And then again, they cost $$$$$, because PHB's keep on buying them.
Tapes themselves are incredible cheap, in fact they are the cheapest storage available right now beating both HDDs and DVD-Rs by costing half or a third as much. It's the drive price that is killing it and make them useless for the average consumer. That of course makes HDDs the medium of choice for backup, but it still kind of irks me that we don't have a cheap backup solution right now that keeps drive and media separate (well, DVD-R do that, but are to small to be practical).
I just don't get why they are so keen to get rid off that faeture...
They are not. This is not about replacing X, it's simply about splitting the graphics and input code out of X and moving it over to the kernel and Wayland. You can still run X11 apps under Wayland. So it's really more a pragmatic approach to getting the Xorg code based cleaned up, then an abandonment of network transparencey.
Also in general X11 network transparency is bit overrated in my opinion, as while it is good for things such as basic thin client computing, it is completely useles for basic everyday uses like screen sharing or moving applications from one display to another. It's also rather useless for multimedia, be it video due to the lack of bandwidth or sound due to the X11 simply not handling sound. Wayland won't solve them, but a clean code base means that it will be easier for other people to attack those problems.
I am an advocate of organisations having a degree of responsibility for anything they intend to profit from
Thinking like this leads straight to cooperation enforced censorship, as companies like to avoid risk and thus stuff like the Anonymous Coward function you used to post this, would very likely be the first things to go. Easy account creation as well, so welcome Facebook login everywhere. That's not even theoretically speaking, over here in Germany we have tons of shitty rules like this, one of the results of that is that free anonymous Wifi essentially does not exist over here, because it's simply to risky to offer it.
I think what you are asking for now is simply impossible.
Yes, that's the point. As said, it's not a hard problem, just adding relocation support to binaries would solve 99% of the problems. Current binary packaging infrastructure simply doesn't do that, manually compiling from source does. That's why it's trivial to keep as many different versions of Gimp around when I compile manually from source, while it's impossible to do the same with binary packages. That's why the current way to package binaries in Linux sucks.
And that's been my point, what you want exists in Linux.
You don't get it. What should be there is the ability to "apt-get install debian" on a Ubuntu system. That's not the case. That does not exist. Yes, you could build it, but nobody has done so.
The idea of a distribution after the initial layer has always been to offer a collection of packages designed to work together. In other words to make choice.
Um, no. The idea of distributions was never to make a choice and force it on it's user, it was simply to provide easy installable binaries for Linux software and most distributions try to offer as much software as they can with the man power they have available. Thus you don't just get KDE or Gnome, but XFCE, FVWM, TWM, WindowMaker and dozens of other window manager or desktop environments all at the same time. Distributions might pick what they install by default and what they throw most of their money at to fine tune it, but that's about it, you are always free to install something else if you don't like the default.
In a perfect world a distribution would simply be a meta-package, a collection of software for a given task or taste, that you could freely mix and match with other distributions. But right now, that's not the case, instead each distribution has to repackage all the software you ever might want to use again and there is no cross-distribution compatibility due to the inflexibilities of the package manager (i.e. install Xubuntu on Ubuntu works, installing Ubuntu on Debian does not, as the package manager wouldn't be able to handle the conflicts). You can't even install old Ubuntu packages on a new Ubuntu system without running into all kinds of trouble.
And no, there is no distribution that does what I want, even Gentoo's SLOT system is little more then a glorified way to avoid renaming packages, it still doesn't allow free mixing and matching of different versions.
The closest thing to what I want is actually outside of the Linux world: Fink and MacPorts. They still fail at all the advanced handling of namespaces as they use the same tools distributions on Linux do, but thanks to them being an "add-on" to the system, not the core, they don't try to take control over "/" instead they go to "/fink" or "/macports" and one can simply install and use both at the same time without conflict. That small change of not polluting the global namespace, but keeping things in a separate directory adds a lot of flexibility, a proper system for handling namespaces could of course go much further.
At a certain point its just easier to switch distributions when it comes to a vertical package like this.
