Sort of, but when you focus on words or letters you're only looking for milliseconds at a time. Even when you are staring your eyes tend to move around (saccades). It's difficult to explain if you haven't done it, but if you have a chance, use an eye tracker as an input device for more than 10 minutes and you'll notice how tired you get.
Perhaps yes. I think the other problem is that rendering engines are generally not optimized to render a given spatial area at a high resolution, while other areas are rendered at low res. I think rendering quality is usually specified on a per-object basis. So 3d engines would have to be substantially reorganized to take advantage of that kind of thing..
Oh dear.. so much blind buyin to the Microsoft marketing.
>> No mention or acknowledgement of WPF/WCF or the new APIs that are and 'set' to replace Win32/Win64
Those API's haven't been proven. Yes they look promising, but until there are a significant set of apps out there using them fully, we won't know if they're any good. And there are signs of that not everything is peachy there: http://www.istartedsomething.com/20071206/yahoo-messenger-vista-launches/
>> Go to Channel 10 and watch videos on why XAML/XPS was created and how it trumps every aspect of other display/print technologies
Shocking that microsoft would say their technologies are better than the competition.
>> Let alone how it is an integrated aspect of the video API system in Vista, making programming freaky simple for advanced features and new UI platforms like 3D.
Blah blah blah. We heard plenty of hype about how the compositing in Vista was so wickedly advanced and it allowed effects that couldn't be done on other platforms. That may or may not be true, I haven't studied it in detail so I'm not going to say. And yet in reality it doesn't make a damn bit of difference. Vista doesn't have anything that doesn't exist on other platforms as well, requiring less resources. So even if it is more advanced, the actual benefits are not there.
>> doesn't mention OS Xs lack of keyboard support
Bullshit and you know it.
>> Office 2007 is a new direction in GUI paradigms, and is WELL accepted in the business world
And you know this via your crystal ball? Stop making statements you can't back up.
>> Menus were a hack to make features available in a GUI context, but are a draw back to non-graphical UIs.
Armchair usability experts are funny...
>> Let alone OS X is still a hybrid 64bit OS, using 32bit code throughout the OS, unlike Vista x64
Of course every windows app is still 32 bit, so your 64bit OS doesn't really make much difference in performance. The only way to really take advantage of 64 bit these days is to run Linux, where ~95% of your apps will be native 64bit (aside from some proprietary apps).
I think you're right in that.NET is a huge step up for Windows devs. But it's still basically for windows devs, and it still requires a virtual machine. Even though the rhetoric is that performance doesn't matter in most cases, every single project I've worked on came to a point where performance was important, and it would have been extremely difficult to achieve the desired level of performance with a JIT compiled language. The other problem is that you're more or less locked onto windows platforms. Yes, mono exists, but it's not an official port and the legal situation is still hazy, so I would never rely on it for real projects delivered to real clients.
So all my stuff is done with C++/Qt. I get C++ speed,.NET/Java style simplicity in the framework, seamless portability to all desktop and now also some embedded platforms (Linux/WinCE), and good platform integration. For my own business I pay for the proprietary license, and at work we use the open source version and write OSS apps (we're a non-profit). Works fine for us.. The license fee is peanuts anyway compared to the time savings in development and cross platform porting effort.
Except that combination is not a shovel. I wouldn't use GCC/Vim, instead preferring an IDE like qdevelop, but Qt is the most productive way to do anything truly cross platform these days. Java lacks integration and requires a VM,.NET is not very cross platform and requires a VM as well, and scripting languages lack integration, performance, and features for large scale development.
Qt with C++ (or Java/Python/Ruby) is really the only choice left. For that I gladly pay the license fee for my proprietary development. I can't be bothered to use tools that I have to fight with.
Yep, the effect you are describing is called the "midas touch problem" (for obvious reasons) and is the major problem with any eye tracking interface. The other major problem is that eyes are not meant to be used to select things (especially small things), and doing so will result in extreme fatigue very quickly.
