I doubt if you have ever seen or used a XO Laptop.
For reference this is how the thing looks with backlight off in normal indoor lighting. Not exactly what I would call usable. Now given, I haven't ever touched a Kindle myself, so I have no idea how good they really are, but all the reports indicate that they are very close to real paper, something the OLPC doesn't even get close to, not even in actual sunlight.
The problem with the OLPC screens is that their "sunlight readable" really means "only readable in sunlight". Black&white mode is mostly useless indoors, as you get way to little light onto the screen to be able to read anything (if anybody remembers the original GBA, kind of like that). Now of course in actual sunlights the screen are great and their 200dpi resolution is still pretty impressive to this day, but even there its more like reading something printed on really dark recycling paper then on actual white paper. With $100 Kindle's being easily available (much lighter, much longer battery life, etc.) it's also not quite as impressive as it used to be. In color mode the screens are pretty bad and extremely depended on viewing angle, so much that it's hard to even find a position where the screen looks proper, in B&W mode that problem however completely disappears. I find the screens most usable when having a lot of environmental light and backlight set to the minimum, this reduces the view angle dependency and leads to pretty readable results.
All that said, the new PixelQi screens are not OLPC screens, so they should be better, but judging from the reviews, the core problems seem to be still the same.
Because there's nothing more that kids love more than history lessons.
The problem is that when you skip history, every invention and science will look like magic and all you end up teaching is bare facts that will leave the chidrens brain faster then you can put them back in.
That of course also means that you shouldn't teach history in the history-lesson sense of having them remember dates and such, those are meaningless. The important part is teaching the process that those people who made those discoveries went through. How did people come up the the original idea, what steps did they take to implement it, what are the fundamentals they are building on, etc.
The problem with examples like this is that the game already had a regular successful commercial run four years ago. It already made it's money back long before it went free to play. Same is true for many MMORPGs, they only went free-to-play after they stopped making money via regular subscriptions and as a MMORPG needs players to function, free-to-play is simply an alternative to revitalize it for a little more before it's completely dead.
I believe Zynga took in more money than EA this past year. And a couple years ago Zynga didn't even exist. EA has been scrambling to copy Zynga's freemium model.
The problem with that business model is that it is not build around giving away a good product for free, but around a product that exploits human psychology and lack of knowledge to essentially turn every player into a human spambot for the company.
And companies like IBM, Google, etc. make billions on the back of open source software.
Google isn't making their money with Open Source software, by far most of their stuff is proprietary. And their free services are all full of advertisement. IBM isn't making money by giving away all their stuff for free either. So those aren't exactly freemium business models, but simply regular businesses which are rich enough to give a bit of their stuff away for free when it makes sense, but can keep everything proprietary when needed.
Getting rid of copyright would mean that the GPL wouldn't need to exist anymore.
The GPL is about giving you the rights, the tools and the code needed to modify the programs you have. Without copyright you don't get any of that, you only get the right to copy. Modifications, while also allowed, would often be to impractical to actual do.
A further problem with absolutely no copyright could be that things simply shift to contracts and water-markings. Copyright might allow you to copy every movie you watched, but if you signed into a contract that revokes that right, you are back to square one. DRM might also make it impossible to run modified code on the devices you own. Of course, many of those measures are already in place anyway, but lack of copyright could encourage distributors to focus even more on them.
That's a bit like suggesting that because I've replaced the individual parts in my car that I no longer have the same car and that it's exactly the same as if I were to just buy a new car.
It is, at least for the car. It only makes a difference to you because you don't lose continuity when you replace your car piece by piece instead of all at once, as you are always left with something that reminds you of your old car. By the time the last part of your old car gets replaced, you have gotten used enough to the new parts to consider them "your car". However for the car it makes no difference, your old car is on the scrapyard and the thing you are driving is all new.
Exoskeletons and robotic limbs are kind of like self-driving cars. Every few years, you see a news report on supposed progress made. Some prototype is demonstrated.