Yes, and that's exactly why I say the package managers we have today are crap. There is nothing in principle that would prevent you from stuffing everything that depends on Gnome2 in one namespace and everything that wants Gnome3 in another. But our current package managers don't support any kind of clean namespace separation, everything goes into one tree and when there are two packages with the same name, only one can continue to exist, the other gets overwritten or needs to be manually renamed by the package maintainer. "Switch the distribution" should simply never be the answer to a Linux problem.
It is however annoying, and labor intensive.
Compared to all the time and effort that gets wasted on the users side, it's pretty minimal.
That's not the fault of the package manager. Gnome is designed for integration.
And what the fuck is "designed for integration" supposed to mean? Throwing buzzwords around isn't an argument. Give me some examples of where there would be insurmountable hurdles to allowing Gnome2 and Gnome3 to exist in parallel. I have all the RAM and HDD space I need for that. So why can't I do it? Why do I have to switch the whole distribution to get that, when the distribution doesn't really care about the DE I am using (i.e. I can replace Gnome3 with XFCE just fine)?
At the source level it is completely trivial to install software of different versions in parallel, you just configure it with a different install prefix. Relocatable binaries require a tiny bit of extra work, as most software doesn't support relocation and relies and hard compiled path, so one would need to fix that. At worst you might also need to adjust some Dbus names in case the Gnome people broke the API. But there is really no magic to it, just fix the naming conflicts and call it a day. The only reason we don't have that is because nobody bothered to implement it and instead continues to do things the way they have always been done on Linux, barf all over/usr/, shut your eyes and hope for the best.
I understand but that has always been there because having absolute open levels DOES NOT MAKE SENSE.
It am not arguing about the actual level design, just about it's presentation. When you had a door in Doom, you could walk into it. If you see a door in a modern game, there is like a 80% chance that you can't walk into it. The reason for that is simply that when Doom needed a wall, the designers used a wall. In modern games that try to present realistic locations however the developers don't use walls, as that would look unrealistic, instead they use houses, cars, trashcans or any other object they can think of. Houses however have doors that I expect to open and cars and trashcans are small enough that I expect to be able to jump onto them, yet when I try I bounce off. With using real world objects come real world expectations on how those objects should behave.
Games of course try to work around that. SciFi games always have green lights above doors that can open, real world games often remove the doorknobs or make the windows of a house opaque to signal that it's a non-interactive location. Sometimes that works reasonably well, but oftentimes you still end up with a something that looks like you should be able to crawl under or jump over, but simply can't. One get used to it, but it's still annoying.
We could just design and build thorium reactors for a lower cost.
Problem is that building a nuclear power takes a ton of time, 5 years minimum, more likely 10-15 years, especially with the fun of a new design and all the law suits you will run into. Grid parity for wind and solar is projected for 2015-2020, so before your shiny plant is finished, renewables will be the cheaper alternatives.
The problem isn't in the looks, but in the interaction. In highly detailed games you often have tons of stuff that looks like you should be able to interact with it, but you cannot. So while the graphics have gotten more detailed, the interactivity has not. It drives me nuts when I run into doors I can not open, "walls" I can not jump over, holes I can not duck through, items I cannot pickup and all that stuff. With simpler graphics there was a much clearer communication as what is interactive and was is not, as there simply wasn't the computing power available to little the rooms with tons of uninteractive decal.
You can't just mix and match in one part of a major subsystem and you certainly can't mix Debian and Fedora packages beyond the very end user applications that have no connections into the different plumbing that seperates the two trees of development and that basically static link everything.
That's a symptom not a cause. Of course when every packages just barfs all over/usr/ and spreads itself everywhere you can't just install another package of the same name and not expect things to explode. But that's a problem of inflexible namespace management and nothing else.
Every few weeks some kid shows up on a Linux forum demanding that we rebuild everything to support a binary only cross distro 'app' model.
Yeah, and guess what, that kid is right, maybe not in the way it should be implemented, but in the features the system should provide.
The reason we have different distros is because they aren't all alike except for the package manager,
Distros are like 99% alike, because they all run all the same software, just in slightly different incompatible ways. It's an idiotic duplication of work that we should get rid of, not celebrate.
You can't even install a stable package on an unstable distri without running into major version conflicts. Debian right now has no way to handle mixing and matching packages from random sources. PPA don't change that, as they still have to be build against specific versions of specific distributions and can't just be mixed willy nilly without running into tons of conflicts.