Yes, this has been done and works quite well (the user doesn't notice any difference). The problem is in the reaction time that is necessary. The last study I read found that the high detail rendering must be performed within 5ms of a fixation to make the experience seamless to the user. That's a problem for most applications, as they won't be able to react that quickly.
It will definitely interfere. Depending on the glasses (reflectivity and material of the frames), the environment (ambient light, glare), and the system, you will get different results. I have some experience with a $40000 eye tracker from Tobii (they're pretty much top of the line trackers) and it still has issues with glasses.
Well I work on these kinds of systems, and that 5mm is not a limitation of the system, it's a limitation of the eyes.
The fovea (dense area of rods and cones) in the retina is large enough to give you approximately a 1 deg cone of "focus". Which means depending on the distance, you can focus on an area of a given size on the screen. So even with a perfect eye tracker, you cannot pinpoint gaze location exactly just by measuring eye orientation. Accuracy depends on distance from the screen, but 5mm is in the ballpark for what you can achieve (and that is with a perfectly calibrated system, real accuracy will be worse).
Yeah, and +5 insightful too. All with an argument of "tech is getting cheaper, therefore yahoo is undervalued" which makes about no sense whatsoever. Yahoo's customer's have nothing to do with their overhead costs in the data center. And technology advances will reduce the cost for everyone, not just Yahoo. There will be more customers total, but nothing says they will be on Yahoo.
You mean in the title bar of the apps? I guess so, but I don't see how that will confuse anyone. By the time you have the application open you will know what it is.
I guess it would be ok if it was named something like KDE File Manager, but then people get pissed about everything being KDE Something, just like the pseudo standard of kname was ridiculed for so long.
If you just name it "File manager" or "Archive manager" then you can't keep different apps apart. Sure most people should only have one of each type of application installed, but realistically you might have two different apps for the same task due to one app doing a task better than another one. So now you have the gnome File Manager and the kde File Manager with the same name and you can't keep them apart.
And where do you draw the line? Photo Editor? Well there are tons of those with different capabilities. Or Web Browser? Well same problem. For some basic utilities it might work, but the more complex the app gets, the less useful a generic name will be. Anyway, people on other platforms don't seem to be getting confused by meaningless application names, so I don't see the problem.
Yay for comparing a random screenshot of an alpha to a carefully put together screenshot of a released version, with different sizes and levels of detail. Lots can be seen from that!
I think the 4.0.x releases were a mistake, but 4.1 snapshots are a lot better already, to the point where I wouldn't want to go back to anything else. Give it another go when 4.1 comes out.
>> My favourite apps are all Linux; digikam, amarok, k3b but non-kde 4 versions of these are a dead end now.
What are you talking about? Amarok 1.4.9 was released very recently, Digikam is still getting a release for KDE3, and k3b works just fine. None of those apps are released for KDE4 yet. The premier versions are still KDE3 apps.
When the KDE4 versions are released, you can run them in your KDE3 desktop with no problems. I don't see what you're complaining about.
That's the preview for plasma themes. It only shows up when you click the combo box, so screen space usage is irrelevant. Not great, but better than no preview.
So you gazed into your crystal ball and figured out that 4.1 will not be stable? Since it hasn't been released that's quite the accomplishment. I use a snapshot of 4.1 at work, and I haven't seen it crash in weeks.
Partially due to incomplete dialog design on the snapshot they are testing. This isn't released software. Also it depends on what you're comparing it to. Mac UI also has a lot of empty space, mostly because of the icon+text toolbars.
>> Why the faded titles in the panel? What are they intended to signify?
I don't see this anywhere in the link you gave. I think in the ars screenshot, it's just a bug in the theme.
>> Why are the minimise and maximise icons raised, tiny and 'stuck on' rather graphically integrated into the window title?
Yeah I don't like those either. I usually change the titlebar to something more sensible, like Plastique.
>> Why is the panel so g i g a n t i c?