20 something years VW was demoning a self parking car prototype, then nothing every came of that... till a little while ago when automatic parking starting becoming a feature seen in regular everyday cars. Also ePaper, years and years of prototype and tech demo and nothing usable of even buyable, then Kindle happened. New technology simply takes a while to get from first prototype to mass market and when you sit at the sidelines reading news reports about it all the time, it might seem like there isn't any progress, yet it still happens.
Now exoskeletons might not have quite the mass market appeal as a Kindle, but there is still plenty of demand (i.e. elderly people) that I have little double that 10 or 20 years down the road those things won't be something you only see on some tech demo video, but something that walks by you on the street.
The overarching issue isn't really CLI vs GUI, but that the OS provides the user with very little semantic information, instead you simply get pretty pixel graphics. Case in point: Look at your screen right now, how much of the text you see can you select and copy as text? The answer will of course vary depending on what you do, but you can be pretty sure that it will be a good bit lower then 100% (i.e. window titles, menus, etc. can't be selected). There is really no good reason for that being that way, other then that being the way it has always been. The text is available to the OS and the applications, but there are no tools to get it out or at least not easily. Now that's of course just a very basic case, the issues goes of course much deeper when it comes to active parts of the GUI. When your filemanager is displaying a list, can you copy it into a spreadsheet? Can you move the play button of your MP3 player over to your iPhone? etc. Some of those use cases are of course a little far fetched, but essentially what you want is a rich and flexible way to interact with your computer. Neither CLIs nor GUIs really provide that and both of them don't really mix well (i.e. double clicking on the output of 'ls' should allow you to open a file).
I'd like to scrape that list and do some kind of throwaway mashup for myself. It's painful.
That's not so much a problem with the semantic web, but simply a lack of a more powerful copy&paste in your webbrowser/OS. The information is already there, nicely structured in a list and all, but your browser provides no way to get it out of the dropdown menu easily. If you go low-tech and use Lynx, you can just copy&paste the thing right out of your terminal with little problem. Nice benefit of everything being text in a terminal, even GUI elements.
Now of course when it comes to building more permanent things, not just throw away copy&paste, RESTful APis are the way to go, as they give you direct access to the raw undelying data, not the pretty-printed HTML gibberish that is split across dozens of pages. What is missing there is again a bit of browser support, as while browsers have no problem retrieving the info, they have no userfriendly way to dealing with a JSON or XML structure in any meaningful way. There is also no standard way of linking that data up into the webpage, i.e. saying "this table/article/whatever can be accessed as JSON (here)". Lack of standard formats for common data, aside from the structure that XML/JSON provide, is of course also an issue.
On top of that however the biggest issue is simply that most webpages don't want to provide semantic information, as that would mean the user would have raw access and couldn't be bothered so easy with adverts and other junk.
Then in another 5 years, the fee doubles. 5 more years it doubles again...
Yes, and in practice that means individual authors get 10 years of copyright while cooperations get 70 years (or whatever). I simply don't see how such a scheme could end up not punishing the individual author and have little to no effect on big cooperations.
At the 15 or 20 year mark, if your work is a smash hit and still selling well, then you can afford to pay the increasing fee...
And again, I just don't see the benefit. As that's exactly the opposite of how things should be. Why should the already successful work be the only one that gets even more protection? It already was successful and made it's money back. Copyright should encourage new works, not companies milking past successes.
Others have suggested requiring appraisals for any copyrights a company is holding, and taxing them on the estimated value.
In theory that sounds like a good idea, but in practice that could turn into a heap load of paper work, which again, won't bother the big cooperation with it's staff of lawyers much.
Big companies won't have to be subject to public domain right away, but they'll have to pay a hell of a lot of money for the privlidge.
They won't. If you don't want to force every independent author into the public domain, copyright extension has to be extremely cheap and easy, which means all cooperations can extend the copyright on all their work for essentially free, as whatever the fee will be, it will be to tiny for them to care about.
