Err, it does manage windows to the extent that it handles creation, deletion, and display. The window manager is just reponsible for movement and ordering.
Re:Usability isnt the issue, Quality is the issue.
on
Fresco M1 Released
·
· Score: 2
First, the ability of windows to have different alpha values *is* transparency (more appropriately translucency). It *is* useless.
This is VERY useful, look at OSX and see how its used, even Windows XP uses it when you move your icons, the icons become transparent so you can see where you are moving them. >>>>>>>>>> Um, how exactly is this useful? First, I doubt you can make out two tiny icon-sized images alpha-blended together. Second, if you're drop target is the size of an icon, there is some serious design error. Besides, Linux has it too. I don't have desktop icons in KDE (I hate icons) but if I did, they'd be translucent when moved.
The cursor also needs to become transparent so that it can have proper shadows and look professional, >>>>>>>>> Um, shadows on cursors looks cheesy, not professional. Are you telling me that before Windows 2000 and it's drop shadow effect, there was no professional looking GUI? Beyond that, you're just wrong on so many levels. All cursors (since like Windows 2.0) have transparency (they're two bitmaps, a color one and a mask). That's why it looks like an arrow rather than a big square. Besides, the XCursor extension (already in XFree CVS) does drop shadows and animation and all that.
the fonts would also look better, along with the windows. >>>>>>>> Um, translucency at the screen level (which X doesn't support) has nothing to do with translucency at the window level (which X does support, via XRender). Translucency at the screen level is needed for window drop-shadows, while translucency at the window level works just fine for anti-aliased text. On top of that, my AA fonts look better in X than in Windows, and a hell of a lot better than OS X.
This isnt about usability, geeks care about usability, WindowsXP isnt the most usable, neither is OSX, you have to balance usability and presentation. >>>>>>> Usability is king. If something can look nice and still be usable, great. If looks interferes with usability, you've fucked up the design.
KDE3 can add all these nice effects but if they all are fake, the whole thing looks like a hacker OS that it still is. >>>>>>>>> How is it fake? I've got transparent menus and cool eye candy in KDE, and I sure as a hell can't tell the difference between it and OS X, aside from the god-damned window drop-shadows. There are two rendering models at work here, and despite what Apple would have you believe, Aqua's isn't better than X's.
X has a model that maps very well to current hardware, and is very fast. Do benchmarks if you don't believe me on the 'fast' part. I've done them, and X whips some ass. As a trade-off for this, it has to implement certain eye-candy tricks as hacks. This is just fine, because eye candy isn't drawn that often anyway. If the user doesn't notice, it's working just fine.
Aqua uses a model that does not map well at all to current hardware. It's coupling to DisplayPDF precludes a lot of hardware acceleration possibilities, and it uses inordinate amounts of RAM. However, it enables certain things like window shadows and the genie effect to be done "naturally."
The major problem with Aqua is it trades "real work" performance for "stupid eye-candy" performance. That's a no-no.
(Note, I'm just comparing the window rendering methods. Aqua's *real* killer feature is the Quartz vector API, which Apple unfortunately does half-assedly by rendering everything to giant bitmaps...)
Things need to look professional, and this is the purpose of eye candy. >>>>>>>>>>. Um, even Apple realized that it had to cut down on eye candy to look more professional. Hence, Aqua Graphite.
Amen to that. I run a 1600x1200 15" LCD screen, and KDE doesn't give me a single bit of trouble with it. Everything looks exactly like it should, only sharper. Well designed web pages look just as great. In comparison, Windows just doesn't handle it as well. It certainly doesn't look bad, but you can definately notice that you're using larger than normal fonts because of layout glitches.
Um, who would these "open source guys" be? The software I write is open sourced. Does that make me an "open source guy"? Do you think I give a fuck what *you* think Linux needs to compete with Windows? Do you think these Berlin guys do either (though, they'd probably phrase it more diplomatically!) They're working on this because they enjoy it. They're not employees of some Open Source Corporation that have product deadlines and focus groups and such shit.
PS> As for software installation, you're using the wrong distro. Software installation for me is a two second (literally) process where I type "emerge " Same thing for Debian users. Silly Windows users spend precious minutes scouring around the net, downloading a program, navigating to the installer, and clicking 'next' a bazillion times before the program is installed!
