It's been a long time, but if I recall correctly that was Johnson's price for supporting the space program, when he was Senator. Why should Florida get all the business?
The newer, especially bigger & higher resolution, screens do a pretty large amount of signal processing to turn that old POS television show into visual gold. Among other things most hi-res screens are doing sophisticated upsampling to turn a 480P show into something that looks closer to a 1080P. Otherwise that I Love Lucy re-run would look like crap (although in actuality that's unfair - Lucille Ball insisted on filming her shows in 35MM, not videotape, which now means that her shows have much more visual quality that most shows of that era.)
IMHO it depends on what you're doing with it. I'm running dual 1680x1050 (effectively 3360x1050 total) monitors on my workstation as I type this. If you want fine detail, the resolution really counts. I'm shopping for a new TV, and I will be using it for PC output at least as much as for TeeVee. I've been staring at flat screens at the stores, and from eight feet I can see quite a difference between the 46" and 55", even watching standard TV and movies. One less-obvious item I've noticed is that some 1080P screens don't have fast enough processors, so fast motion shows up as blocky and shifty (technical terms, there.) But the store demos rarely show anything that is very good at really testing the TVs, and when they do it's only on for a second so you can't compare and analyze.
I just read that a rapidly increasing number of movie producers are moving to shooting in 4K. I expect a lot of that will be 3D 4K. Interestingly, the higher the resolution the better compression works.
I'll just add that long ago I used to teach a quality assurance workshop, and the data back then was that black box testing could only every find about 20% of all the bugs. IIRC there is actual theory behind this. Also, given a 'good' software practices methodology (design and code reviews, structured programming, walk throughs, etc.), about 70% of all the bugs are built into the original design (not the code). And, oh yeah - at each stage of the development process from proposal to design to various code releases to alpha, beta and final release, the cost of fixing a bug multiplies by 10. (In the extreme ends, a bug may cost $40,000 to fix after release, while it would have cost the time to move an eraser on the whiteboard in the design review.)
So that people can be charged for use of public restrooms depending on the excrement mass they release. what the hell. lets just put it in streetspeak : for charging people per ounce of shit..............
Satisfaction guaranteed or double your shit back!:D
No, they are complaining that they finally got their car customized just the way they wanted, and now the car maker has come back and said that they are taking the car back and replacing it with one that has five wheels and no speedometer or tach. If they want to have something that does what they need (and already have), they have to spend the time to go back and build a car from the frame up from the parts bin. And it's about loss of power. I've been using X for 30 years, and every new desktop manager in the last 10 years seems to be oriented toward taking away configuration options and the facility with which things could be done and forcing everyone into a certain way of doing things - just as the big commercial-backed distributions have been overlaying proprietary Wysi tools over the OS configuration, making everything except the simplest things harder. I'm surprised that it's still possible to edit/etc/hosts.
Bottom line is that it's about time - I've got sixteen other things to do besides build and maintain a distro myself. But I'm seriously considering it. Many particular desktop features that I used in Gnome 1 were removed - made not difficult but impossible - from Gnome 2. And from the reading, it appears that this is happening again.
I for one have been torn about using Gnome for quite a while - most features of my desktop are based on Compiz and Cairo Dock. I'm probably going to go to an XFCE+compiz desktop, or maybe Compiz alone if I can get it to do what I want.
Perhaps the most frustrating thing is that each of these environments comes with its own library baggage, so if (for example) you want to use Gnome's Terminal then it pulls in a bunch of other cycle burners, and if you want to run a KDE app at the same time, it also pulls in some stuff, and so you are paying in performance for all this stuff that you're not really using.
That's the idea, but I was taking it farther - imagine an 'infinite' number of workspaces in (for example) a spherical shell around you. And, like Enlightenment, you could aim the viewport at any point, not just on workspace boundaries - smooth panning across them. (Like a mapping application). The shake would mean you wouldn't actually have to turn to the right to see the workspace 90 degrees to the 'right'.