Not sure, but it's been like that for ages (in the KDE 3.x series as well). I always resize it much smaller.
Or they could use the keyboard and forego that mouse they're having so much difficulty with completely. So you're saying you haven't. They certainly don't know about the keyboard shortcuts to do this stuff, and it's also not nearly as flexible as using the mouse.
However, I don't see the relevance of this to being able to drag windows out of the visible desktop, something the post I responded to was trying to pass off as "basic window management". If a part of your window does not fit on the screen, then moving it so you can see it is basic window management. Also moving/resizing it with a bigger hit area is pretty basic. I use it all the time.
No, it's better when the software does the right thing (ie: makes it impossible to create oversized dialog boxes and makes it impossible to move them somewhere where they can't be seen) Impossible to create oversized dialog boxes? For what screen size? That makes no sense. THere will always be devices that have limited resolution like the Eee and not all standard software will fit on it. That's unavoidable, and you can't expect developers to consider a screen only 480 pixels high, because that would make the experience worse for most other users.
So if a dialog is too big, you need a way to view it. The fact is, Linux has an easy way to do that (could be even easier) and Windows makes it impossible.
>> seeing no reason to switch from a fully patched IE 7 running as non-administrator on Vista
Not to start a flame war, but security is really the least of my reasons for choosing Firefox over IE. Firefox is faster, more standards compliant, has way better functionality and flexibility due to extensions.
>> In no way, shape, or form, is that "basic window management".
Ever seen an elderly person trying to hit a small target with the mouse? It's just about impossible. The title bar is ok, but trying to hit the window borders to resize a window is damn near impossible for some people. KDE (probably gnome too) makes that dead easy. ALT+Drag anywhere on the window to move it, and ALT+right drag anywhere to resize it.
>> Indeed, there's a pretty strong UI argument that it shouldn't even be possible.
Riiight.. It sure is great when the software knows better and makes it impossible to use the system. No trust me, It's better this way, you don't know what's good for you.
What are you on? On my glossy laptop screen, I can see the light behind me, and it's pretty annoying. On my matte LCD, I can't see any reflections, which is way better. There is no glare from that same light.
"Light from any vector to the user will create glare" ?? What are you smoking?
Slashdot Mirror
Sort of, but when you focus on words or letters you're only looking for milliseconds at a time. Even when you are staring your eyes tend to move around (saccades). It's difficult to explain if you haven't done it, but if you have a chance, use an eye tracker as an input device for more than 10 minutes and you'll notice how tired you get.
Perhaps yes. I think the other problem is that rendering engines are generally not optimized to render a given spatial area at a high resolution, while other areas are rendered at low res. I think rendering quality is usually specified on a per-object basis. So 3d engines would have to be substantially reorganized to take advantage of that kind of thing..
Oh dear.. so much blind buyin to the Microsoft marketing.
>> No mention or acknowledgement of WPF/WCF or the new APIs that are and 'set' to replace Win32/Win64
Those API's haven't been proven. Yes they look promising, but until there are a significant set of apps out there using them fully, we won't know if they're any good. And there are signs of that not everything is peachy there: http://www.istartedsomething.com/20071206/yahoo-messenger-vista-launches/
>> Go to Channel 10 and watch videos on why XAML/XPS was created and how it trumps every aspect of other display/print technologies
Shocking that microsoft would say their technologies are better than the competition.
>> Let alone how it is an integrated aspect of the video API system in Vista, making programming freaky simple for advanced features and new UI platforms like 3D.
Blah blah blah. We heard plenty of hype about how the compositing in Vista was so wickedly advanced and it allowed effects that couldn't be done on other platforms. That may or may not be true, I haven't studied it in detail so I'm not going to say. And yet in reality it doesn't make a damn bit of difference. Vista doesn't have anything that doesn't exist on other platforms as well, requiring less resources. So even if it is more advanced, the actual benefits are not there.
>> doesn't mention OS Xs lack of keyboard support
Bullshit and you know it.