The independent author on the other side will run into issues all the time, as even with an easy and cheap system, the overhead to register every blog post, twitter post, images on flickr and what ever will be to much to bother with.
That could lead to situations where a publisher simply refuses to publish a work to avoid paying the author any royalties, i.e. way pay now when you can get the thing for free a few years down the road? I'd like that solution better if it would be limited to non-commercial use and only allowed commercial use until regular copyright expired (which a shorter overall copyright term of course).
Say if you want to keep your copyright after X years pay X fee.
While that would be a nice trick to get a lot of today abandoned stuff into the public domain, I really don't like the idea in the long run, as it would mean that all the big cooperations simply let their lawyers handle things and get copyright protection for as long as the law allows, while the stuff of the little guy will slip into public domain against their will.
I think a much better solution to copyright would be staged copyright, i.e. 15 years of copyright as is, after that another 15 years where the work is free-for-non-commecial-use, then full public domain.
Apparently you live on a different planet than I do. On this planet, I have used free software to get all my work done for the last six years or so.
I have done so for the last 12 or so years, I am however not under any disillusion that Free Software had much to do with that. Proprietary software would have essentially almost always done a much better job at getting the work done. Have there been cases where there I used the source to fix problems? Sure, but those where problems I would never have had with good proprietary software. And speaking about spreadsheets, OpenOffice can't even import Gnumeric files.
There are two ways to deal with the problem of too much information to process: less information or better processing.
The problem with choices isn't just about "to much", but that the choices often actually conflict with each other and that there isn't a clear best choice. If you compare two MP3 players and each of them has half the features you want, but each a different half, you are kind of stuck between a rock and a hard place. The solution here wouldn't be more information, but building a better product.
That's why the existance of 10,000,000 sites on a particular subject doesn't cause us anxiety.
The reason those million sites don't cause much of a problem isn't because of Google, but because you don't have to commit to any of those sites. You go there, pick the information you want and move on. Don't like the page? Go to another one. Shifting through to much redundant information might get tiresome, but it's additive, you always learn a little more of what you want to know. When you actually have to commit to a choice and it locks you out of the other choices, then it starts to become a real problem.
Users do want freedom, they just don't realize it until they've completely lost it and then have a use for it.
The thing users want is practical freedom, the freedom that lets them get their work done, let them get access to their data, let them have fun and consume what they want.
The problem with Free Software is that it generally only gives the user theoretical freedom. Having access to the source and a general computing device sounds all nice in theory, but when you want to get work done, it's hardly ever of any use. Fixing the software is too complicated and hiring somebody would be to expensive. Thus the user not only doesn't get his work done, he also has freedom that has no benefit for him.
If Free Software wants to matter for end users it needs to start focusing on actually producing working solutions for user problems, not just random bits and pieces that the user has to plug together himself.
If you want to get precise: There is no non-deadly, you can kill people with a tooth pick or with a spoon, heck, you can kill them with your bare hands, deadliness is a matter of degree. There however is a big difference between a tool that is optimized for killing and one where the killing part is an undesirable side-effect of it's mass and shape.
There is a substantial difference between 'there are no guns' and 'guns are banned.'
While that is a perfectly good argument against a radical ban of guns right for the USA, it's not much of a good argument against a long term goal of a "no guns". Which is perfectly valid and achievable, not completely 100%, but close enough as demonstrated by plenty of other countries. One could for example start by limit gun production and disallowing them for certain uses where safer alternatives exist.
Again, explain to me how you can make a javelin non-lethal.
You can't, but you can make them less deadly, which you might have easily deduced by using a bit common sense, but hey, attacking strawman is so much more fun.
Remove P2P from the equation, and people will share copyright videos using other means (just like they did before P2P).
While the removal of P2P wouldn't stop piracy, it could certainly change it scale and scope, as peoples behavior is influences a lot by the convenience and abilities of the tools they have at hand.