... that happens to be the group that creates the applications and systems that the other 99.96% use...
I learned something a long time ago about high tech. The companies that were most successful in tech were the ones that had a pipeline into the most advanced bleeding edge research. For example, a company building oscilloscopes always needs a faster oscilloscope to test the ones they are building.
This continued to be the case in computing when all of the most bleeding edge research was being done in universities. The companies that were most successful were the ones that sponsored university research and had those students come to work for them.
The broad market users are one, two or three generations behind the front edge of the tech. While it is perfectly reasonable, wise and essential for an OS to accommodate such users, it should not be designed to constrain all users to that paradigm - that is permanently turning the product into an also-ran. Today's bleeding edge UI development should be oriented toward management of Big Data and many simultaneous inputs. Single input can easily be accommodated as a special case.
Today, both in uni and in the commercial side, the developers (and some other folks) are the ones who need the performance, the flexibility, the ability to work with complex data sets and multiple applications more than anyone else. The successful 'companies' (taken broadly and including FOSS groups) will be the ones that learn from the real power users, who are pushing their desktops and laptops and tablets and handhelds to the max - not the ones that dumb things down for the common user.
That is not to say that complexity should be visible - some folks like that but most don't. Steve Jobs was all about building complex systems underneath _in order to_ make the user interface seem simple and elegant.
I think the phone/tablet paradigm is merely a stepping stone to eyewear-based displays, where the screen space is effectively unlimited, and apps can be hung on anything in your visual field. That is pretty much the opposite of this full-screen thing.
Compiz and Synergy FTW!:D I've read that it's possible to set up a non-Gnome Compiz desktop where Compiz becomes the window manager. I might try that next.
Mouse wraparound would be nice. Hmmm, it's probably an option..
I am generally running synergy on my desktop, which (if set up that way) does allow wraparound, and also allows your keyboard and mouse to be shared across multiple machines - even linux and windows. In the rare occasions when I have to do some work on the POS windows box in my cube, I like being able to cut and paste between two machines!
Sorry, I misconstrued your remark. The way I work I don't distinguish between instances of the program and windows within the program. In fact I have my dock set to collect all instances of a single program under a single icon. In general those are technically separate processes. Most of these programs support both the 'new window' paradigm (menu item within the program's menu) to launch an instance and the 'click on the icon in the dock' to do the same thing. The big difference is that (for example) using the Terminal's menu will launch the second instance in the same context (directory and history), while clicking the Dock icon will launch a new one with the GUI's default context.
Oh, come on. How common do you think it is for users to want a second instance of an application, rather than just another window? I mean, I've only wanted to do it maybe once or twice in the five or so years I've had this Mac, and I'm very much a power user.
WTF? I _normally_ have three separate Firefox windows, each of which has 1/2 dozen tabs; I _normally_ have two to six GVim windows, each of which may have several frames within. I always have one Terminal windows with three tabs (three different servers), plus I commonly pop up additional Terminals for one-shot tasks so I don't have to interrupt stuff. And that's just on one of four virtual desktops. (I use Compiz desktop cube.)
I just had an idea - many tablets now notice when you rotate the thing 90 degrees, and reformat the screen accordingly. So how about a swift side-to-side shake would move the window to the next thing to the right or left, as appropriate? Then the tablet can interact with an entire borderless worldview. Yes, this is patentable (these days, for good or ill...) Thanks to me, it's now published prior art!:D
Good point. I make a habit of leaving windows not-quite-overlapping, so by mousing over to that visible edge of a covered window, it pops to the front. I have gotten a pretty good working interface with Ubuntu 10.04 and Compiz for virtual deskop duties - I'll stick with Compiz, and apparently it's now possible to use it for the window manager as well.
The problem of having all these different windows is managing them. It's like having a tiny desk with a hundred sheets of paper on top of one another, except those pieces of paper are resizable and highly mutable. Occasionally a post-it note appears out of nowhere on top of your mess to remind you to update your virus scanner.
