to convert the jar to an exe that use the included jre bundle. That way you always know and control exactly what java version your application is running on, and you don't mess with any of the users existing java installation if he got one. Only problem is really that your installer will be ~8MB larger.
Mainframes are breathing their last gasp; they will soon exist only in cases where you need very fast access to all of very large data sets.
Why would you need a mainframe just for that? A fully upgraded SGI Altix 4700 will be faster then IBM's fastest mainframe and I also think that HP's superdrome is up there in performance.
How the hell do you get a linux desktop to become unresponsive? I have used Linux on my desktop for many years, and have newer seen my desktop become unresponsive for even a single second*. Some applications(Hi firefox) may be unresponsive, but X and linux always respond.
*With the exception of when Kde/Plasma crashes. If they do that most thing become unresponsive a few seconds until the reload is complete.
But can other operation systems do that when running in "NON fullscrenn" mode? (This is, when using normal programs?).
Can windows XP for example do a "swap buffer on frame breaks" in a normal program window, while other software is also displaying?
And while we are talking about it does it matter anymore, now that anyone is using lcd screens.
I might have missed something importent about lcd, but i always thought that lcd screens were without any frame break(vblank) so that the time to change any pixel would be the same, no matter the display state and no matter what have just changed. (With the exception, that it take longer time when changing a pixel, that have just changed).
Or is it the case that the lcd hardware can do it, but that the graphics card still send the image to the screen, one line at a time. (That would give the worst of both worlds).
I am currently on a Windows XP(Radeon X600 graphics card should be fast enough) computer home with my parents, and both browsers I have tested in(IE, and FireFox) don't take frame breaks into account at all when scrolling. (At least as far as I can see)
Don't the application(Or atleast the gfx subsystem) need to know the screen response time in order to swap buffer when all pixels have changed color. Hmmm so many questions, I guess I need to read up on how lcd screens handle pixel change.
In May 2008, Adobe launched the Open Screen Project (Adobe link), which made the SWF specification available without restrictions. Previously, developers couldn't use the specification for making SWF-compatible players, but only for making SWF-exporting authoring software. The specification remains incomplete, however, as it does not include any details regarding RTMP or Sorenson Spark,[27] both of which are widely used to distribute video through Flash.
So the only missing piece is the video encoding and that can be handled by mplayer already.
But the question is: Is it worth paying 100000$ more for a sun box with the mentioned features? I mean yes they are nice features but still.
They don't say much about their software, but I guess they run the boxes in some kind of raid 1/storage rotation, where all data are stored on more then 1 box. One thing I do not understand in their setup, is the lack of spare drives.
As an owner of an old Mac laptop that still Run OS X 10.1.8 let me say, that I still think that 10.2 should have been free, because 10.1.8 is so buggy and it will newer be fixed. Using NFS to mount a disk will almost always crash my kernel within an hour.
A big problem with the way that Apple does upgrades is that to get bugfixes, you often do need to buy the newest OS X and it's seldom free. I wish they would split the os from their applications, so the os bugfixes/upgrades were free, but they could charge you if you really wanted the i* software. I don't really like having to buy a new os, just to get working NFS.
That is ironic enough, one of the reasons that people keep their windows XP boxes(Instead of 'upgrading' to Vista. Windows XP was buggy when Microsoft released it, but Microsoft have used the latest many year bug fixing it, and all those bug fixes are free.
And XP bugfixes don't ever require new hardware unlike Mac OS X, where even if I wanted to buy the newest Mac OS X upgrade for Power PC, my laptop could not use it because its graphics chip is to slow. So I can't get all the bug fixes that Apple have developed and released because my Gfx chip is to slow. (Unless I buy new hardware. I begin to see why people call Apple a hardware company).
They were issues in the past. ID newer made any money on any of their linux ports. The only reason they ported to linux was that they thought their software should be available anyware.
So it's more a "There were newer any economical motivation for ID to port their software to Linux. With the new requirements of their 3d Engine, even less linux users will be able to use it, and the cost to port and support it will be larger".