Heh. I love how Apple has turned people into mindless zombies.
"Quartz Extreme" is'nt analagous to "Quartz." "Quartz Extreme" is a way to hardware accelerate window compositing while Quartz is (mostly) a software renderer to draw stuff inside those windows.
But then again, how would Fresco support games running in a window? >>>>>>>> Most likely how X supports it: bypassing the protocol via something like the DRI.
What, praytell, does X have to do with copying and pasting of images or formatted text, or divx codecs? X is like the GDI. It does drawing and manages windows. That's it! KDE or GNOME provide the desktop environment. If applications don't interoperate at that level, it's the job of KDE and GNOME to fix it.
Why don't you buy from pricewatch? Like getting ripped off at CompUSA? I buy from Pricewatch all the time, have the sense to go to comparably large online retailers (except for RAM!) and have yet (in 3 or 4 years) to get a bad deal a single time!
If you hadn't been paying attention, memory bandwidth has gotten a huge shot in the arm lately. Graphics cards are up in the 20GB/sec and in a couple of months, main memory will be at 6.4 GB/sec (dual channel DDR-400). That's a *huge* jump in memory bandwidth over what was the case only a year ago. No, CPU speeds definately need to go up, because as of late, memory bandwidth has been keeping pace. As for why home users need 3.0 GHz, I can rattle off a list of several things:
1) Not everyone uses just Word. A large chunk of the population actually does demanding work on their PCs, and this includes video/audio editing, 3D, programming, scientific computing, desktop publishing, etc. 2) Gaming! 3) Natural language processing/artificial intelligence. 4) Windows Longhorn and KDE 3.1!
Personally, I really don't see the advantages of Java or C# outside a specific application domain. For heavy systems level work, C++ is superior, because once your needs branch away from what is provided in the standard libraries of Java or.NET, there is a far deeper and broader collection of libraries available for C++. Plus, both Java and C# have an overhead that C++ simply does not have. For applications where performance isn't critical, a *real* high level language is in order. Java and C# claim to be high level, but in all truth, outside of their standard libraries, there is not a whole lot (aside from removed features) in the base language that seperate them from C++. Assuming you have the same libraries for both, and are a competent C++ programming (meaning you use the STL to the point where new/delete are rare in your code) Java/C# code is just as long (or, in the case of Java longer). Python, in comparison, feels like a real high level language. Python code is an order of magnitude shorter than most Java/C# code and much easier to read. Once the new Parrot VM comes out (supposedly it has JIT compiling) these higher level languages will get a significant performance boost, and will be far more appropriate for many applications than Java or C#.
And as a person who's done a fair bit of hardware programming, I'm sick of broken hardware that requires software to go through contortions to get it to work:)
No, they should construct laws around what is right and what is wrong. Unfortunately, right and wrong gradiate into each other, and there is no clear distinction. And remember, "A foolish consistency is the hobgoblin of a small mind."
I'm bent over, take me now! What wider? Sure, anything you want! Long as I get my daily dose of FOX PrimeTime programming, I'll do whatever you want me to!
You know, I think they should be allowed to sell information about you to their affiliates. They just need to redefine 'affiliate' to mean "any company whose CEO has personally slept with with the wife of Verizon's CEO." *That* makes you an 'affiliate' in my book. None of this "if I sell something to you then you're my affiliate" bullshit!
There is a great deal of difference between what meal you order and who you call, and anybody who doesn't realize the distinction is seriously deluded. Remember, proof by analogy is no proof at all. The harsh reality is that life is complicated, and in certain cases, companies should use information about you to optimize their services (for example, like Amazon.com does for your buying habits) and in other cases (especially with something as critical and personal as phone service) they shouldn't. Then you have to take into account historical precedence. Amazon.com doesn't have a history of spamming me with marketing crap. The 'affiliates' of the phone company do. Thus, I'm okay with Amazon.com using my buying habits while I have a problem with the phone company doing the same.
I have no problem with her being a girl. I think teenage guys who act like airheads are equally stupid. Unfortunately, our culture tends to have a bias that tells girls it's okay to act like that, which is why the stereotype is associated with girls.
