Could anybody give some details on how this was done? How do you hack multiplayer into a closed source game that doesn't have multiplayer support it? Decompile the thing? Find some unused engine hooks that allow multiplayer? Something completely different?
This won't be the future of movies (unless you count movies about this tech like Strange Days), but the future of Americas Funniest Home Videos. We are already see this trend with cars, some countries require dash cams for insurance purpose and thus we have a rise of all kinds of car videos on the net. Head mountable cameras like the GoPro also already do the same thing for sports and recreational activities.
But anyway, I consider all those to be side effects, that will not be why people are wearing those cameras, it's just stuff that will happen when people wear those cameras. Real reason to wear those will probably be as memory enhancers and just like for cars as insurance, a theft might have a much harder time getting away when he is filmed by dozens of glasses on the street and the eye whiteness of the future might also be one with glasses on his head. With glasses you can start recording your life 24/7 non-stop, it will no longer be about making a picture, but simply about tagging a moment in the recording of your life. 99.99% of that footage will of course always useless and never be watched again (who has the time to rewatch their life after all?), but the remaining 0.01% could turn out quite useful if the right software to search it is provided.
Another interesting application for those glasses could be Telepresence, if you have multiple people wearing those glasses, it could be possible to link them up and then you see a life video feed of what is happening somewhere else. Could be quite useful in some situations to jump into somebodies head that way and guide them through some problem.
We currently use around 12 class A networks per year of which there are only 255 in total (many of which are unrelocatable due to being reserved for localhost, multicast and so on) . Whenever you hear people complaining about IBM or whoever holding a large chunk of IP addresses, that refers to a single class A network. So getting IBM, HP or Xerox to restructure their network and give back their IPs would buy you one month each time. There aren't a whole lot of companies holding class A networks, so you could at maximum get probably 2 years or so, realistically much less.
A little extra time to shake out the bugs from any infrastructure upgrade seems couldn't hurt, too.
We already had 14 years to do that, another one or two won't make a difference. IPv6 doesn't need time, it needs something that forces people to make the switch, running out of IPv4 seems to slowly building up to be that force.
If the P2P software uploaded 9,250 copies of each song to other users, then Thomas isn't paying more than retail.
Yes, if he did upload that much. Except there is no proof that he uploaded that much and not only that, it's also highly implausible. 9250 copies per song, 15 songs, a song say 3MB in size and we are talking about 400GB upload volume. That so much that it should be not only unlikely, but proofable impossible for the average user.
Try backing up your 3TB HDD over a 1Mbit/s upload link, it will take you a year, at 1Gbit/s you might be able to do it overnight if you drive is fast enough. The point with super fast Internet isn't that you can now watch Youtube a little better, but that it will allow applications that would have been impossible before. And yes, that of course includes things such as P2P, as with 1Gbit/s up and downstream, a content addressable anonymous network such as Freenet could get really interesting and provide some serious competition to services like Youtube, as you no longer would need to have a central service do the hosting, it could be hosted by everybody. It could bring the Internet back to what it once was, a network between equal peers, not between producers and consumers.
There's no way to measure whether a measurement has been performed. However there's a way to determine whether the measurement result has been predetermined by the state before the measurement.
But aren't those the same thing? Say you have two physicists. One does his little quantum teleportation experiment and writes down the states of the photons. Then he hands of the photons to another physicist, but doesn't tell him that the photons come from a teleportation experiment. The second guy now does all those fancy other experiments to check if they have a predefined state. So how can the second guy come out negative, but the first guy can have all the states written down on a piece of paper?
Only when you make the measurement does this happen.
How do we know that? Is there any way to figure out if a quanta has bean measured or not? Don't think so, as otherwise we could use it to transmit information via quantum teleportation, which we can't.
Yep, and in turn: Give every GUI tool a meaningful command line interface. The CLI isn't the enemy, but you should you be forced to use it. But neither should you be forced to use the GUI for tasks that could be easily done from CLI (e.g. batch conversion of documents).
There is nothing fundamentally wrong with the Linux desktop, there are icons, you click on them and they make programs start, it's not rocket science, it doesn't need to. But there are tons of small issues and interoperability problems. Just fix them one by one instead of constantly reinventing the wheel.
Things will get even more interesting when it comes to 'printing' DNA. DNA synthesis is becoming quite affordable, which means in the not so distant future people can recreate any virus they like on the cheap. Will be tricky to control that.