>> Office 2007 is a new direction in GUI paradigms, and is WELL accepted in the business world
And you know this via your crystal ball? Stop making statements you can't back up.
>> Menus were a hack to make features available in a GUI context, but are a draw back to non-graphical UIs.
Armchair usability experts are funny...
>> Let alone OS X is still a hybrid 64bit OS, using 32bit code throughout the OS, unlike Vista x64
Of course every windows app is still 32 bit, so your 64bit OS doesn't really make much difference in performance. The only way to really take advantage of 64 bit these days is to run Linux, where ~95% of your apps will be native 64bit (aside from some proprietary apps).
I think you're right in that .NET is a huge step up for Windows devs. But it's still basically for windows devs, and it still requires a virtual machine. Even though the rhetoric is that performance doesn't matter in most cases, every single project I've worked on came to a point where performance was important, and it would have been extremely difficult to achieve the desired level of performance with a JIT compiled language.
.NET/Java style simplicity in the framework, seamless portability to all desktop and now also some embedded platforms (Linux/WinCE), and good platform integration. For my own business I pay for the proprietary license, and at work we use the open source version and write OSS apps (we're a non-profit). Works fine for us.. The license fee is peanuts anyway compared to the time savings in development and cross platform porting effort.
The other problem is that you're more or less locked onto windows platforms. Yes, mono exists, but it's not an official port and the legal situation is still hazy, so I would never rely on it for real projects delivered to real clients.
So all my stuff is done with C++/Qt. I get C++ speed,
Except that combination is not a shovel. I wouldn't use GCC/Vim, instead preferring an IDE like qdevelop, but Qt is the most productive way to do anything truly cross platform these days. Java lacks integration and requires a VM, .NET is not very cross platform and requires a VM as well, and scripting languages lack integration, performance, and features for large scale development.
Qt with C++ (or Java/Python/Ruby) is really the only choice left. For that I gladly pay the license fee for my proprietary development. I can't be bothered to use tools that I have to fight with.
Yep, the effect you are describing is called the "midas touch problem" (for obvious reasons) and is the major problem with any eye tracking interface. The other major problem is that eyes are not meant to be used to select things (especially small things), and doing so will result in extreme fatigue very quickly.
Yes, this has been done and works quite well (the user doesn't notice any difference). The problem is in the reaction time that is necessary. The last study I read found that the high detail rendering must be performed within 5ms of a fixation to make the experience seamless to the user. That's a problem for most applications, as they won't be able to react that quickly.
It will definitely interfere. Depending on the glasses (reflectivity and material of the frames), the environment (ambient light, glare), and the system, you will get different results. I have some experience with a $40000 eye tracker from Tobii (they're pretty much top of the line trackers) and it still has issues with glasses.
Well I work on these kinds of systems, and that 5mm is not a limitation of the system, it's a limitation of the eyes.
The fovea (dense area of rods and cones) in the retina is large enough to give you approximately a 1 deg cone of "focus". Which means depending on the distance, you can focus on an area of a given size on the screen. So even with a perfect eye tracker, you cannot pinpoint gaze location exactly just by measuring eye orientation. Accuracy depends on distance from the screen, but 5mm is in the ballpark for what you can achieve (and that is with a perfectly calibrated system, real accuracy will be worse).
Yeah, and +5 insightful too. All with an argument of "tech is getting cheaper, therefore yahoo is undervalued" which makes about no sense whatsoever. Yahoo's customer's have nothing to do with their overhead costs in the data center. And technology advances will reduce the cost for everyone, not just Yahoo. There will be more customers total, but nothing says they will be on Yahoo.
I guess a whiner will always find something to bitch about..
You mean in the title bar of the apps? I guess so, but I don't see how that will confuse anyone. By the time you have the application open you will know what it is.
I guess it would be ok if it was named something like KDE File Manager, but then people get pissed about everything being KDE Something, just like the pseudo standard of kname was ridiculed for so long.