The simple truth is: The way software is bundled right now is not hacker friendly. It's ugly, slow and complicated and just flat out doesn't allow a lot of stuff without weeks of extra work. Remember when people complained about Gnome3 in Ubuntu? Why did they do that? Because there was no easy way to get Gnome2 back. How the heck is that acceptable behavior for a package manger in 2012? Handling situations like that should be a non-issue, just keep the old stuff around and call it a day, but you can't do that, as name and version conflicts will give you all kinds of nightmares.
Game: "Now press the green button" Player: "Which one?" Game: "The one that looks like a circle"
Also whats up with the dpad? I thought Nintendo's patent on the cross design expired so that people would now be allowed to build good ones instead of those weird circle things?
I find that the desktop environment or distribution really doesn't matter that much in the end, what matters is that you have the application that you need. So if the users you want to switch over just need a browser, any popular distribution will do (i.e. Ubuntu), but if your users want to play games or use Photoshop or whatever, forget it, it's a waste of time, as Linux doesn't really have anything in those areas that can match the quality of products that Windows offers.
Also know what you are up against. Windows98 did had some issues with stability and such that made Linux an interesting alternative. Newer Windows do not. My Linux boxes crash a heck of a lot more (mostly thanks to buggy GPU drivers) then my rusty copy of Vista, which I don't think I have ever seen crashing. So there is no benefit in switching over for stability. Only areas where I still see Linux having a big advantage is in old hardware support (i.e. drivers rarely become obsolete with a new version) and lower system requirements when you display all the OpenGL desktop effects.
So in the end Linux is good for poor people and Open Source hippies. For everybody else there really is not much reason to switch on the desktop.
That's why Wikipedia is the best online teacher, without all the unnecessary personal involvements.
Kind of, but more by accident then by design. The Wikipedia rules explicitly state that Wikipedia is not supposed to be a text book, thus content in Wikipedia isn't allowed to explicitly teach the reader, but instead just state facts. There is of course still tons and tons of valuable information in Wikipedia, but there is plenty more that is not (i.e. try learning C++ just with Wikipedia, won't work, as the descriptions of it are not detailed enough). It's thus more a medium about supplementing other teaching tools then replacing them.
The hardest problem with learning is that you get lost in all the unresolved references, when the teacher (or book) assumes knowledge from the student that the student doesn't yet have, not the learning new stuff itself. And that's something that could very well be solved by technology and by making use of an interactive medium. Of course it could also be solved by having much smaller classes and more teachers, but I don't quite think that will happen anytime soon.
VGA is on the way out, and in 25 years it's really not unthinkable USB will be too
"on the way out" != unreadable. RS232, parallel ports, PS/2, VGA, zipdrives and 3 1/2 floppies still function perfectly fine on your newest most modern high tech PC thanks to USB, all you have to do is throw $15 or so for an adapter or drive. 5 1/4" is a little harder as your mainboard might not have a floppy connector, but it's not so hard to find one that has one or a drive. 8" is probably the hardest, but they never had near the spread in consumer space like the other tech, they never where ubiquitous and thus not much of a good idea for long term storage to begin with.
Anyway, I'd worry a hell of a lot more about the durability of the medium, then about the hardware getting obsolete. And this ain't limited to digital stuff, a lot of the stuff I printed out 15 years ago on a HP550C deskjet is now completely faded, the printouts that had exposure to light are completely blank. Same is true for a lot of photos, old black&white stuff from my grandparents that is 100 years old is still in prestine conditions, yet color photos from the 70's are completely faded.
USB is pretty much everywhere these days, in your phone, your game console, in your MP3 player, in your harddrive, in your TV, maybe even in your toaster. In Europe there it's even part of law, as they want to get rid of all the custom phone chargers. For mouse and keyboards you don't need anything faster then USB1.0, so there is no need to upgrade, so I would expect that to be around for a pretty damn long time, especially given that right now there is nothing on a the horizon to replace it and even if, whatever will replace it will very likely be either compatible or can be made compatible with a cheap adapter, just like you can still get a serial port and an IDE apdopter for your computer today.