If enough people conclude you're right, your way is incorporated.
The problem is that that approach only works for monolithic software. When it comes to Linux on the desktop almost none of the core issue are just in a monolithic piece of of software, they are all in the communication between different pieces of software and in the base libraries. When every application uses it's own file dialog, you can't just fix that with a patch, you first need some agreement on what the proper file dialog should look like and all the applications have to agree on it before you can start sending out patches. Same is true for drag&drop, applets, package management and all that other stuff. The problem isn't that application X has an issue, but that application X and Y behave slightly different, neither of which behaves wrong, it's them being different what causes the problems.
Freedesktop.org has done plenty of good in bring things closer together, but from where we are now, to a fully unified desktop experience is a pretty damn long way, especially as right now we don't seem to have anybody actually doing conformance testing (i.e. XDG Base Directory Spec is all fine and good, yet most apps still store their config files in ~/.appname/ and not ~/.config/appname/). I think it would also help when those specs would all get a reference implementation, as while confirming to the XDG Base Directory Spec is reasonable simple, if that spec ever needs to get changed, you are back at square one and have thousands of apps to fix, instead of just a single library to adjust.
Tons and tons of games were released from 95-01 for that OS and sadly waaaay too many relied on "hacks" to boost the then weak performance of the systems.
That's easy to fix with emulation.
and of course the guts of Win9X were such a hodge podge and have so many copyrights and patents to deal with a DOSBox style solution for Win9X will most likely be impossible.
Wine already exist and if you don't like that, you can just use a real Windows95 and install it on a PC emulator.
As long as the CD don't corrupt before somebody makes a backup, Window95 games won't have an issue, as preserving them and playing them is not very difficult.
The topic you complain of (complexity of installing software) is a topic that can be mastered in very little time.
Just because you know how to workaround an issue doesn't make that issue go away. Software installation, outside of the main repositories, is a complete night mare in Linux and it doesn't get any better when you know everything there is to know about how to install software, you still will end up wasting tons and tons of time and fix problems that shouldn't even exist in the first place.
The readers and players are part of the experience, just like they joysticks, keyboard and all the rest of the hardware. Just backing up the code and throwing it in an emulator works reasonably well, but you lose a lot in the process and unless you have actual hardware to compare to, you can't even be sure your emulator is even working correctly.
I would be hugely pissed if some amateur took those poems and transmogrified them into some horrible high school poetry... and released the changed text under my name.
For the avoidance of doubt, You may only use the credit required by this Section for the purpose of attribution in the manner set out above and, by exercising Your rights under this License, You may not implicitly or explicitly assert or imply any connection with, sponsorship or endorsement by the Original Author, Licensor and/or Attribution Parties, as appropriate, of You or Your use of the Work, without the separate, express prior written permission of the Original Author, Licensor and/or Attribution Parties.
It is this latter one that I feel warrants the "no commercial" verbiage, even today.
Zinga is big enough, they can take care of themselves and just recreate the artwork. The problems actually comes more from the smaller developers, who will grab "free" things from the web, use them in their iPhone apps and then sell them on the AppStore without ever giving to much considerations about the license. While I haven't yet used NC due to all problems it will cause with Linux distributions, that kind of stuff happened to me often enough to seriously consider it. It wouldn't completely stop it, but it would make it much easier to just point at the license and say "You are not allowed to do that", as CC-by-sa leaves to many confusing questions (i.e. You use SA sprites and backgrounds in a game, is the game Share-Alike?). I would also really like it when we somehow could mush CC-by-sa together with the GPL, as I really hate it to have my artwork used in non-Open Source games, but there really is right now no good way to prevent that right now with CC licenses.
How exactly is anyone on this team supposed to get an anti-user feature embedded into the VLC codebase past the eyes of the rest of the team?
See the Debian OpenSSH incident. A very high profile software package with an obvious and easy to detect bug that rendered all newly installed machines wide open, yet nobody found it for two years. So how did they get past the OpenSSH maintainers? Easy, they didn't, Debian people introduced the bug while packaging OpenSSH, where none of the original developers where looking. There are a lot of orphaned and abandoned Open Source packages around that one could adopt, introduce a backdoor and then slip into a distribution without anybody noticing.