I was 15 once (and not that many years ago!) and I never used 'like' every five sentences. I don't think I can remember anybody I knew who did either. There is a difference between using slang older people don't approve of (guilty as charged) and not knowing how to speak properly. Have you read high-school student writing lately? It's phenomenally bad. I'm talking pre-literate Amazon rain-forest civilization bad. The fact that kids speak poorly is just a reflection of the fact, that in general, the language skills of our country are heading down the shitter.
I can't claim to know what the developers were thinking, but I can guess at it. They're workstations were running NT4, a Windows platform, so it would be natural for them to use Visual Studio, not necessarily for any advantages it had, but becauase it's the de-facto standard on Windows (and the fact that even with Cygwin, the integration of UNIX tools on Windows just isn't that natural). UNIX on the desktop at that time was even less common than it is now, and thus the developers at the startup were probably much more familiar with Visual Studio than anything else. Thus, it would make sense for them to use what they already knew how ot, rather than have them switch to something different. There are any of a number of explanations of why they used Visual Studio that don't imply that Visual Studio is technically superior.
Because ATI obviously is having problems doing this. ATI's not a small company. They can obviously make some good hardware, so they've got engineering talent. If they've been unable to get a good solid set of drivers together after this long, there is obviously a problem in their driver development team, and studying binary code doesn't seem to be working. If I was NVIDIA, I wouldn't take the risk and give ATI any more help than was necessary.
There is a difference between implementing the whole pipeline in hardware and not having any software support on top. Putting the OpenGL pipeline in hardware means that the 'model' OpenGL pipeline is in hardware. This means everything from transform and lighting to rasterization. It means that the whole API (all of its functionality) is accelerated, not that the API maps directly to the hardware. You need a lot of software support on top to interface with user level applications. The job of this software is to map OpenGL calls to what is best accelerated by hardware. The hardware might have one acceleration primitive that three different OpenGL features all map to. The software also has to optimize the order of OpenGL calls to best take advantage of the hardware. Without this functionality, performance would suck. Not only would the hardware be poorly utilized, but calling the hardware for each API-level call (instead of batching lists and sending them in bulk over the bus) would kill performance. Besides all of that, this software has to manage things totally outside the realm of hardware, like implementing the GLX and WGL APIs, allocating memory on the card, managing DMA buffers, etc.
I have yet to meet template code that gcc 3.1+ won't compile. Particularly, it compiles Loki and Boost just fine, which Visual C++ doesn't. Anything that doesn't compile is a bug, not a known lack of adherence to the standard (unlike with Visual C++, where it's limitations are known standards non-complience). The 'export' keyword is a known exception, because the standard UNIX compile/link model doesn't easily allow for it to be easily implemented. As for EDG and the export keyword, two things: first, EDG isn't a compiler, it's just a front-end. While it supports export, compilers based on it don't. In particular, Intel C++ 6.0 (EDG-based) doesn't support export. Second, the only full compiler that supports 'export' is Comeau C++. While it's great that Visual C++.NET finally replaces the POS that was the Visual C++ 6.x STL, you can't really say that Dinkumware's STL is any better than STLport or even libg++3.2's STL. I wouldn't say g++ falls far short in its support for standard C++. GCC on Linux compiles all of Boost correctly, while Visual C++ 7.0 on Windows fails many of the tests. What's far more of a problem with GCC is that compiling with agressive optimizations (mostly for athlon-xp and P4) can still cause the compiler to choke. What I'd like to see is an example of standards complient C++ code that you've actually used that failed to compile under g++.
Running through binary code isn't exactly easy to do. It's doable when you're trying to reverse engineer how the hardware works, but if you're trying to figure out how NVIDIA does some sophisticated high level OpenGL optimizations, it's decidedly non-trival. And given that the NVIDIA driver is some 7 megabytes of code, including the kernel driver, GLX and XAA modules, and libGL and libGLcore libraries, it would be impossible. Saying that the should open source it because somebody could dissasemble it is just like saying that all software should be open sourced, because all software can be disassembled.
Slashdot Mirror
Err, it does manage windows to the extent that it handles creation, deletion, and display. The window manager is just reponsible for movement and ordering.