But you only need to acknowledge each ip package, not each ethernet package. And a ip package can be 64K. When I download using http with max speed, I get something like 4.5MBit down and use 2.6K up*, giving me a ration of download/upload at more then 100
*According to the linux network transfer monitor tool included with kde3/fedora 6.
The problem with the Amiga floppy was not really the drive itself but the floppy controller. You can read amiga disks on a standard pc diskdrive, but it require either a special disk controller, or 2 diskdrives and some REALLY clever software*.
*Googling for why it require 2 diskdrives for a pc to read 1 amiga disk, will really show some great software hacking.
Have you fixed battle.net so that 2(or more) players behind nat can play together (Very importent now that you removed lan play), and play against other players in a 2 vs 2?
The Starcraft/battle.net that is online now can't do this.
I have newer used "Adobe Connect"* but a major difference is that Google wave is a protocol.
This mean that anyone can make their own wave server**, and you can make your own graphics interface to the wave, so you don't need to use whatever google make. (I think that this will be importent, because in the end someone will make a much better "non browser" gui to wave).
*So this analysis is based on 2 minutes wiki reading
**Your own wave server can still talk to all the other wave servers. This is importent because it mean that Wave is(Or atleast can be) total independent of any google hardware/software. So there is no fear of being locked to a singel vendor who will not implement what you need.
In particular, you *can* write a halting problem solver for arbitrary programs of length n
For this to be true i think n should be the size of the state space, not the size of the initial program. The complexity does not really depend on the size of the start program, but the size of the state space.
Just think of something simple like "Game of life". The program itself is small, but the state space is large.
There are no specific limit to the size of a kernel. This mean that I can always make a kernel that is larger then the largest existing kernel. This kernel will de unique(Because it's larger then any other kernels).
As I can thus always make a new unique kernel the number of different kernels are not finite. So it must be infinite.
Godel's Incompleteness Theorem just say(In this context) that there exists infinite many kernels that are correct but which can not be proven correct. It does not say that no kernel can be proven correct.
So they just happen to write one of the kernels that could be proven.
But you have to know the other person's account number. Which presumably means you have to trust them. I'm not giving my bank account number and routing number to many people; well mostly because you can use those numbers to withdraw money as well since there's no "deposit only" routing number.
??? I have newer considered my account number a secret, and there is no way you can get any money from my account, just by knowing my account number*. How would you propose to use my account number to take money from the account?
Example: I have my own small company(Just me) and I always include all account information in the invoice. That way the customer can just pay me by doing a direct deposit to my account with a payment note, saying this is payment for invoice XX
*Unless the bank really fuck up, but in that case they do cover the costs of their own mistakes.
There doesn't seem to be any reasonable way for ordinary bank customers to transfer money to customers of other banks
Really? That sound insane. Here in Danmark I can just walk down to the ATM enter a target account and transfer the money that way(It's even free I think). It will even print a slip, with the transfer details so I can prove that I did transfer money.
Or I could use this fancy new internet bank thing, which allow me to view all my accounts, and transfer money to anyone with a bank account in EU. And it's all free(Well atleast internal in denmark, it might cost a bit to transfer to other countries), because its so much cheaper for the bank when I use the internet, compared to me having to walk down to the branch building and talking to them in person. (And they even support Both Linux and MacOS)
Facebook can easily implement wave without paying Google. They just need to run their own wave server. Wave have been designed with the explicit goal to allow users on different servers(Called federations) to talk to each other as if they were with the same Wave provider.
So facebook just have to develop their own wave server, or compile and install the opensource(bsd style) wave server that google have released.
And how could I forgot.net. The system where Microsoft once again wrote an entire widget system to replace the widgets in win32. A.net button is not a win32 button.
It seems to me that the standard way(As in: done by most) to develop a gui for Windows is to first find an alternate for the win32 widgets. And I don't see the difference between using the hellspawn that Adobe/Microsoft/Adobe use, or using Qt. It's all just a new gui system, which use gdi+ (And a few win32 calls) as a backend.