On the contrary, imagine running a power plant (all those displays and switches!) with one window and a single switch that you have to put in the proper mode to set anything. I gave a lengthy description of my working environment above, but in short, my normal working environment has typically over a dozen 'live' windows and frames-within-windows that I am working with at the same time. I may be editing six different files simultaneously (frames within GVim). I usually have three rolled-up Firefox windows, each of which has 1/2 dozen tabs. And other stuff. And that's just on one face of my virtual desktop cube in Compiz (with dual 1680x1050 screens, times four = a virtual workspace of 13440x1050).
If you have a tiny desktop or a touchpad or whatever, then yes - you are forced to flip back and forth between windows. But IMHO that's not a work environment, that's a surfing-facebook environment. But that would make Compiz much more valuable, as you could put different sets of things on the different faces, and just slide between them.
Sorry, I take exception.:) I'm one of those old farts, and I've been using and advocating multiple screens since at least 1978. Some folks are visual thinkers, some are linear. I'm definitely in the visual group. Those others, I think, are in the linear group. And Emacs has supported multiple windows since the beginning, IIRC. So even the text-mode types are not necessarily linear either.
In 1979 I was using three graphics terminals side by side (each of them 640x480 to 1280x1024). I hacked up a custom RS232 switch to direct output from the mainframe output to each one as needed, while input to the mainframe was always from one of them for the keyboard. On one terminal I had the code I was editing, on the second was my command line interaction, and on the third was the 3D graphics output that resulted from running the code.
Today in my normal workspace (a Compiz cube on dual 1680x1050 monitors) I have four virtual 3360x1050 desktops, all visible in the background in my transparent cube (when there aren't working windows in the way). I can spin the cube with one middle click & pan. The first desktop has housekeeping - mail, timeclock, Pidgin, sometimes a web page open, sometimes a terminal as I deal with email and office matters. My second 'working desktop' has one (sometimes two) Terminal, usually with three tabs for three different machines I'm logged into, two GVim windows one of which is broken into from one to several separate subwindows (vertically and horizontally) for different class files and the other of which contains one to three output log files. At any given time there may be diffs of log files or diffs of code files. Then, because I don't have a third screen, I keep three Firefox windows rolled up except when I'm using them, each of which has several tabs. One of the three, visible on all four sides of the virtual desktop, contains database interfaces for two machines (phpMyAdmin), dotProject, Trac, Mercurial, and the development portal. The second contains tabs for various sorts of documentation, the third contains reference material for the project I'm working on - usually web pages that I'm either scraping or reviewing.
If I'm working on more than one project this week, I will have a similar setup open on the third face of the cube, and the fourth usually has some more casual stuff such as a webpage that shows Slashdot, the news, etc.
I'm seriously considering going to a third screen (and 1920x1200 monitors), so I don't have to flip between windows for the Firefox stuff and the logfile views. Why should I have to flip between windows instead of just scanning my eyes over to the right? I want CONTEXT, dammit!:D I guess my workspace is more analogous to the bridge of a ship than a computer terminal. There's a lot going on, and I want access to all of it right now, and a visual indication of everything that's going on while I work on each individual task.
If you have sufficient resolution, the only reason to use a single window full screen is to get the maximum amount of data for one application on there, temporarily. I sometimes do this with an editing file, because I need just 'one more line' of text for context.
I think the ideal progression would be to stop trying to squeeze everything into a single screen, and instead make that screen a true viewport into an unlimited virtual space. As we move to head-up displays, we should be able to hang a window anywhere in space. The real world is a big space that surrounds us - why not our 'desktop' as well? And why can't I read a virtual newspaper the same way I do a real one, with the full spread visible and readable? And other parts of my environment visible around it - the stove, the clock, the coffee pot, my SO, etc.
I'm talking out my hat here, since I have not done anything with Android or iOS, but the following would work for any generic *nix.
In Linux, BSD or any *nix (does iOS run a form of BSD, like Mac OSX does?), one can make any program the shell (the thing that comes up when you log in). So as soon as the tablet boots up, it will just be running the web client.