First, the ability of windows to have different alpha values *is* transparency (more appropriately translucency). It *is* useless.
This is VERY useful, look at OSX and see how its used, even Windows XP uses it when you move your icons, the icons become transparent so you can see where you are moving them.
>>>>>>>>>>
Um, how exactly is this useful? First, I doubt you can make out two tiny icon-sized images alpha-blended together. Second, if you're drop target is the size of an icon, there is some serious design error. Besides, Linux has it too. I don't have desktop icons in KDE (I hate icons) but if I did, they'd be translucent when moved.
The cursor also needs to become transparent so that it can have proper shadows and look professional,
>>>>>>>>>
Um, shadows on cursors looks cheesy, not professional. Are you telling me that before Windows 2000 and it's drop shadow effect, there was no professional looking GUI? Beyond that, you're just wrong on so many levels. All cursors (since like Windows 2.0) have transparency (they're two bitmaps, a color one and a mask). That's why it looks like an arrow rather than a big square. Besides, the XCursor extension (already in XFree CVS) does drop shadows and animation and all that.
the fonts would also look better, along with the windows.
>>>>>>>>
Um, translucency at the screen level (which X doesn't support) has nothing to do with translucency at the window level (which X does support, via XRender). Translucency at the screen level is needed for window drop-shadows, while translucency at the window level works just fine for anti-aliased text. On top of that, my AA fonts look better in X than in Windows, and a hell of a lot better than OS X.
This isnt about usability, geeks care about usability, WindowsXP isnt the most usable, neither is OSX, you have to balance usability and presentation.
>>>>>>>
Usability is king. If something can look nice and still be usable, great. If looks interferes with usability, you've fucked up the design.
KDE3 can add all these nice effects but if they all are fake, the whole thing looks like a hacker OS that it still is.
>>>>>>>>>
How is it fake? I've got transparent menus and cool eye candy in KDE, and I sure as a hell can't tell the difference between it and OS X, aside from the god-damned window drop-shadows. There are two rendering models at work here, and despite what Apple would have you believe, Aqua's isn't better than X's.
X has a model that maps very well to current hardware, and is very fast. Do benchmarks if you don't believe me on the 'fast' part. I've done them, and X whips some ass. As a trade-off for this, it has to implement certain eye-candy tricks as hacks. This is just fine, because eye candy isn't drawn that often anyway. If the user doesn't notice, it's working just fine.
Aqua uses a model that does not map well at all to current hardware. It's coupling to DisplayPDF precludes a lot of hardware acceleration possibilities, and it uses inordinate amounts of RAM. However, it enables certain things like window shadows and the genie effect to be done "naturally."
The major problem with Aqua is it trades "real work" performance for "stupid eye-candy" performance. That's a no-no.
(Note, I'm just comparing the window rendering methods. Aqua's *real* killer feature is the Quartz vector API, which Apple unfortunately does half-assedly by rendering everything to giant bitmaps...)
Things need to look professional, and this is the purpose of eye candy.
>>>>>>>>>>.
Um, even Apple realized that it had to cut down on eye candy to look more professional. Hence, Aqua Graphite.
Amen to that. I run a 1600x1200 15" LCD screen, and KDE doesn't give me a single bit of trouble with it. Everything looks exactly like it should, only sharper. Well designed web pages look just as great. In comparison, Windows just doesn't handle it as well. It certainly doesn't look bad, but you can definately notice that you're using larger than normal fonts because of layout glitches.
Um, who would these "open source guys" be? The software I write is open sourced. Does that make me an "open source guy"? Do you think I give a fuck what *you* think Linux needs to compete with Windows? Do you think these Berlin guys do either (though, they'd probably phrase it more diplomatically!) They're working on this because they enjoy it. They're not employees of some Open Source Corporation that have product deadlines and focus groups and such shit.
PS> As for software installation, you're using the wrong distro. Software installation for me is a two second (literally) process where I type "emerge " Same thing for Debian users. Silly Windows users spend precious minutes scouring around the net, downloading a program, navigating to the installer, and clicking 'next' a bazillion times before the program is installed!
Heh. I love how Apple has turned people into mindless zombies.