The whole obsolete hardware craze is really a little overrated, as when it happened in the past, it was always with pretty damn obscure hardware. Of course not everybody has a machine around to read some old NASA tapes, so you will have trouble reading those in a few decades, but pretty damn near everybody has something around to read USB.
The real shame is the decline of the adventure genre,
The adventure genre is alive and kicking. Aside from all the stuff TellTale does, the recent Kickstarter revival of almost all classic adventure game developers, the remakes of Monkey Island, you also got plenty of European developers such as Dedalic, House of Tales, Pedulo or Amanita cranking out quality games on a regular basis. You even have a few high profile titles with full AAA budget, such as Heavy Rain or LA Noire. The only thing missing from the adventure genre is the press attention, which is more focused on the third person shooter titles, as there are simply more high budget titles of them.
Modern western games got most of their bad ideas from Half Life, not from JRPGs. The JRPG model of endless cutscene is pretty rarely seen outside of the JRPG genre these days and even there it's not exactly popular given the numerous complaints about the last Final Fantasy. The problem with western games instead is that they are way to focused on setpieces, pre-scripting game events instead of having game events be the result of game mechanics, making everything feel fake and predictable. Trying to cram all your story into one linear unbroken sequence focusing only on a single hero character also leads to rather flat stories (group of military dudes running from point A to B, shooting stuff inbetween, the end).
Not using Facebook does not stop others from uploading photos of you and tagging them with your name. The only secure way to stop that is to stop being around other people.
What's *really* weird is that the iPhone has some of those same limitations and yet it is wildly successful...
The difference is that the iPhone got there first, so whatever problems remain, people learned to live with them. The whole trouble with the Windows Phone is that it's late to the party, so to actually be accepted it would need to be superior to the iPhone, not just on par, as just being on par won't make people switch. Why waste time learning a new phone OS when it has no advantage over the old one and still a lot of the same problems?
Slashdot Mirror
That would be trivial to solve by just caching the file on a HDD. Does LTFS do that?
And then again, they cost $$$$$, because PHB's keep on buying them.
Tapes themselves are incredible cheap, in fact they are the cheapest storage available right now beating both HDDs and DVD-Rs by costing half or a third as much. It's the drive price that is killing it and make them useless for the average consumer. That of course makes HDDs the medium of choice for backup, but it still kind of irks me that we don't have a cheap backup solution right now that keeps drive and media separate (well, DVD-R do that, but are to small to be practical).
I just don't get why they are so keen to get rid off that faeture...
They are not. This is not about replacing X, it's simply about splitting the graphics and input code out of X and moving it over to the kernel and Wayland. You can still run X11 apps under Wayland. So it's really more a pragmatic approach to getting the Xorg code based cleaned up, then an abandonment of network transparencey.
Also in general X11 network transparency is bit overrated in my opinion, as while it is good for things such as basic thin client computing, it is completely useles for basic everyday uses like screen sharing or moving applications from one display to another. It's also rather useless for multimedia, be it video due to the lack of bandwidth or sound due to the X11 simply not handling sound. Wayland won't solve them, but a clean code base means that it will be easier for other people to attack those problems.
I am an advocate of organisations having a degree of responsibility for anything they intend to profit from
Thinking like this leads straight to cooperation enforced censorship, as companies like to avoid risk and thus stuff like the Anonymous Coward function you used to post this, would very likely be the first things to go. Easy account creation as well, so welcome Facebook login everywhere. That's not even theoretically speaking, over here in Germany we have tons of shitty rules like this, one of the results of that is that free anonymous Wifi essentially does not exist over here, because it's simply to risky to offer it.
I think what you are asking for now is simply impossible.
Yes, that's the point. As said, it's not a hard problem, just adding relocation support to binaries would solve 99% of the problems. Current binary packaging infrastructure simply doesn't do that, manually compiling from source does. That's why it's trivial to keep as many different versions of Gimp around when I compile manually from source, while it's impossible to do the same with binary packages. That's why the current way to package binaries in Linux sucks.
And that's been my point, what you want exists in Linux.