Most of the time the trouble is of course not even security, but just basic consistency. Does an app store it's config data in ~/.appname, in ~/config/appname, maybe in ~/.local/share/appname/? Or how about maybe in ~/.gnome/appname/ or in the depths of ~/.kde/? With a proper sandbox model that's not a questions you would ever need to ask, because the sandbox would only allow a single location. File system hierarchy would thus be no longer just a policy that an application could adopt or not, but something that is actively enforced.
Has everyone forgotten the days when your computer actually belonged to you?
To make a counterpoint: We need a certain level of restrictions so that the computer actually does what I want instead of does what the application wants. Current OSs offer very little in the way of actually restricting applications. If you execute an application from a third party, it can do quite literally what it wants. Even Open Source doesn't help much as you have no time to audit it all. At best it might not have root rights, but that still doesn't stop it from searching through your personal photo collection, your credit card info, your mail and all that stuff. In contrast you have Apples AppStore, it's a walled garden with all it's problems, but it also comes with a sandbox. If you execute an App from the AppStore it can't do anything to your system, as it's stuck in it's little box. It can't get your photos, credit card info or anything like that unless you explicitly allow it. The AppStore thus shift control away from the application and back to the user where it belongs.
The problem already starts long before death, as how to you even share those books and other digital goods with the rest of your family? Do you give your Kindle device to your kids? Do they get their own but reuse your account? What if they get their own account? What about games on Steam, etc.? A lot of digital services right now don't really have a clean way of sharing digital goods with the rest of the family and many form of sharing might be considered a TOS violation. What about when it comes to a divorce who gets the library? All those issues start long before death comes into play, they essentially start right when you want to interact with another human being.
Ok, please explain under what kind of circumstances constant switching between Metro fullscreen apps and regular windowed desktop apps makes any kind of sense. I wouldn't mind it much when Metro apps could just be started in a window like everything else, but clicking on a PDF and then being forced into fullscreen without even a way out really serves no purpose whatsoever, it's just distracting and annoying.
The problem with DLC isn't one of money, but that it fractures the game. Instead of one well designed product that fits together, you have a dozens of products that sort of kind of fit together. Bonus DLC weapons are often out of balance, the story DLC often doesn't quite fit in, multiplayer maps just make it more troublesome to find people to play with and costumes are just useless. "Don't buy" isn't a solution, as the main game still might reference the missing content, so when you don't buy it, you still know there is a piece missing from the game that makes it feel incomplete and sometimes you don't even know you got it, as it automatically comes with your Game of the Year edition. There is also a general lack of quality in DLC, as most of it ends up being far less quality then the main game.
Sometimes there is of course DLC that takes more the form of an expansion pack and truly does tell a piece of story that is completely optional to the main game and still doesn't feel like a useless addition, but even those can be a little annoying, as DLC means they have to be short and cheap. Thus even when they are actually good, they just end up feeling short and a full blown expansion pack would have been better.
In the end DLC is nothing but a money grab and nobody can honestly tell me that it's somehow the best way to deliver a game to give the user the best experience. DLC has done nothing but make gaming more complicated and frustrating. From a business standpoint it makes sense, from an entertainment point of view it's a really shitty idea.
"Back in the day" you didn't have voice acting,
Voice acting has been a normal feature since 1993, it has been around for a while. Network play existed even earlier and plenty games on my Amiga had Modem support. 1080? Nothing all to special, a good CRT could give you 1280x1024 long before any HD-TV was around.
But the majority of the PC gaming market are not those people.
I think even the majority of PC gamers wouldn't mind the ability to continue playing their games on the go on their phone or have the ability to check out a game demo within seconds instead of an hour to download a few gigabytes. Or how about playing that latest Crysis game without having to by a new PC? OnLive would have allowed all of that.
The problem with OnLive was simply lack of integration, when you have to compete with Steam, than sure as hell you will have a hard time. If OnLive would have been simply a part of Steam that you could use when you want to, I am sure as hell PC gamers would have loved it. Another problem was that they simply a little to early, bandwidth are still to limited to make high quality video streaming possible for everybody. They also entered at a bad point into the PC market when most PC games were targeted for 7 year old consoles and thus had very low hardware requirements.
Either way, OnLive as a company might be toast, but I that kind of streaming technology is here to stay and I wouldn't be surprised if another company will be successful with it in the not so distance future.
JavaScript can be an ugly language to work with, but the interactive editor they have looks really good and nicely works around all the issue one frequently runs into when using raw JavaScript (catches missing semicolon, catches typos of function names, allows editing color values via GUI, automatically runs the code on each change, etc.).