"Quartz Extreme" is'nt analagous to "Quartz." "Quartz Extreme" is a way to hardware accelerate window compositing while Quartz is (mostly) a software renderer to draw stuff inside those windows.
But then again, how would Fresco support games running in a window?
>>>>>>>>
Most likely how X supports it: bypassing the protocol via something like the DRI.
What, praytell, does X have to do with copying and pasting of images or formatted text, or divx codecs? X is like the GDI. It does drawing and manages windows. That's it! KDE or GNOME provide the desktop environment. If applications don't interoperate at that level, it's the job of KDE and GNOME to fix it.
That's why you use this little thing called the 'Internet' to do this little thing called 'Research'
Why don't you buy from pricewatch? Like getting ripped off at CompUSA? I buy from Pricewatch all the time, have the sense to go to comparably large online retailers (except for RAM!) and have yet (in 3 or 4 years) to get a bad deal a single time!
If you hadn't been paying attention, memory bandwidth has gotten a huge shot in the arm lately. Graphics cards are up in the 20GB/sec and in a couple of months, main memory will be at 6.4 GB/sec (dual channel DDR-400). That's a *huge* jump in memory bandwidth over what was the case only a year ago. No, CPU speeds definately need to go up, because as of late, memory bandwidth has been keeping pace. As for why home users need 3.0 GHz, I can rattle off a list of several things:
1) Not everyone uses just Word. A large chunk of the population actually does demanding work on their PCs, and this includes video/audio editing, 3D, programming, scientific computing, desktop publishing, etc.
2) Gaming!
3) Natural language processing/artificial intelligence.
4) Windows Longhorn and KDE 3.1!
Personally, I really don't see the advantages of Java or C# outside a specific application domain. For heavy systems level work, C++ is superior, because once your needs branch away from what is provided in the standard libraries of Java or .NET, there is a far deeper and broader collection of libraries available for C++. Plus, both Java and C# have an overhead that C++ simply does not have. For applications where performance isn't critical, a *real* high level language is in order. Java and C# claim to be high level, but in all truth, outside of their standard libraries, there is not a whole lot (aside from removed features) in the base language that seperate them from C++. Assuming you have the same libraries for both, and are a competent C++ programming (meaning you use the STL to the point where new/delete are rare in your code) Java/C# code is just as long (or, in the case of Java longer). Python, in comparison, feels like a real high level language. Python code is an order of magnitude shorter than most Java/C# code and much easier to read. Once the new Parrot VM comes out (supposedly it has JIT compiling) these higher level languages will get a significant performance boost, and will be far more appropriate for many applications than Java or C#.
And as a person who's done a fair bit of hardware programming, I'm sick of broken hardware that requires software to go through contortions to get it to work :)
No, it's better to do a recall and give all those customers properly working chips!
No, they should construct laws around what is right and what is wrong. Unfortunately, right and wrong gradiate into each other, and there is no clear distinction. And remember, "A foolish consistency is the hobgoblin of a small mind."
I'm bent over, take me now! What wider? Sure, anything you want! Long as I get my daily dose of FOX PrimeTime programming, I'll do whatever you want me to!
You know, I think they should be allowed to sell information about you to their affiliates. They just need to redefine 'affiliate' to mean "any company whose CEO has personally slept with with the wife of Verizon's CEO." *That* makes you an 'affiliate' in my book. None of this "if I sell something to you then you're my affiliate" bullshit!
There is a great deal of difference between what meal you order and who you call, and anybody who doesn't realize the distinction is seriously deluded. Remember, proof by analogy is no proof at all. The harsh reality is that life is complicated, and in certain cases, companies should use information about you to optimize their services (for example, like Amazon.com does for your buying habits) and in other cases (especially with something as critical and personal as phone service) they shouldn't. Then you have to take into account historical precedence. Amazon.com doesn't have a history of spamming me with marketing crap. The 'affiliates' of the phone company do. Thus, I'm okay with Amazon.com using my buying habits while I have a problem with the phone company doing the same.
I have no problem with her being a girl. I think teenage guys who act like airheads are equally stupid. Unfortunately, our culture tends to have a bias that tells girls it's okay to act like that, which is why the stereotype is associated with girls.