You don't get it. What should be there is the ability to "apt-get install debian" on a Ubuntu system. That's not the case. That does not exist. Yes, you could build it, but nobody has done so.
The idea of a distribution after the initial layer has always been to offer a collection of packages designed to work together. In other words to make choice.
Um, no. The idea of distributions was never to make a choice and force it on it's user, it was simply to provide easy installable binaries for Linux software and most distributions try to offer as much software as they can with the man power they have available. Thus you don't just get KDE or Gnome, but XFCE, FVWM, TWM, WindowMaker and dozens of other window manager or desktop environments all at the same time. Distributions might pick what they install by default and what they throw most of their money at to fine tune it, but that's about it, you are always free to install something else if you don't like the default.
In a perfect world a distribution would simply be a meta-package, a collection of software for a given task or taste, that you could freely mix and match with other distributions. But right now, that's not the case, instead each distribution has to repackage all the software you ever might want to use again and there is no cross-distribution compatibility due to the inflexibilities of the package manager (i.e. install Xubuntu on Ubuntu works, installing Ubuntu on Debian does not, as the package manager wouldn't be able to handle the conflicts). You can't even install old Ubuntu packages on a new Ubuntu system without running into all kinds of trouble.
And no, there is no distribution that does what I want, even Gentoo's SLOT system is little more then a glorified way to avoid renaming packages, it still doesn't allow free mixing and matching of different versions.
The closest thing to what I want is actually outside of the Linux world: Fink and MacPorts. They still fail at all the advanced handling of namespaces as they use the same tools distributions on Linux do, but thanks to them being an "add-on" to the system, not the core, they don't try to take control over "/" instead they go to "/fink" or "/macports" and one can simply install and use both at the same time without conflict. That small change of not polluting the global namespace, but keeping things in a separate directory adds a lot of flexibility, a proper system for handling namespaces could of course go much further.
At a certain point its just easier to switch distributions when it comes to a vertical package like this.
Yes, and that's exactly why I say the package managers we have today are crap. There is nothing in principle that would prevent you from stuffing everything that depends on Gnome2 in one namespace and everything that wants Gnome3 in another. But our current package managers don't support any kind of clean namespace separation, everything goes into one tree and when there are two packages with the same name, only one can continue to exist, the other gets overwritten or needs to be manually renamed by the package maintainer. "Switch the distribution" should simply never be the answer to a Linux problem.
It is however annoying, and labor intensive.
Compared to all the time and effort that gets wasted on the users side, it's pretty minimal.
That's not the fault of the package manager. Gnome is designed for integration.
And what the fuck is "designed for integration" supposed to mean? Throwing buzzwords around isn't an argument. Give me some examples of where there would be insurmountable hurdles to allowing Gnome2 and Gnome3 to exist in parallel. I have all the RAM and HDD space I need for that. So why can't I do it? Why do I have to switch the whole distribution to get that, when the distribution doesn't really care about the DE I am using (i.e. I can replace Gnome3 with XFCE just fine)?
At the source level it is completely trivial to install software of different versions in parallel, you just configure it with a different install prefix. Relocatable binaries require a tiny bit of extra work, as most software doesn't support relocation and relies and hard compiled path, so one would need to fix that. At worst you might also need to adjust some Dbus names in case the Gnome people broke the API. But there is really no magic to it, just fix the naming conflicts and call it a day. The only reason we don't have that is because nobody bothered to implement it and instead continues to do things the way they have always been done on Linux, barf all over /usr/, shut your eyes and hope for the best.
I understand but that has always been there because having absolute open levels DOES NOT MAKE SENSE.
It am not arguing about the actual level design, just about it's presentation. When you had a door in Doom, you could walk into it. If you see a door in a modern game, there is like a 80% chance that you can't walk into it. The reason for that is simply that when Doom needed a wall, the designers used a wall. In modern games that try to present realistic locations however the developers don't use walls, as that would look unrealistic, instead they use houses, cars, trashcans or any other object they can think of. Houses however have doors that I expect to open and cars and trashcans are small enough that I expect to be able to jump onto them, yet when I try I bounce off. With using real world objects come real world expectations on how those objects should behave.