You are however brushing p2p with rather broad strokes you have to admit...
Might be, but I am just going by personal experience. I consume a ton of legal open content, be podcast, videos or free software. Yet the amount of that that comes over P2P is pretty damn close to zero. If P2P is so awesome for legal uses, I really don't see them much in the wild, I have to go search for them with a magnifying glass.
It has not. Go read the intro books from the early 90s. They all talk about how everyone can talk to each other equally.
Modern P2P isn't just about talking equally, but about sharing resources and becoming an provider for the stuff you consume. That has never been how the Internet operated. Internet was always about having servers that provide data and clients that consume it, it always was a one way street aside from manual efforts such as setting up a classic mirror server.
The reason that your code, podcasts, and online videos are hosted in special purpose web applications is that they understand the format and purpose of your data.
Yes, but those have little to do with the actual transfer protocol. There is in principle nothing that would stop you from using P2P for these services instead of HTTP, as the core protocol doesn't care much about what data you ship with it. Yet my Podcast tool doesn't even support torrent or any other kind of P2P and my browser doesn't either.
Rsync is optimised for replicating frequently changing collections of small files
Rsync is not P2P, it's classic client/server like HTTP and FTP, just with a bit of delta/incremental updates thrown in.
You must not edit video collaboratively, or develop and distribute indie games, or train neural nets, or share language corpora, or any number of other possible legal uses.
I do and bittorrent or really any kind of P2P, is a shitty tool for handling regularly updating data.
World of Warcraft uses the bittorrent protocol to update their clients. That's far more than a "proof of concept".
Not really, as that's not the use case I was talking about. WoW patches are still a large blobs of mostly static data, not a large collection of frequently changing small files. Furthermore Blizzard has bittorrent wrapped up in their own client, which also allows them to bypass limitations of the protocol (i.e. chumbersome management of managing.torrent files).
Slashdot Mirror
I doubt if you have ever seen or used a XO Laptop.
For reference this is how the thing looks with backlight off in normal indoor lighting. Not exactly what I would call usable. Now given, I haven't ever touched a Kindle myself, so I have no idea how good they really are, but all the reports indicate that they are very close to real paper, something the OLPC doesn't even get close to, not even in actual sunlight.
The problem with the OLPC screens is that their "sunlight readable" really means "only readable in sunlight". Black&white mode is mostly useless indoors, as you get way to little light onto the screen to be able to read anything (if anybody remembers the original GBA, kind of like that). Now of course in actual sunlights the screen are great and their 200dpi resolution is still pretty impressive to this day, but even there its more like reading something printed on really dark recycling paper then on actual white paper. With $100 Kindle's being easily available (much lighter, much longer battery life, etc.) it's also not quite as impressive as it used to be. In color mode the screens are pretty bad and extremely depended on viewing angle, so much that it's hard to even find a position where the screen looks proper, in B&W mode that problem however completely disappears. I find the screens most usable when having a lot of environmental light and backlight set to the minimum, this reduces the view angle dependency and leads to pretty readable results.
All that said, the new PixelQi screens are not OLPC screens, so they should be better, but judging from the reviews, the core problems seem to be still the same.
Because there's nothing more that kids love more than history lessons.
The problem is that when you skip history, every invention and science will look like magic and all you end up teaching is bare facts that will leave the chidrens brain faster then you can put them back in.
That of course also means that you shouldn't teach history in the history-lesson sense of having them remember dates and such, those are meaningless. The important part is teaching the process that those people who made those discoveries went through. How did people come up the the original idea, what steps did they take to implement it, what are the fundamentals they are building on, etc.
And then you have something like TF2...
The problem with examples like this is that the game already had a regular successful commercial run four years ago. It already made it's money back long before it went free to play. Same is true for many MMORPGs, they only went free-to-play after they stopped making money via regular subscriptions and as a MMORPG needs players to function, free-to-play is simply an alternative to revitalize it for a little more before it's completely dead.