Win32? The gui system which don't even have a split pane. (Yes I have developed for Win32, an interesting job but not one I care to repeat).
I can not mention a major application released within the last 4 year, with an interface based on pure win32 widgets. Not even microsofts own software. Win32 miss far to many things. (Even the toolbar in Microsoft internet explorer 6 is different from the Win32 toolbar).
Here is an other example: The menu that popup when you press the right mouse button: Does it come when you click or release the button? Answer: That depend on the application, because win32 don't even handle that stuff.
So if the only standard interface for windows is win32, then I can't mention any major application released within the last 5 years which have used the standard windows interface. Nothing released from Microsoft. Nothing released by Adobe. Not a single 3D editor, nothing released from Apple. Not a single browser. Well nothing (Except google earth, that is the only pure major win32 application I can mention).
Would you call MFC* part of the windows interface? If yes, then how does it differ from Qt? Developers who use MFC/QT both have to download MFC/QT, and both does a static link to (part of) the library, so the user don't have to download anything.
MFC is not part of windows. It comes with Visual studio and you are not allowed to static link to it, unless you have a license to Visual Studio which is in no way part of Windows.
I would say that windows don't have a standard GUI application development platform either.
But how do you define standard anyway, and in what way is QT less standard then the "roll your own gui, because Win32 sucks" that so much software(Including anything from Adobe and microsoft office*)
And what is non-standard about QT? I have not seen QT break anything since between the switch between QT3 and QT4. (Unlike MacOS X where an upgrade suddenly required some gui callback methods to be reentrent, breaking lots of software.
*Just take the ribbon interface as an example. They developed an entire new gui system, just for office.
Slashdot Mirror
It's not tough to deploy java applications on windows, you just bundle the jre with the app, and use a small tool to convert the jar to an exe that use the included jre bundle. That way you always know and control exactly what java version your application is running on, and you don't mess with any of the users existing java installation if he got one. Only problem is really that your installer will be ~8MB larger.
I am pretty sure that the ribbon code used in office is not available as any kind of WPF or WinForm library.
What I could find were ribbon re-implementations done by 3.parties in .net but I can't imagine Firefox using the .net framework for widgets.
The Ribbon documentation is available, but I don't think there is any code available at all. Or have that changed in the last few months?
Mainframes are breathing their last gasp; they will soon exist only in cases where you need very fast access to all of very large data sets.
Why would you need a mainframe just for that? A fully upgraded SGI Altix 4700 will be faster then IBM's fastest mainframe and I also think that HP's superdrome is up there in performance.
How the hell do you get a linux desktop to become unresponsive? I have used Linux on my desktop for many years, and have newer seen my desktop become unresponsive for even a single second*. Some applications(Hi firefox) may be unresponsive, but X and linux always respond.
*With the exception of when Kde/Plasma crashes. If they do that most thing become unresponsive a few seconds until the reload is complete.
It frees apple from needing Microsoft software on the client.
And they did license the access to exchange from Microsoft, so they can't just lock them out.
But can other operation systems do that when running in "NON fullscrenn" mode? (This is, when using normal programs?).
Can windows XP for example do a "swap buffer on frame breaks" in a normal program window, while other software is also displaying?
And while we are talking about it does it matter anymore, now that anyone is using lcd screens.
I might have missed something importent about lcd, but i always thought that lcd screens were without any frame break(vblank) so that the time to change any pixel would be the same, no matter the display state and no matter what have just changed. (With the exception, that it take longer time when changing a pixel, that have just changed).
Or is it the case that the lcd hardware can do it, but that the graphics card still send the image to the screen, one line at a time. (That would give the worst of both worlds).
I am currently on a Windows XP(Radeon X600 graphics card should be fast enough) computer home with my parents, and both browsers I have tested in(IE, and FireFox) don't take frame breaks into account at all when scrolling. (At least as far as I can see)
Don't the application(Or atleast the gfx subsystem) need to know the screen response time in order to swap buffer when all pixels have changed color. Hmmm so many questions, I guess I need to read up on how lcd screens handle pixel change.