Slashdot Mirror
Could anybody give some details on how this was done? How do you hack multiplayer into a closed source game that doesn't have multiplayer support it? Decompile the thing? Find some unused engine hooks that allow multiplayer? Something completely different?
This won't be the future of movies (unless you count movies about this tech like Strange Days), but the future of Americas Funniest Home Videos. We are already see this trend with cars, some countries require dash cams for insurance purpose and thus we have a rise of all kinds of car videos on the net. Head mountable cameras like the GoPro also already do the same thing for sports and recreational activities.
But anyway, I consider all those to be side effects, that will not be why people are wearing those cameras, it's just stuff that will happen when people wear those cameras. Real reason to wear those will probably be as memory enhancers and just like for cars as insurance, a theft might have a much harder time getting away when he is filmed by dozens of glasses on the street and the eye whiteness of the future might also be one with glasses on his head. With glasses you can start recording your life 24/7 non-stop, it will no longer be about making a picture, but simply about tagging a moment in the recording of your life. 99.99% of that footage will of course always useless and never be watched again (who has the time to rewatch their life after all?), but the remaining 0.01% could turn out quite useful if the right software to search it is provided.
Another interesting application for those glasses could be Telepresence, if you have multiple people wearing those glasses, it could be possible to link them up and then you see a life video feed of what is happening somewhere else. Could be quite useful in some situations to jump into somebodies head that way and guide them through some problem.
how much longer would they have?
We currently use around 12 class A networks per year of which there are only 255 in total (many of which are unrelocatable due to being reserved for localhost, multicast and so on) . Whenever you hear people complaining about IBM or whoever holding a large chunk of IP addresses, that refers to a single class A network. So getting IBM, HP or Xerox to restructure their network and give back their IPs would buy you one month each time. There aren't a whole lot of companies holding class A networks, so you could at maximum get probably 2 years or so, realistically much less.
A little extra time to shake out the bugs from any infrastructure upgrade seems couldn't hurt, too.
We already had 14 years to do that, another one or two won't make a difference. IPv6 doesn't need time, it needs something that forces people to make the switch, running out of IPv4 seems to slowly building up to be that force.
If the P2P software uploaded 9,250 copies of each song to other users, then Thomas isn't paying more than retail.
Yes, if he did upload that much. Except there is no proof that he uploaded that much and not only that, it's also highly implausible. 9250 copies per song, 15 songs, a song say 3MB in size and we are talking about 400GB upload volume. That so much that it should be not only unlikely, but proofable impossible for the average user.
Try backing up your 3TB HDD over a 1Mbit/s upload link, it will take you a year, at 1Gbit/s you might be able to do it overnight if you drive is fast enough. The point with super fast Internet isn't that you can now watch Youtube a little better, but that it will allow applications that would have been impossible before. And yes, that of course includes things such as P2P, as with 1Gbit/s up and downstream, a content addressable anonymous network such as Freenet could get really interesting and provide some serious competition to services like Youtube, as you no longer would need to have a central service do the hosting, it could be hosted by everybody. It could bring the Internet back to what it once was, a network between equal peers, not between producers and consumers.
Why not ban text descriptions, or ban stories which encourage child abuse?
They are already banned, see Manga Collection Ruled “Child Pornography By US Court.
There's no way to measure whether a measurement has been performed. However there's a way to determine whether the measurement result has been predetermined by the state before the measurement.
But aren't those the same thing? Say you have two physicists. One does his little quantum teleportation experiment and writes down the states of the photons. Then he hands of the photons to another physicist, but doesn't tell him that the photons come from a teleportation experiment. The second guy now does all those fancy other experiments to check if they have a predefined state. So how can the second guy come out negative, but the first guy can have all the states written down on a piece of paper?
Only when you make the measurement does this happen.
How do we know that? Is there any way to figure out if a quanta has bean measured or not? Don't think so, as otherwise we could use it to transmit information via quantum teleportation, which we can't.
Yep, and in turn: Give every GUI tool a meaningful command line interface. The CLI isn't the enemy, but you should you be forced to use it. But neither should you be forced to use the GUI for tasks that could be easily done from CLI (e.g. batch conversion of documents).
There is nothing fundamentally wrong with the Linux desktop, there are icons, you click on them and they make programs start, it's not rocket science, it doesn't need to. But there are tons of small issues and interoperability problems. Just fix them one by one instead of constantly reinventing the wheel.