I believe Zynga took in more money than EA this past year. And a couple years ago Zynga didn't even exist. EA has been scrambling to copy Zynga's freemium model.
The problem with that business model is that it is not build around giving away a good product for free, but around a product that exploits human psychology and lack of knowledge to essentially turn every player into a human spambot for the company.
And companies like IBM, Google, etc. make billions on the back of open source software.
Google isn't making their money with Open Source software, by far most of their stuff is proprietary. And their free services are all full of advertisement. IBM isn't making money by giving away all their stuff for free either. So those aren't exactly freemium business models, but simply regular businesses which are rich enough to give a bit of their stuff away for free when it makes sense, but can keep everything proprietary when needed.
Getting rid of copyright would mean that the GPL wouldn't need to exist anymore.
The GPL is about giving you the rights, the tools and the code needed to modify the programs you have. Without copyright you don't get any of that, you only get the right to copy. Modifications, while also allowed, would often be to impractical to actual do.
A further problem with absolutely no copyright could be that things simply shift to contracts and water-markings. Copyright might allow you to copy every movie you watched, but if you signed into a contract that revokes that right, you are back to square one. DRM might also make it impossible to run modified code on the devices you own. Of course, many of those measures are already in place anyway, but lack of copyright could encourage distributors to focus even more on them.
That's a bit like suggesting that because I've replaced the individual parts in my car that I no longer have the same car and that it's exactly the same as if I were to just buy a new car.
It is, at least for the car. It only makes a difference to you because you don't lose continuity when you replace your car piece by piece instead of all at once, as you are always left with something that reminds you of your old car. By the time the last part of your old car gets replaced, you have gotten used enough to the new parts to consider them "your car". However for the car it makes no difference, your old car is on the scrapyard and the thing you are driving is all new.
Exoskeletons and robotic limbs are kind of like self-driving cars. Every few years, you see a news report on supposed progress made. Some prototype is demonstrated.
20 something years VW was demoning a self parking car prototype, then nothing every came of that... till a little while ago when automatic parking starting becoming a feature seen in regular everyday cars. Also ePaper, years and years of prototype and tech demo and nothing usable of even buyable, then Kindle happened. New technology simply takes a while to get from first prototype to mass market and when you sit at the sidelines reading news reports about it all the time, it might seem like there isn't any progress, yet it still happens.
Now exoskeletons might not have quite the mass market appeal as a Kindle, but there is still plenty of demand (i.e. elderly people) that I have little double that 10 or 20 years down the road those things won't be something you only see on some tech demo video, but something that walks by you on the street.
The overarching issue isn't really CLI vs GUI, but that the OS provides the user with very little semantic information, instead you simply get pretty pixel graphics. Case in point: Look at your screen right now, how much of the text you see can you select and copy as text? The answer will of course vary depending on what you do, but you can be pretty sure that it will be a good bit lower then 100% (i.e. window titles, menus, etc. can't be selected). There is really no good reason for that being that way, other then that being the way it has always been. The text is available to the OS and the applications, but there are no tools to get it out or at least not easily. Now that's of course just a very basic case, the issues goes of course much deeper when it comes to active parts of the GUI. When your filemanager is displaying a list, can you copy it into a spreadsheet? Can you move the play button of your MP3 player over to your iPhone? etc. Some of those use cases are of course a little far fetched, but essentially what you want is a rich and flexible way to interact with your computer. Neither CLIs nor GUIs really provide that and both of them don't really mix well (i.e. double clicking on the output of 'ls' should allow you to open a file).
I'd like to scrape that list and do some kind of throwaway mashup for myself. It's painful.
That's not so much a problem with the semantic web, but simply a lack of a more powerful copy&paste in your webbrowser/OS. The information is already there, nicely structured in a list and all, but your browser provides no way to get it out of the dropdown menu easily. If you go low-tech and use Lynx, you can just copy&paste the thing right out of your terminal with little problem. Nice benefit of everything being text in a terminal, even GUI elements.