The problem with that would be how much the web client needs the graphical user interface login. If it needs that, then you can make rsh or another restricted shell environment, which only allows the programs that you specify in a chrooted directory.
There are probably some complications to getting such a thing set up right to run the embedded things, but I would presume that the Android (which supposedly has a Linux kernel) and possibly iPad tablets could be set up that way.
Also, I think some versions of Linux have a 'kiosk' mode that does most or all of what you want.
Then, to prevent going outside, configure the tablets to only work with a single WiFi access point, and that access point is connected to your data server, and that server is not connected to anything else. The user would have to login via an external boot loader or debugger to get around the OS and modify the WiFi configuration.
You might want to have another application running in the background that screams bloody murder if the tablet gets out of range of the WiFi access point. Maybe something like 'beep... beep... Beep.. Beep.. Please return to the office... BEEPBEEPBEEP HELP I'M BEING REMOVED FROM THE OFFICE!!!! HELP!! HELP!! HELP!! AIEEEE!!! '
I think "Do Not Track" will only stop (or slow down) the war of escalation between advertisers. With Do Not Track, Google et al can sell 'more targeted ads' to advertisers that are willing to pay more for that level of targeting. If we assume that the average user/consumer is only going to click on some maximum number of ads per day then the advertising business on the internet is basically a zero-sum game (which I think is pretty close to true at this point). So if we continue to allow tracking, those advertisers benefit at the expense of those who don't spend the extra money.
So with tracking, Google makes more money, advertisers spend more for the same aggregate sales, users lose privacy. Google makes plenty of money without tracking.
Therefore IMHO tracking has no net benefit to anyone but Google.
Slashdot Mirror
It's been a long time, but if I recall correctly that was Johnson's price for supporting the space program, when he was Senator. Why should Florida get all the business?
Yep, all of them do that.
The newer, especially bigger & higher resolution, screens do a pretty large amount of signal processing to turn that old POS television show into visual gold. Among other things most hi-res screens are doing sophisticated upsampling to turn a 480P show into something that looks closer to a 1080P. Otherwise that I Love Lucy re-run would look like crap (although in actuality that's unfair - Lucille Ball insisted on filming her shows in 35MM, not videotape, which now means that her shows have much more visual quality that most shows of that era.)
IMHO it depends on what you're doing with it. I'm running dual 1680x1050 (effectively 3360x1050 total) monitors on my workstation as I type this. If you want fine detail, the resolution really counts. I'm shopping for a new TV, and I will be using it for PC output at least as much as for TeeVee. I've been staring at flat screens at the stores, and from eight feet I can see quite a difference between the 46" and 55", even watching standard TV and movies. One less-obvious item I've noticed is that some 1080P screens don't have fast enough processors, so fast motion shows up as blocky and shifty (technical terms, there.) But the store demos rarely show anything that is very good at really testing the TVs, and when they do it's only on for a second so you can't compare and analyze.
I just read that a rapidly increasing number of movie producers are moving to shooting in 4K. I expect a lot of that will be 3D 4K. Interestingly, the higher the resolution the better compression works.
$5 DVD bin at WalMart FTW! :D Just gotta pay attention to the format - widescreen, fullscreen, etc.
I'll just add that long ago I used to teach a quality assurance workshop, and the data back then was that black box testing could only every find about 20% of all the bugs. IIRC there is actual theory behind this. Also, given a 'good' software practices methodology (design and code reviews, structured programming, walk throughs, etc.), about 70% of all the bugs are built into the original design (not the code). And, oh yeah - at each stage of the development process from proposal to design to various code releases to alpha, beta and final release, the cost of fixing a bug multiplies by 10. (In the extreme ends, a bug may cost $40,000 to fix after release, while it would have cost the time to move an eraser on the whiteboard in the design review.)
So that people can be charged for use of public restrooms depending on the excrement mass they release. .............
what the hell. lets just put it in streetspeak :
for charging people per ounce of shit.