Things will get even more interesting when it comes to 'printing' DNA. DNA synthesis is becoming quite affordable, which means in the not so distant future people can recreate any virus they like on the cheap. Will be tricky to control that.
If enough people conclude you're right, your way is incorporated.
The problem is that that approach only works for monolithic software. When it comes to Linux on the desktop almost none of the core issue are just in a monolithic piece of of software, they are all in the communication between different pieces of software and in the base libraries. When every application uses it's own file dialog, you can't just fix that with a patch, you first need some agreement on what the proper file dialog should look like and all the applications have to agree on it before you can start sending out patches. Same is true for drag&drop, applets, package management and all that other stuff. The problem isn't that application X has an issue, but that application X and Y behave slightly different, neither of which behaves wrong, it's them being different what causes the problems.
Freedesktop.org has done plenty of good in bring things closer together, but from where we are now, to a fully unified desktop experience is a pretty damn long way, especially as right now we don't seem to have anybody actually doing conformance testing (i.e. XDG Base Directory Spec is all fine and good, yet most apps still store their config files in ~/.appname/ and not ~/.config/appname/). I think it would also help when those specs would all get a reference implementation, as while confirming to the XDG Base Directory Spec is reasonable simple, if that spec ever needs to get changed, you are back at square one and have thousands of apps to fix, instead of just a single library to adjust.
Seriously, what's the problem?
Here is a nice and detailed List of Major Linux Problems.
Tons and tons of games were released from 95-01 for that OS and sadly waaaay too many relied on "hacks" to boost the then weak performance of the systems.
That's easy to fix with emulation.
and of course the guts of Win9X were such a hodge podge and have so many copyrights and patents to deal with a DOSBox style solution for Win9X will most likely be impossible.
Wine already exist and if you don't like that, you can just use a real Windows95 and install it on a PC emulator.
As long as the CD don't corrupt before somebody makes a backup, Window95 games won't have an issue, as preserving them and playing them is not very difficult.
The topic you complain of (complexity of installing software) is a topic that can be mastered in very little time.
Just because you know how to workaround an issue doesn't make that issue go away. Software installation, outside of the main repositories, is a complete night mare in Linux and it doesn't get any better when you know everything there is to know about how to install software, you still will end up wasting tons and tons of time and fix problems that shouldn't even exist in the first place.
The readers and players are part of the experience, just like they joysticks, keyboard and all the rest of the hardware. Just backing up the code and throwing it in an emulator works reasonably well, but you lose a lot in the process and unless you have actual hardware to compare to, you can't even be sure your emulator is even working correctly.
I would be hugely pissed if some amateur took those poems and transmogrified them into some horrible high school poetry... and released the changed text under my name.
That's already covered by the license:
It is this latter one that I feel warrants the "no commercial" verbiage, even today.
Zinga is big enough, they can take care of themselves and just recreate the artwork. The problems actually comes more from the smaller developers, who will grab "free" things from the web, use them in their iPhone apps and then sell them on the AppStore without ever giving to much considerations about the license. While I haven't yet used NC due to all problems it will cause with Linux distributions, that kind of stuff happened to me often enough to seriously consider it. It wouldn't completely stop it, but it would make it much easier to just point at the license and say "You are not allowed to do that", as CC-by-sa leaves to many confusing questions (i.e. You use SA sprites and backgrounds in a game, is the game Share-Alike?). I would also really like it when we somehow could mush CC-by-sa together with the GPL, as I really hate it to have my artwork used in non-Open Source games, but there really is right now no good way to prevent that right now with CC licenses.
How exactly is anyone on this team supposed to get an anti-user feature embedded into the VLC codebase past the eyes of the rest of the team?
See the Debian OpenSSH incident. A very high profile software package with an obvious and easy to detect bug that rendered all newly installed machines wide open, yet nobody found it for two years. So how did they get past the OpenSSH maintainers? Easy, they didn't, Debian people introduced the bug while packaging OpenSSH, where none of the original developers where looking. There are a lot of orphaned and abandoned Open Source packages around that one could adopt, introduce a backdoor and then slip into a distribution without anybody noticing.
Most of the time the trouble is of course not even security, but just basic consistency. Does an app store it's config data in ~/.appname, in ~/config/appname, maybe in ~/.local/share/appname/? Or how about maybe in ~/.gnome/appname/ or in the depths of ~/.kde/? With a proper sandbox model that's not a questions you would ever need to ask, because the sandbox would only allow a single location. File system hierarchy would thus be no longer just a policy that an application could adopt or not, but something that is actively enforced.