Not anymore. (From http://en.wikipedia.org/wiki/Adobe_Flash)
In May 2008, Adobe launched the Open Screen Project (Adobe link), which made the SWF specification available without restrictions. Previously, developers couldn't use the specification for making SWF-compatible players, but only for making SWF-exporting authoring software. The specification remains incomplete, however, as it does not include any details regarding RTMP or Sorenson Spark,[27] both of which are widely used to distribute video through Flash.
So the only missing piece is the video encoding and that can be handled by mplayer already.
But the question is: Is it worth paying 100000$ more for a sun box with the mentioned features? I mean yes they are nice features but still.
They don't say much about their software, but I guess they run the boxes in some kind of raid 1/storage rotation, where all data are stored on more then 1 box. One thing I do not understand in their setup, is the lack of spare drives.
As an owner of an old Mac laptop that still Run OS X 10.1.8 let me say, that I still think that 10.2 should have been free, because 10.1.8 is so buggy and it will newer be fixed. Using NFS to mount a disk will almost always crash my kernel within an hour.
A big problem with the way that Apple does upgrades is that to get bugfixes, you often do need to buy the newest OS X and it's seldom free. I wish they would split the os from their applications, so the os bugfixes/upgrades were free, but they could charge you if you really wanted the i* software. I don't really like having to buy a new os, just to get working NFS.
That is ironic enough, one of the reasons that people keep their windows XP boxes(Instead of 'upgrading' to Vista. Windows XP was buggy when Microsoft released it, but Microsoft have used the latest many year bug fixing it, and all those bug fixes are free.
And XP bugfixes don't ever require new hardware unlike Mac OS X, where even if I wanted to buy the newest Mac OS X upgrade for Power PC, my laptop could not use it because its graphics chip is to slow. So I can't get all the bug fixes that Apple have developed and released because my Gfx chip is to slow. (Unless I buy new hardware. I begin to see why people call Apple a hardware company).
They were issues in the past. ID newer made any money on any of their linux ports. The only reason they ported to linux was that they thought their software should be available anyware.
So it's more a "There were newer any economical motivation for ID to port their software to Linux. With the new requirements of their 3d Engine, even less linux users will be able to use it, and the cost to port and support it will be larger".
But you only need to acknowledge each ip package, not each ethernet package. And a ip package can be 64K. When I download using http with max speed, I get something like 4.5MBit down and use 2.6K up*, giving me a ration of download/upload at more then 100
*According to the linux network transfer monitor tool included with kde3/fedora 6.
The problem with the Amiga floppy was not really the drive itself but the floppy controller. You can read amiga disks on a standard pc diskdrive, but it require either a special disk controller, or 2 diskdrives and some REALLY clever software*.
*Googling for why it require 2 diskdrives for a pc to read 1 amiga disk, will really show some great software hacking.
Have you fixed battle.net so that 2(or more) players behind nat can play together (Very importent now that you removed lan play), and play against other players in a 2 vs 2?
The Starcraft/battle.net that is online now can't do this.
I have newer used "Adobe Connect"* but a major difference is that Google wave is a protocol.
This mean that anyone can make their own wave server**, and you can make your own graphics interface to the wave, so you don't need to use whatever google make. (I think that this will be importent, because in the end someone will make a much better "non browser" gui to wave).
*So this analysis is based on 2 minutes wiki reading
**Your own wave server can still talk to all the other wave servers. This is importent because it mean that Wave is(Or atleast can be) total independent of any google hardware/software. So there is no fear of being locked to a singel vendor who will not implement what you need.
In particular, you *can* write a halting problem solver for arbitrary programs of length n
For this to be true i think n should be the size of the state space, not the size of the initial program. The complexity does not really depend
on the size of the start program, but the size of the state space.
Just think of something simple like "Game of life". The program itself is small, but the state space is large.
Let me try to turn the argument around.
There are no specific limit to the size of a kernel. This mean that I can always make a kernel that is larger then the largest existing kernel. This kernel will de unique(Because it's larger then any other kernels).