Games of course try to work around that. SciFi games always have green lights above doors that can open, real world games often remove the doorknobs or make the windows of a house opaque to signal that it's a non-interactive location. Sometimes that works reasonably well, but oftentimes you still end up with a something that looks like you should be able to crawl under or jump over, but simply can't. One get used to it, but it's still annoying.
We could just design and build thorium reactors for a lower cost.
Problem is that building a nuclear power takes a ton of time, 5 years minimum, more likely 10-15 years, especially with the fun of a new design and all the law suits you will run into. Grid parity for wind and solar is projected for 2015-2020, so before your shiny plant is finished, renewables will be the cheaper alternatives.
The problem isn't in the looks, but in the interaction. In highly detailed games you often have tons of stuff that looks like you should be able to interact with it, but you cannot. So while the graphics have gotten more detailed, the interactivity has not. It drives me nuts when I run into doors I can not open, "walls" I can not jump over, holes I can not duck through, items I cannot pickup and all that stuff. With simpler graphics there was a much clearer communication as what is interactive and was is not, as there simply wasn't the computing power available to little the rooms with tons of uninteractive decal.
You can't just mix and match in one part of a major subsystem and you certainly can't mix Debian and Fedora packages beyond the very end user applications that have no connections into the different plumbing that seperates the two trees of development and that basically static link everything.
That's a symptom not a cause. Of course when every packages just barfs all over /usr/ and spreads itself everywhere you can't just install another package of the same name and not expect things to explode. But that's a problem of inflexible namespace management and nothing else.
Every few weeks some kid shows up on a Linux forum demanding that we rebuild everything to support a binary only cross distro 'app' model.
Yeah, and guess what, that kid is right, maybe not in the way it should be implemented, but in the features the system should provide.
The reason we have different distros is because they aren't all alike except for the package manager,
Distros are like 99% alike, because they all run all the same software, just in slightly different incompatible ways. It's an idiotic duplication of work that we should get rid of, not celebrate.
You can already do all of this
You can't even install a stable package on an unstable distri without running into major version conflicts. Debian right now has no way to handle mixing and matching packages from random sources. PPA don't change that, as they still have to be build against specific versions of specific distributions and can't just be mixed willy nilly without running into tons of conflicts.
The simple truth is: The way software is bundled right now is not hacker friendly. It's ugly, slow and complicated and just flat out doesn't allow a lot of stuff without weeks of extra work. Remember when people complained about Gnome3 in Ubuntu? Why did they do that? Because there was no easy way to get Gnome2 back. How the heck is that acceptable behavior for a package manger in 2012? Handling situations like that should be a non-issue, just keep the old stuff around and call it a day, but you can't do that, as name and version conflicts will give you all kinds of nightmares.
It surely will be fun for the color blind:
Game: "Now press the green button"
Player: "Which one?"
Game: "The one that looks like a circle"
Also whats up with the dpad? I thought Nintendo's patent on the cross design expired so that people would now be allowed to build good ones instead of those weird circle things?
I find that the desktop environment or distribution really doesn't matter that much in the end, what matters is that you have the application that you need. So if the users you want to switch over just need a browser, any popular distribution will do (i.e. Ubuntu), but if your users want to play games or use Photoshop or whatever, forget it, it's a waste of time, as Linux doesn't really have anything in those areas that can match the quality of products that Windows offers.
Also know what you are up against. Windows98 did had some issues with stability and such that made Linux an interesting alternative. Newer Windows do not. My Linux boxes crash a heck of a lot more (mostly thanks to buggy GPU drivers) then my rusty copy of Vista, which I don't think I have ever seen crashing. So there is no benefit in switching over for stability. Only areas where I still see Linux having a big advantage is in old hardware support (i.e. drivers rarely become obsolete with a new version) and lower system requirements when you display all the OpenGL desktop effects.
So in the end Linux is good for poor people and Open Source hippies. For everybody else there really is not much reason to switch on the desktop.
That's why Wikipedia is the best online teacher, without all the unnecessary personal involvements.
Kind of, but more by accident then by design. The Wikipedia rules explicitly state that Wikipedia is not supposed to be a text book, thus content in Wikipedia isn't allowed to explicitly teach the reader, but instead just state facts. There is of course still tons and tons of valuable information in Wikipedia, but there is plenty more that is not (i.e. try learning C++ just with Wikipedia, won't work, as the descriptions of it are not detailed enough). It's thus more a medium about supplementing other teaching tools then replacing them.