Until Mac OS X, there really was no good UI for a desktop Unix-y OS
>>>>>>>>>>
KDE is a great desktop. Suck it Mac-bitch!
I was 15 once (and not that many years ago!) and I never used 'like' every five sentences. I don't think I can remember anybody I knew who did either. There is a difference between using slang older people don't approve of (guilty as charged) and not knowing how to speak properly. Have you read high-school student writing lately? It's phenomenally bad. I'm talking pre-literate Amazon rain-forest civilization bad. The fact that kids speak poorly is just a reflection of the fact, that in general, the language skills of our country are heading down the shitter.
I can't claim to know what the developers were thinking, but I can guess at it. They're workstations were running NT4, a Windows platform, so it would be natural for them to use Visual Studio, not necessarily for any advantages it had, but becauase it's the de-facto standard on Windows (and the fact that even with Cygwin, the integration of UNIX tools on Windows just isn't that natural). UNIX on the desktop at that time was even less common than it is now, and thus the developers at the startup were probably much more familiar with Visual Studio than anything else. Thus, it would make sense for them to use what they already knew how ot, rather than have them switch to something different. There are any of a number of explanations of why they used Visual Studio that don't imply that Visual Studio is technically superior.
Because ATI obviously is having problems doing this. ATI's not a small company. They can obviously make some good hardware, so they've got engineering talent. If they've been unable to get a good solid set of drivers together after this long, there is obviously a problem in their driver development team, and studying binary code doesn't seem to be working. If I was NVIDIA, I wouldn't take the risk and give ATI any more help than was necessary.
There is a difference between implementing the whole pipeline in hardware and not having any software support on top. Putting the OpenGL pipeline in hardware means that the 'model' OpenGL pipeline is in hardware. This means everything from transform and lighting to rasterization. It means that the whole API (all of its functionality) is accelerated, not that the API maps directly to the hardware. You need a lot of software support on top to interface with user level applications. The job of this software is to map OpenGL calls to what is best accelerated by hardware. The hardware might have one acceleration primitive that three different OpenGL features all map to. The software also has to optimize the order of OpenGL calls to best take advantage of the hardware. Without this functionality, performance would suck. Not only would the hardware be poorly utilized, but calling the hardware for each API-level call (instead of batching lists and sending them in bulk over the bus) would kill performance. Besides all of that, this software has to manage things totally outside the realm of hardware, like implementing the GLX and WGL APIs, allocating memory on the card, managing DMA buffers, etc.
I have yet to meet template code that gcc 3.1+ won't compile. Particularly, it compiles Loki and Boost just fine, which Visual C++ doesn't. Anything that doesn't compile is a bug, not a known lack of adherence to the standard (unlike with Visual C++, where it's limitations are known standards non-complience). The 'export' keyword is a known exception, because the standard UNIX compile/link model doesn't easily allow for it to be easily implemented. .NET finally replaces the POS that was the Visual C++ 6.x STL, you can't really say that Dinkumware's STL is any better than STLport or even libg++3.2's STL.
As for EDG and the export keyword, two things: first, EDG isn't a compiler, it's just a front-end. While it supports export, compilers based on it don't. In particular, Intel C++ 6.0 (EDG-based) doesn't support export. Second, the only full compiler that supports 'export' is Comeau C++.
While it's great that Visual C++
I wouldn't say g++ falls far short in its support for standard C++. GCC on Linux compiles all of Boost correctly, while Visual C++ 7.0 on Windows fails many of the tests. What's far more of a problem with GCC is that compiling with agressive optimizations (mostly for athlon-xp and P4) can still cause the compiler to choke. What I'd like to see is an example of standards complient C++ code that you've actually used that failed to compile under g++.
Running through binary code isn't exactly easy to do. It's doable when you're trying to reverse engineer how the hardware works, but if you're trying to figure out how NVIDIA does some sophisticated high level OpenGL optimizations, it's decidedly non-trival. And given that the NVIDIA driver is some 7 megabytes of code, including the kernel driver, GLX and XAA modules, and libGL and libGLcore libraries, it would be impossible. Saying that the should open source it because somebody could dissasemble it is just like saying that all software should be open sourced, because all software can be disassembled.