If you just name it "File manager" or "Archive manager" then you can't keep different apps apart. Sure most people should only have one of each type of application installed, but realistically you might have two different apps for the same task due to one app doing a task better than another one. So now you have the gnome File Manager and the kde File Manager with the same name and you can't keep them apart.
And where do you draw the line? Photo Editor? Well there are tons of those with different capabilities. Or Web Browser? Well same problem. For some basic utilities it might work, but the more complex the app gets, the less useful a generic name will be. Anyway, people on other platforms don't seem to be getting confused by meaningless application names, so I don't see the problem.
Yay for comparing a random screenshot of an alpha to a carefully put together screenshot of a released version, with different sizes and levels of detail. Lots can be seen from that!
I think the 4.0.x releases were a mistake, but 4.1 snapshots are a lot better already, to the point where I wouldn't want to go back to anything else. Give it another go when 4.1 comes out.
>> My favourite apps are all Linux; digikam, amarok, k3b but non-kde 4 versions of these are a dead end now.
What are you talking about? Amarok 1.4.9 was released very recently, Digikam is still getting a release for KDE3, and k3b works just fine. None of those apps are released for KDE4 yet. The premier versions are still KDE3 apps.
When the KDE4 versions are released, you can run them in your KDE3 desktop with no problems. I don't see what you're complaining about.
That's the preview for plasma themes. It only shows up when you click the combo box, so screen space usage is irrelevant. Not great, but better than no preview.
So you gazed into your crystal ball and figured out that 4.1 will not be stable? Since it hasn't been released that's quite the accomplishment. I use a snapshot of 4.1 at work, and I haven't seen it crash in weeks.
>> Why such vast tracts of grey?
Partially due to incomplete dialog design on the snapshot they are testing. This isn't released software. Also it depends on what you're comparing it to. Mac UI also has a lot of empty space, mostly because of the icon+text toolbars.
>> Why the faded titles in the panel? What are they intended to signify?
I don't see this anywhere in the link you gave. I think in the ars screenshot, it's just a bug in the theme.
>> Why are the minimise and maximise icons raised, tiny and 'stuck on' rather graphically integrated into the window title?
Yeah I don't like those either. I usually change the titlebar to something more sensible, like Plastique.
>> Why is the panel so g i g a n t i c?
Not sure, but it's been like that for ages (in the KDE 3.x series as well). I always resize it much smaller.
>> Dolphin? Why not just call it 'file browser'?
It is called the file browser in the menu.
So if a dialog is too big, you need a way to view it. The fact is, Linux has an easy way to do that (could be even easier) and Windows makes it impossible.
>> seeing no reason to switch from a fully patched IE 7 running as non-administrator on Vista
Not to start a flame war, but security is really the least of my reasons for choosing Firefox over IE.
Firefox is faster, more standards compliant, has way better functionality and flexibility due to extensions.
>> In no way, shape, or form, is that "basic window management".
Ever seen an elderly person trying to hit a small target with the mouse? It's just about impossible. The title bar is ok, but trying to hit the window borders to resize a window is damn near impossible for some people. KDE (probably gnome too) makes that dead easy. ALT+Drag anywhere on the window to move it, and ALT+right drag anywhere to resize it.
>> Indeed, there's a pretty strong UI argument that it shouldn't even be possible.
Riiight.. It sure is great when the software knows better and makes it impossible to use the system.
No trust me, It's better this way, you don't know what's good for you.
>> there are dialog windows that run off the screen in Linux apps too.
Of course on Linux you can easily hold the ALT key and drag the window to make the buttons visible. Not possible on windows without third party hacks.
What are you on? On my glossy laptop screen, I can see the light behind me, and it's pretty annoying. On my matte LCD, I can't see any reflections, which is way better. There is no glare from that same light.
"Light from any vector to the user will create glare" ?? What are you smoking?
On the other hand their risk of dying in a fire will decrease drastically..