Learning is hard.
The hardest problem with learning is that you get lost in all the unresolved references, when the teacher (or book) assumes knowledge from the student that the student doesn't yet have, not the learning new stuff itself. And that's something that could very well be solved by technology and by making use of an interactive medium. Of course it could also be solved by having much smaller classes and more teachers, but I don't quite think that will happen anytime soon.
Don't seem many these days, do you?
USB to R232 adapters cost like $4 on Amazon, not exactly an unobtainable rarity.
VGA is on the way out, and in 25 years it's really not unthinkable USB will be too
"on the way out" != unreadable. RS232, parallel ports, PS/2, VGA, zipdrives and 3 1/2 floppies still function perfectly fine on your newest most modern high tech PC thanks to USB, all you have to do is throw $15 or so for an adapter or drive. 5 1/4" is a little harder as your mainboard might not have a floppy connector, but it's not so hard to find one that has one or a drive. 8" is probably the hardest, but they never had near the spread in consumer space like the other tech, they never where ubiquitous and thus not much of a good idea for long term storage to begin with.
Anyway, I'd worry a hell of a lot more about the durability of the medium, then about the hardware getting obsolete. And this ain't limited to digital stuff, a lot of the stuff I printed out 15 years ago on a HP550C deskjet is now completely faded, the printouts that had exposure to light are completely blank. Same is true for a lot of photos, old black&white stuff from my grandparents that is 100 years old is still in prestine conditions, yet color photos from the 70's are completely faded.
USB is pretty much everywhere these days, in your phone, your game console, in your MP3 player, in your harddrive, in your TV, maybe even in your toaster. In Europe there it's even part of law, as they want to get rid of all the custom phone chargers. For mouse and keyboards you don't need anything faster then USB1.0, so there is no need to upgrade, so I would expect that to be around for a pretty damn long time, especially given that right now there is nothing on a the horizon to replace it and even if, whatever will replace it will very likely be either compatible or can be made compatible with a cheap adapter, just like you can still get a serial port and an IDE apdopter for your computer today.
The whole obsolete hardware craze is really a little overrated, as when it happened in the past, it was always with pretty damn obscure hardware. Of course not everybody has a machine around to read some old NASA tapes, so you will have trouble reading those in a few decades, but pretty damn near everybody has something around to read USB.
The real shame is the decline of the adventure genre,
The adventure genre is alive and kicking. Aside from all the stuff TellTale does, the recent Kickstarter revival of almost all classic adventure game developers, the remakes of Monkey Island, you also got plenty of European developers such as Dedalic, House of Tales, Pedulo or Amanita cranking out quality games on a regular basis. You even have a few high profile titles with full AAA budget, such as Heavy Rain or LA Noire. The only thing missing from the adventure genre is the press attention, which is more focused on the third person shooter titles, as there are simply more high budget titles of them.
You can blame that model on JRPGs
Modern western games got most of their bad ideas from Half Life, not from JRPGs. The JRPG model of endless cutscene is pretty rarely seen outside of the JRPG genre these days and even there it's not exactly popular given the numerous complaints about the last Final Fantasy. The problem with western games instead is that they are way to focused on setpieces, pre-scripting game events instead of having game events be the result of game mechanics, making everything feel fake and predictable. Trying to cram all your story into one linear unbroken sequence focusing only on a single hero character also leads to rather flat stories (group of military dudes running from point A to B, shooting stuff inbetween, the end).
Not using Facebook does not stop others from uploading photos of you and tagging them with your name. The only secure way to stop that is to stop being around other people.
What's *really* weird is that the iPhone has some of those same limitations and yet it is wildly successful ...
The difference is that the iPhone got there first, so whatever problems remain, people learned to live with them. The whole trouble with the Windows Phone is that it's late to the party, so to actually be accepted it would need to be superior to the iPhone, not just on par, as just being on par won't make people switch. Why waste time learning a new phone OS when it has no advantage over the old one and still a lot of the same problems?