Now of course when it comes to building more permanent things, not just throw away copy&paste, RESTful APis are the way to go, as they give you direct access to the raw undelying data, not the pretty-printed HTML gibberish that is split across dozens of pages. What is missing there is again a bit of browser support, as while browsers have no problem retrieving the info, they have no userfriendly way to dealing with a JSON or XML structure in any meaningful way. There is also no standard way of linking that data up into the webpage, i.e. saying "this table/article/whatever can be accessed as JSON (here)". Lack of standard formats for common data, aside from the structure that XML/JSON provide, is of course also an issue.
On top of that however the biggest issue is simply that most webpages don't want to provide semantic information, as that would mean the user would have raw access and couldn't be bothered so easy with adverts and other junk.
Then in another 5 years, the fee doubles. 5 more years it doubles again...
Yes, and in practice that means individual authors get 10 years of copyright while cooperations get 70 years (or whatever). I simply don't see how such a scheme could end up not punishing the individual author and have little to no effect on big cooperations.
At the 15 or 20 year mark, if your work is a smash hit and still selling well, then you can afford to pay the increasing fee...
And again, I just don't see the benefit. As that's exactly the opposite of how things should be. Why should the already successful work be the only one that gets even more protection? It already was successful and made it's money back. Copyright should encourage new works, not companies milking past successes.
Others have suggested requiring appraisals for any copyrights a company is holding, and taxing them on the estimated value.
In theory that sounds like a good idea, but in practice that could turn into a heap load of paper work, which again, won't bother the big cooperation with it's staff of lawyers much.
Big companies won't have to be subject to public domain right away, but they'll have to pay a hell of a lot of money for the privlidge.
They won't. If you don't want to force every independent author into the public domain, copyright extension has to be extremely cheap and easy, which means all cooperations can extend the copyright on all their work for essentially free, as whatever the fee will be, it will be to tiny for them to care about.
The independent author on the other side will run into issues all the time, as even with an easy and cheap system, the overhead to register every blog post, twitter post, images on flickr and what ever will be to much to bother with.
That could lead to situations where a publisher simply refuses to publish a work to avoid paying the author any royalties, i.e. way pay now when you can get the thing for free a few years down the road? I'd like that solution better if it would be limited to non-commercial use and only allowed commercial use until regular copyright expired (which a shorter overall copyright term of course).
Say if you want to keep your copyright after X years pay X fee.
While that would be a nice trick to get a lot of today abandoned stuff into the public domain, I really don't like the idea in the long run, as it would mean that all the big cooperations simply let their lawyers handle things and get copyright protection for as long as the law allows, while the stuff of the little guy will slip into public domain against their will.
I think a much better solution to copyright would be staged copyright, i.e. 15 years of copyright as is, after that another 15 years where the work is free-for-non-commecial-use, then full public domain.
Apparently you live on a different planet than I do. On this planet, I have used free software to get all my work done for the last six years or so.
I have done so for the last 12 or so years, I am however not under any disillusion that Free Software had much to do with that. Proprietary software would have essentially almost always done a much better job at getting the work done. Have there been cases where there I used the source to fix problems? Sure, but those where problems I would never have had with good proprietary software. And speaking about spreadsheets, OpenOffice can't even import Gnumeric files.
That sadly fails on like 40% of the services out there, as they don't allow passwords longer then 20 or so characters.
There are two ways to deal with the problem of too much information to process: less information or better processing.
The problem with choices isn't just about "to much", but that the choices often actually conflict with each other and that there isn't a clear best choice. If you compare two MP3 players and each of them has half the features you want, but each a different half, you are kind of stuck between a rock and a hard place. The solution here wouldn't be more information, but building a better product.
That's why the existance of 10,000,000 sites on a particular subject doesn't cause us anxiety.