Satisfaction guaranteed or double your shit back! :D
sorry, moderated by accident, replying to cancel.
No, they are complaining that they finally got their car customized just the way they wanted, and now the car maker has come back and said that they are taking the car back and replacing it with one that has five wheels and no speedometer or tach. If they want to have something that does what they need (and already have), they have to spend the time to go back and build a car from the frame up from the parts bin. And it's about loss of power. I've been using X for 30 years, and every new desktop manager in the last 10 years seems to be oriented toward taking away configuration options and the facility with which things could be done and forcing everyone into a certain way of doing things - just as the big commercial-backed distributions have been overlaying proprietary Wysi tools over the OS configuration, making everything except the simplest things harder. I'm surprised that it's still possible to edit /etc/hosts.
Bottom line is that it's about time - I've got sixteen other things to do besides build and maintain a distro myself. But I'm seriously considering it. Many particular desktop features that I used in Gnome 1 were removed - made not difficult but impossible - from Gnome 2. And from the reading, it appears that this is happening again.
I for one have been torn about using Gnome for quite a while - most features of my desktop are based on Compiz and Cairo Dock. I'm probably going to go to an XFCE+compiz desktop, or maybe Compiz alone if I can get it to do what I want.
Perhaps the most frustrating thing is that each of these environments comes with its own library baggage, so if (for example) you want to use Gnome's Terminal then it pulls in a bunch of other cycle burners, and if you want to run a KDE app at the same time, it also pulls in some stuff, and so you are paying in performance for all this stuff that you're not really using.
That's the idea, but I was taking it farther - imagine an 'infinite' number of workspaces in (for example) a spherical shell around you. And, like Enlightenment, you could aim the viewport at any point, not just on workspace boundaries - smooth panning across them. (Like a mapping application). The shake would mean you wouldn't actually have to turn to the right to see the workspace 90 degrees to the 'right'.
40% of 1% of the computer market.
I learned something a long time ago about high tech. The companies that were most successful in tech were the ones that had a pipeline into the most advanced bleeding edge research. For example, a company building oscilloscopes always needs a faster oscilloscope to test the ones they are building.
This continued to be the case in computing when all of the most bleeding edge research was being done in universities. The companies that were most successful were the ones that sponsored university research and had those students come to work for them.
The broad market users are one, two or three generations behind the front edge of the tech. While it is perfectly reasonable, wise and essential for an OS to accommodate such users, it should not be designed to constrain all users to that paradigm - that is permanently turning the product into an also-ran. Today's bleeding edge UI development should be oriented toward management of Big Data and many simultaneous inputs. Single input can easily be accommodated as a special case.
Today, both in uni and in the commercial side, the developers (and some other folks) are the ones who need the performance, the flexibility, the ability to work with complex data sets and multiple applications more than anyone else. The successful 'companies' (taken broadly and including FOSS groups) will be the ones that learn from the real power users, who are pushing their desktops and laptops and tablets and handhelds to the max - not the ones that dumb things down for the common user.
That is not to say that complexity should be visible - some folks like that but most don't. Steve Jobs was all about building complex systems underneath _in order to_ make the user interface seem simple and elegant.
I think the phone/tablet paradigm is merely a stepping stone to eyewear-based displays, where the screen space is effectively unlimited, and apps can be hung on anything in your visual field. That is pretty much the opposite of this full-screen thing.
Compiz and Synergy FTW! :D
I've read that it's possible to set up a non-Gnome Compiz desktop where Compiz becomes the window manager. I might try that next.
Mouse wraparound would be nice. Hmmm, it's probably an option..
I am generally running synergy on my desktop, which (if set up that way) does allow wraparound, and also allows your keyboard and mouse to be shared across multiple machines - even linux and windows. In the rare occasions when I have to do some work on the POS windows box in my cube, I like being able to cut and paste between two machines!