As I can thus always make a new unique kernel the number of different kernels are not finite. So it must be infinite.
Godel's Incompleteness Theorem just say(In this context) that there exists infinite many kernels that are correct but which can not be proven correct. It does not say that no kernel can be proven correct.
So they just happen to write one of the kernels that could be proven.
But you have to know the other person's account number. Which presumably means you have to trust them. I'm not giving my bank account number and routing number to many people; well mostly because you can use those numbers to withdraw money as well since there's no "deposit only" routing number.
??? I have newer considered my account number a secret, and there is no way you can get any money from my account, just by knowing my account number*. How would you propose to use my account number to take money from the account?
Example: I have my own small company(Just me) and I always include all account information in the invoice. That way the customer can just pay me by doing a direct deposit to my account with a payment note, saying this is payment for invoice XX
*Unless the bank really fuck up, but in that case they do cover the costs of their own mistakes.
There doesn't seem to be any reasonable way for ordinary bank customers to transfer money to customers of other banks
Really? That sound insane. Here in Danmark I can just walk down to the ATM enter a target account and transfer the money that way(It's even free I think). It will even print a slip, with the transfer details so I can prove that I did transfer money.
Or I could use this fancy new internet bank thing, which allow me to view all my accounts, and transfer money to anyone with a bank account in EU. And it's all free(Well atleast internal in denmark, it might cost a bit to transfer to other countries), because its so much cheaper for the bank when I use the internet, compared to me having to walk down to the branch building and talking to them in person. (And they even support Both Linux and MacOS)
Facebook can easily implement wave without paying Google. They just need to run their own wave server. Wave have been designed with the explicit goal to allow users on different servers(Called federations) to talk to each other as if they were with the same Wave provider.
So facebook just have to develop their own wave server, or compile and install the opensource(bsd style) wave server that google have released.
And how could I forgot .net. The system where Microsoft once again wrote an entire widget system to replace the widgets in win32. A .net button is not a win32 button.
It seems to me that the standard way(As in: done by most) to develop a gui for Windows is to first find an alternate for the win32 widgets. And I don't see the difference between using the hellspawn that Adobe/Microsoft/Adobe use, or using Qt. It's all just a new gui system, which use gdi+ (And a few win32 calls) as a backend.
Win32? The gui system which don't even have a split pane. (Yes I have developed for Win32, an interesting job but not one I care to repeat).
I can not mention a major application released within the last 4 year, with an interface based on pure win32 widgets. Not even microsofts own software. Win32 miss far to many things. (Even the toolbar in Microsoft internet explorer 6 is different from the Win32 toolbar).
Here is an other example: The menu that popup when you press the right mouse button: Does it come when you click or release the button? Answer: That depend on the application, because win32 don't even handle that stuff.
So if the only standard interface for windows is win32, then I can't mention any major application released within the last 5 years which have used the standard windows interface. Nothing released from Microsoft. Nothing released by Adobe. Not a single 3D editor, nothing released from Apple. Not a single browser. Well nothing (Except google earth, that is the only pure major win32 application I can mention).
Would you call MFC* part of the windows interface? If yes, then how does it differ from Qt? Developers who use MFC/QT both have to download MFC/QT, and both does a static link to (part of) the library, so the user don't have to download anything.
MFC is not part of windows. It comes with Visual studio and you are not allowed to static link to it, unless you have a license to Visual Studio which is in no way part of Windows.
But when that would only work for Windows 7. What about users of Windows XP, and Windows Vista?
I would say that windows don't have a standard GUI application development platform either.
But how do you define standard anyway, and in what way is QT less standard then the "roll your own gui, because Win32 sucks" that so much software(Including anything from Adobe and microsoft office*)
And what is non-standard about QT? I have not seen QT break anything since between the switch between QT3 and QT4. (Unlike MacOS X where an upgrade suddenly required some gui callback methods to be reentrent, breaking lots of software.
*Just take the ribbon interface as an example. They developed an entire new gui system, just for office.