The reason those million sites don't cause much of a problem isn't because of Google, but because you don't have to commit to any of those sites. You go there, pick the information you want and move on. Don't like the page? Go to another one. Shifting through to much redundant information might get tiresome, but it's additive, you always learn a little more of what you want to know. When you actually have to commit to a choice and it locks you out of the other choices, then it starts to become a real problem.
Users do want freedom, they just don't realize it until they've completely lost it and then have a use for it.
The thing users want is practical freedom, the freedom that lets them get their work done, let them get access to their data, let them have fun and consume what they want.
The problem with Free Software is that it generally only gives the user theoretical freedom. Having access to the source and a general computing device sounds all nice in theory, but when you want to get work done, it's hardly ever of any use. Fixing the software is too complicated and hiring somebody would be to expensive. Thus the user not only doesn't get his work done, he also has freedom that has no benefit for him.
If Free Software wants to matter for end users it needs to start focusing on actually producing working solutions for user problems, not just random bits and pieces that the user has to plug together himself.
If you want to get precise: There is no non-deadly, you can kill people with a tooth pick or with a spoon, heck, you can kill them with your bare hands, deadliness is a matter of degree. There however is a big difference between a tool that is optimized for killing and one where the killing part is an undesirable side-effect of it's mass and shape.
There is a substantial difference between 'there are no guns' and 'guns are banned.'
While that is a perfectly good argument against a radical ban of guns right for the USA, it's not much of a good argument against a long term goal of a "no guns". Which is perfectly valid and achievable, not completely 100%, but close enough as demonstrated by plenty of other countries. One could for example start by limit gun production and disallowing them for certain uses where safer alternatives exist.
Again, explain to me how you can make a javelin non-lethal.
You can't, but you can make them less deadly, which you might have easily deduced by using a bit common sense, but hey, attacking strawman is so much more fun.
Remove P2P from the equation, and people will share copyright videos using other means (just like they did before P2P).
While the removal of P2P wouldn't stop piracy, it could certainly change it scale and scope, as peoples behavior is influences a lot by the convenience and abilities of the tools they have at hand.
You are however brushing p2p with rather broad strokes you have to admit...
Might be, but I am just going by personal experience. I consume a ton of legal open content, be podcast, videos or free software. Yet the amount of that that comes over P2P is pretty damn close to zero. If P2P is so awesome for legal uses, I really don't see them much in the wild, I have to go search for them with a magnifying glass.
It has not. Go read the intro books from the early 90s. They all talk about how everyone can talk to each other equally.
Modern P2P isn't just about talking equally, but about sharing resources and becoming an provider for the stuff you consume. That has never been how the Internet operated. Internet was always about having servers that provide data and clients that consume it, it always was a one way street aside from manual efforts such as setting up a classic mirror server.
The reason that your code, podcasts, and online videos are hosted in special purpose web applications is that they understand the format and purpose of your data.
Yes, but those have little to do with the actual transfer protocol. There is in principle nothing that would stop you from using P2P for these services instead of HTTP, as the core protocol doesn't care much about what data you ship with it. Yet my Podcast tool doesn't even support torrent or any other kind of P2P and my browser doesn't either.
Rsync is optimised for replicating frequently changing collections of small files
Rsync is not P2P, it's classic client/server like HTTP and FTP, just with a bit of delta/incremental updates thrown in.
You must not edit video collaboratively, or develop and distribute indie games, or train neural nets, or share language corpora, or any number of other possible legal uses.
I do and bittorrent or really any kind of P2P, is a shitty tool for handling regularly updating data.
World of Warcraft uses the bittorrent protocol to update their clients. That's far more than a "proof of concept".
Not really, as that's not the use case I was talking about. WoW patches are still a large blobs of mostly static data, not a large collection of frequently changing small files. Furthermore Blizzard has bittorrent wrapped up in their own client, which also allows them to bypass limitations of the protocol (i.e. chumbersome management of managing .torrent files).