Sorry, I misconstrued your remark. The way I work I don't distinguish between instances of the program and windows within the program. In fact I have my dock set to collect all instances of a single program under a single icon. In general those are technically separate processes. Most of these programs support both the 'new window' paradigm (menu item within the program's menu) to launch an instance and the 'click on the icon in the dock' to do the same thing. The big difference is that (for example) using the Terminal's menu will launch the second instance in the same context (directory and history), while clicking the Dock icon will launch a new one with the GUI's default context.
Oh, come on. How common do you think it is for users to want a second instance of an application, rather than just another window? I mean, I've only wanted to do it maybe once or twice in the five or so years I've had this Mac, and I'm very much a power user.
WTF? I _normally_ have three separate Firefox windows, each of which has 1/2 dozen tabs; I _normally_ have two to six GVim windows, each of which may have several frames within. I always have one Terminal windows with three tabs (three different servers), plus I commonly pop up additional Terminals for one-shot tasks so I don't have to interrupt stuff. And that's just on one of four virtual desktops. (I use Compiz desktop cube.)
I just had an idea - many tablets now notice when you rotate the thing 90 degrees, and reformat the screen accordingly. So how about a swift side-to-side shake would move the window to the next thing to the right or left, as appropriate? Then the tablet can interact with an entire borderless worldview. Yes, this is patentable (these days, for good or ill...) Thanks to me, it's now published prior art! :D
Good point. I make a habit of leaving windows not-quite-overlapping, so by mousing over to that visible edge of a covered window, it pops to the front. I have gotten a pretty good working interface with Ubuntu 10.04 and Compiz for virtual deskop duties - I'll stick with Compiz, and apparently it's now possible to use it for the window manager as well.
The problem of having all these different windows is managing them. It's like having a tiny desk with a hundred sheets of paper on top of one another, except those pieces of paper are resizable and highly mutable. Occasionally a post-it note appears out of nowhere on top of your mess to remind you to update your virus scanner.
On the contrary, imagine running a power plant (all those displays and switches!) with one window and a single switch that you have to put in the proper mode to set anything. I gave a lengthy description of my working environment above, but in short, my normal working environment has typically over a dozen 'live' windows and frames-within-windows that I am working with at the same time. I may be editing six different files simultaneously (frames within GVim). I usually have three rolled-up Firefox windows, each of which has 1/2 dozen tabs. And other stuff. And that's just on one face of my virtual desktop cube in Compiz (with dual 1680x1050 screens, times four = a virtual workspace of 13440x1050).
If you have a tiny desktop or a touchpad or whatever, then yes - you are forced to flip back and forth between windows. But IMHO that's not a work environment, that's a surfing-facebook environment. But that would make Compiz much more valuable, as you could put different sets of things on the different faces, and just slide between them.
I had my own reply to someone farther up, but in short: "Me too!" :D
Sorry, I take exception. :) I'm one of those old farts, and I've been using and advocating multiple screens since at least 1978. Some folks are visual thinkers, some are linear. I'm definitely in the visual group. Those others, I think, are in the linear group. And Emacs has supported multiple windows since the beginning, IIRC. So even the text-mode types are not necessarily linear either.
In 1979 I was using three graphics terminals side by side (each of them 640x480 to 1280x1024). I hacked up a custom RS232 switch to direct output from the mainframe output to each one as needed, while input to the mainframe was always from one of them for the keyboard. On one terminal I had the code I was editing, on the second was my command line interaction, and on the third was the 3D graphics output that resulted from running the code.
