I see the context-switch as one between text and graphics.
If I'm working on the command line, then most times I have no need to have X11 running. I'm working exclusively with text.
I can boot to a command line and start working. No X11 is needed.
But when a need arises for graphics, e.g., to read a PDF composed of scanned images (not pure Postscript), then I have to "context-switch" to the X11 context.
I find that switching back and forth between these two contexts is not smooth and can easily lead to instability.
There is often a presumption, as in Plan9, that we will just switch once: to the graphical environment. And not return to the original console.
To me, neither an X11 terminal emulator nor the Plan9 environment is "the console". It's another layer of abstraction on top of the console.
That is a lot of overhead I do not need if I'm just working with text.
Because it is tied to the checking system, where checks were routinely flown across the country from bank to bank before they could be settled? (Even after we had electronic systems that should make this unnecessary?)
And all the mainstream messaging services to date have been centralised, at least in the sense that they involve interacting with a third party server.
When each friend can be both a client and/or a server, no third party servers are necessary. In theory (and practice), this is something you can achieve on a small network consisting only of your friends.
What if all your friends want to be online at the same time?
What if they want to share photos and video while online at the same time?
What if they want to play games with each other while online at the same time?
You can currently do these things with the mainstream web-based services like Facebook. But they are recording everything you say and do _and_ selling that information for profit. You don't receive any portion of that profit.
I imagine there might be other software/consulting companies in the business of stealthily collecting GB and PB's of personal data about consumers using wired and mobile networks who are thinking "Am I next?"
And I imagine some of their employees' minds might now be filling with thoughts about how to justify what they do, or to discount the need for anyone to make a big deal about what they do.
Will consumers care about what's booting when they turn on their phone, or what connections their phones are making? This will be very interesting.
With respect to "telling something to some of your friends", and attempting to do so "privately", there are certainly ways to do this without using Facebook.
However, that was only a specific example I chose, in line with veb's example of telling people you rub lettuce on your face, to use to illustrate to scott a point about whether only large networks could be useful in order to stay in touch with a small number of people, i.e., your friends. In theory, I could use any online activity or any service/protocol as an example to illustrate what the "solution" (a small private network) aims to achieve.
Talking (VOIP e.g. SIP), smtp (email), IRC and http (web forums), to use your examples, are examples of services/protocols that can be run over a network. Of course it is not an exhaustive list.
You could run them over the open internet, i.e. a very large, public network (of networks).
You could also run them over a small private network to which only a selection of people belong, e.g., your friends.
In theory, anything you could do with your friends on Facebook you could also do with your friends on your own small private network.
Multiplayer games is something for which this idea of "being on the same network", all at the same time, is well-suited. This is not a new concept. It is a very old one. Consequently, it's time-tested.
But playing games is only one example of what you can do.
The internet supports many services.
Theoretically, so too can your smaller network.
An obvious difference between doing things on the open internet (Facebook) and doing them on your own network is: _privacy_.
You do not have to invite advertisers and countless others to your private network if you do not want to. Might this be important to some people? That is an open question.
_Privacy_, of the kind discussed in the Facebook context, is the goal which the "solution" we are discussing aims to address.
Not simply "private mesaging" but privacy in everything you do with your friends online.
Rest assured, even if such a solution did exist and could be shown to work (NAT and whatever other issues you might predict have been solved), all the Facebook-type user interface doo-dahs are noticeably absent.
As such, it is a non-starter for any friend who cannot use a command line, unless some very good user interface developers got behind it.
You said: "I think it has to be a part of a large _service_."
My question was about the size of the _network_.
In any event, following your line of thought, do you think it's possible to have a many _small_, separate networks that were somehow part of a large service?
Regardless of your answer, does our solution have to be a "service"?
What if it is a "product" that creates small networks as overlays on a larger, existing network such as the one all your friends are connected to: the internet?
You said: "I can see that it's possible to create a service around, say, PGP..."
What if you could see that it's possible to create a service (or product, or both) around, say, a scheme that involved only a single shared password and a single shared encryption key? That is, each friend has to remember only two strings for each network to which she belongs, sort of like, say, a username and password.
What if you could see that such a scheme might not require logging on and logging out as frequently as a web-based service such as Facebook?
Would that change your thoughts at all?
You said, when referring to a PKI scheme like PGP: "But I think that [the] level of conceptual overhead is more than lay-people are willing to deal with."
I once thought the same thing about Amazon's S3 service. When I saw the Dropbox product, my thoughts changed.
What if someone wants to tell a few friends about something but doesn't want to tell all the people working at Facebook, all its clients and potentially any other Facebook user?
As another commenter pointed out, Zuckerberg once called users "dumb fucks" for giving him their personal information. If he does not believe in privacy, as he now claims, why would he say that?
So where does the investor obtain the information to perform due diligence? Are companies required to file reports? If so, with whom do they file them?
So you think that with no SEC the public will be safer from people like Madoff?
Do you think that investors would be able to make sound investment decisions without companies having any reporting requirements? Not every country in the world has such requirements and I think the SEC in that regard alone (call it a "success" if you will) is something to be commended.
It's one thing to say people at the SEC are not doing their jobs. But it's another thing to question the rationale and purpose behind the regulations. Are you doing both? Or just one?
If I'm not mistaken the senator is only doing the later.
Is it unreasonable for a consumer to want complete control over a device from the moment she powers on?
Is there a certain level beneath which it is not reasonable to give consumers (optional) access? (Should consumers be prevented from "rooting" devices? Should we allow companies to maintain control over devices, e.g. having them "phone home", after they sell them?)
If yes, why?
Maybe a rootkit should just be viewed just like the crapware that comes pre-installed on a PC. Sure it will help some company and perhaps the consumer herself, if she decides to use it. But it's _optional_.
Maybe they could give consumers an easy way to opt-out.
I see the context-switch as one between text and graphics.
If I'm working on the command line, then most times I have no need to have X11 running. I'm working exclusively with text. I can boot to a command line and start working. No X11 is needed.
But when a need arises for graphics, e.g., to read a PDF composed of scanned images (not pure Postscript), then I have to "context-switch" to the X11 context.
I find that switching back and forth between these two contexts is not smooth and can easily lead to instability.
There is often a presumption, as in Plan9, that we will just switch once: to the graphical environment. And not return to the original console.
To me, neither an X11 terminal emulator nor the Plan9 environment is "the console". It's another layer of abstraction on top of the console.
That is a lot of overhead I do not need if I'm just working with text.