Has everyone forgotten the days when your computer actually belonged to you?
To make a counterpoint: We need a certain level of restrictions so that the computer actually does what I want instead of does what the application wants. Current OSs offer very little in the way of actually restricting applications. If you execute an application from a third party, it can do quite literally what it wants. Even Open Source doesn't help much as you have no time to audit it all. At best it might not have root rights, but that still doesn't stop it from searching through your personal photo collection, your credit card info, your mail and all that stuff. In contrast you have Apples AppStore, it's a walled garden with all it's problems, but it also comes with a sandbox. If you execute an App from the AppStore it can't do anything to your system, as it's stuck in it's little box. It can't get your photos, credit card info or anything like that unless you explicitly allow it. The AppStore thus shift control away from the application and back to the user where it belongs.
The problem already starts long before death, as how to you even share those books and other digital goods with the rest of your family? Do you give your Kindle device to your kids? Do they get their own but reuse your account? What if they get their own account? What about games on Steam, etc.? A lot of digital services right now don't really have a clean way of sharing digital goods with the rest of the family and many form of sharing might be considered a TOS violation. What about when it comes to a divorce who gets the library? All those issues start long before death comes into play, they essentially start right when you want to interact with another human being.
Ok, please explain under what kind of circumstances constant switching between Metro fullscreen apps and regular windowed desktop apps makes any kind of sense. I wouldn't mind it much when Metro apps could just be started in a window like everything else, but clicking on a PDF and then being forced into fullscreen without even a way out really serves no purpose whatsoever, it's just distracting and annoying.
So don't buy it?
The problem with DLC isn't one of money, but that it fractures the game. Instead of one well designed product that fits together, you have a dozens of products that sort of kind of fit together. Bonus DLC weapons are often out of balance, the story DLC often doesn't quite fit in, multiplayer maps just make it more troublesome to find people to play with and costumes are just useless. "Don't buy" isn't a solution, as the main game still might reference the missing content, so when you don't buy it, you still know there is a piece missing from the game that makes it feel incomplete and sometimes you don't even know you got it, as it automatically comes with your Game of the Year edition. There is also a general lack of quality in DLC, as most of it ends up being far less quality then the main game.
Sometimes there is of course DLC that takes more the form of an expansion pack and truly does tell a piece of story that is completely optional to the main game and still doesn't feel like a useless addition, but even those can be a little annoying, as DLC means they have to be short and cheap. Thus even when they are actually good, they just end up feeling short and a full blown expansion pack would have been better.
In the end DLC is nothing but a money grab and nobody can honestly tell me that it's somehow the best way to deliver a game to give the user the best experience. DLC has done nothing but make gaming more complicated and frustrating. From a business standpoint it makes sense, from an entertainment point of view it's a really shitty idea.
"Back in the day" you didn't have voice acting,
Voice acting has been a normal feature since 1993, it has been around for a while. Network play existed even earlier and plenty games on my Amiga had Modem support. 1080? Nothing all to special, a good CRT could give you 1280x1024 long before any HD-TV was around.
But the majority of the PC gaming market are not those people.
I think even the majority of PC gamers wouldn't mind the ability to continue playing their games on the go on their phone or have the ability to check out a game demo within seconds instead of an hour to download a few gigabytes. Or how about playing that latest Crysis game without having to by a new PC? OnLive would have allowed all of that.
The problem with OnLive was simply lack of integration, when you have to compete with Steam, than sure as hell you will have a hard time. If OnLive would have been simply a part of Steam that you could use when you want to, I am sure as hell PC gamers would have loved it. Another problem was that they simply a little to early, bandwidth are still to limited to make high quality video streaming possible for everybody. They also entered at a bad point into the PC market when most PC games were targeted for 7 year old consoles and thus had very low hardware requirements.
Either way, OnLive as a company might be toast, but I that kind of streaming technology is here to stay and I wouldn't be surprised if another company will be successful with it in the not so distance future.
JavaScript can be an ugly language to work with, but the interactive editor they have looks really good and nicely works around all the issue one frequently runs into when using raw JavaScript (catches missing semicolon, catches typos of function names, allows editing color values via GUI, automatically runs the code on each change, etc.).