Today in my normal workspace (a Compiz cube on dual 1680x1050 monitors) I have four virtual 3360x1050 desktops, all visible in the background in my transparent cube (when there aren't working windows in the way). I can spin the cube with one middle click & pan. The first desktop has housekeeping - mail, timeclock, Pidgin, sometimes a web page open, sometimes a terminal as I deal with email and office matters. My second 'working desktop' has one (sometimes two) Terminal, usually with three tabs for three different machines I'm logged into, two GVim windows one of which is broken into from one to several separate subwindows (vertically and horizontally) for different class files and the other of which contains one to three output log files. At any given time there may be diffs of log files or diffs of code files. Then, because I don't have a third screen, I keep three Firefox windows rolled up except when I'm using them, each of which has several tabs. One of the three, visible on all four sides of the virtual desktop, contains database interfaces for two machines (phpMyAdmin), dotProject, Trac, Mercurial, and the development portal. The second contains tabs for various sorts of documentation, the third contains reference material for the project I'm working on - usually web pages that I'm either scraping or reviewing.
If I'm working on more than one project this week, I will have a similar setup open on the third face of the cube, and the fourth usually has some more casual stuff such as a webpage that shows Slashdot, the news, etc.
I'm seriously considering going to a third screen (and 1920x1200 monitors), so I don't have to flip between windows for the Firefox stuff and the logfile views. Why should I have to flip between windows instead of just scanning my eyes over to the right? I want CONTEXT, dammit! :D I guess my workspace is more analogous to the bridge of a ship than a computer terminal. There's a lot going on, and I want access to all of it right now, and a visual indication of everything that's going on while I work on each individual task.
If you have sufficient resolution, the only reason to use a single window full screen is to get the maximum amount of data for one application on there, temporarily. I sometimes do this with an editing file, because I need just 'one more line' of text for context.
I think the ideal progression would be to stop trying to squeeze everything into a single screen, and instead make that screen a true viewport into an unlimited virtual space. As we move to head-up displays, we should be able to hang a window anywhere in space. The real world is a big space that surrounds us - why not our 'desktop' as well? And why can't I read a virtual newspaper the same way I do a real one, with the full spread visible and readable? And other parts of my environment visible around it - the stove, the clock, the coffee pot, my SO, etc.
I'm talking out my hat here, since I have not done anything with Android or iOS, but the following would work for any generic *nix.
In Linux, BSD or any *nix (does iOS run a form of BSD, like Mac OSX does?), one can make any program the shell (the thing that comes up when you log in). So as soon as the tablet boots up, it will just be running the web client.
The problem with that would be how much the web client needs the graphical user interface login. If it needs that, then you can make rsh or another restricted shell environment, which only allows the programs that you specify in a chrooted directory.
There are probably some complications to getting such a thing set up right to run the embedded things, but I would presume that the Android (which supposedly has a Linux kernel) and possibly iPad tablets could be set up that way.
Also, I think some versions of Linux have a 'kiosk' mode that does most or all of what you want.
Then, to prevent going outside, configure the tablets to only work with a single WiFi access point, and that access point is connected to your data server, and that server is not connected to anything else. The user would have to login via an external boot loader or debugger to get around the OS and modify the WiFi configuration.
You might want to have another application running in the background that screams bloody murder if the tablet gets out of range of the WiFi access point. Maybe something like 'beep ... beep ... Beep .. Beep .. Please return to the office ... BEEPBEEPBEEP HELP I'M BEING REMOVED FROM THE OFFICE!!!! HELP!! HELP!! HELP!! AIEEEE!!! '
I think "Do Not Track" will only stop (or slow down) the war of escalation between advertisers. With Do Not Track, Google et al can sell 'more targeted ads' to advertisers that are willing to pay more for that level of targeting. If we assume that the average user/consumer is only going to click on some maximum number of ads per day then the advertising business on the internet is basically a zero-sum game (which I think is pretty close to true at this point). So if we continue to allow tracking, those advertisers benefit at the expense of those who don't spend the extra money.
So with tracking, Google makes more money, advertisers spend more for the same aggregate sales, users lose privacy. Google makes plenty of money without tracking.
Therefore IMHO tracking has no net benefit to anyone but Google.
"Don't tax me, don't tax thee, tax that man behind the tree!" - attributed to Russell B. Long, circa 1950-1960, describing "tax reform".
Not only that, it also violates my patent on slimy business practices.
Sorry, there's 5000 years of prior art WRT your patent! :P