...interesting, yes, but irrelevant, just to be clear, since a touchpad is not even the same class of device as a touch screen. In other words, it is not as if existing touchscreen games developed for a phone will be playable on the Ouya because they have this touchpad thing.

Yes, but Sherlock Holmes is also a master of disguise.

Sherlock Holmes, is that you?

It's X that has this limitation. With X, when I attach an application to a display (X server) it's forever tied to that display and can't be moved.

There is no such limitation in Windows. The only limitation in Windows is that I have to move all my applications in my session to that display, but this is hardly a noteworthy limitation since it's MY session and I'm not setting at the old display anymore.

With X, I have to stop the application, then start it on the new display. Hooray, I don't have to bring my whole session over if I don't want to, but I would gladly do so if it meant resuming the running instance of my already-running application.

The point of this whole (off-topic) discussion is that these are two completely different use cases. One is remote execution of an application. The other is remote control of an application. X only does the former. Windows is actually capable of either (with the caveat that it requires you to start a remote logon session to remote the execution of a single app).

I'm not "cheering" about Windows, you troll. My point is and has been that I hate Windows and love Linux and yet trying to do this simple task in Linux is still not possible a decade after Windows managed it which is extremely frustrating. The most frustrating part is that every time someone complains about the lack of this feature, someone comes along with a shithead response like yours. "What? You want a feature Windows has? Fuck you, you must be a Microsoft lackey."

No.

This uses controllers and not a touchscreen.

This is not a "dev" box in the sense that if you are a developer, you need one of these boxes. These are boxes that were specifically awarded to backers that wanted to do development. The only difference between these and the retail boxes is that these are early versions and therefore available earlier than the retail boxes. Also these are in "special edition" cases as a thank you to the devs for their support.

You don't. Sorry, didn't mean to suggest this was possible.

To be fair, I'd be perfectly happy to transfer the entire X session from one machine to another (provided X was also made as fast as NX), but that doesn't seem to be possible except through VNC which is unacceptable in its slowness and its inability to change screen dimensions, or xpra which is unacceptably slow also, and often cripples performance and features when running locally.

So the point is that the use case in Windows where I can start an application on my home desktop, then access that application later from my work machine with reasonable performance and at the client-side resolution, then do the same from my neighbor's house, etc, all without exiting the application.... this is all I am referring to.

Actually that's a blockquote. A citation is when you say where you got your info from.

Here's a citation:

See what I mean? That's not really a citation... I just put quote tags around my own words.

I suspect you actually copied and pasted that from somewhere instead of making it up yourself, but if you don't cite that source, then it is impossible to tell the difference between sock puppet and actual citation. I see you've been modded up which means people on /. don't really care whether there's a citation or not as long as it sounds official.

Replying to myself. XRPA does in fact work with the following caveats:

1. It does appear to use bitmap tossing and probably for this reason is painfully slow over low bandwidth connections - far more sluggish than Windows' RDP.
2. Like "screen" for console apps, you'd have to have started the application in the xrpa session in order to be able to attach it to a remote screen.

No, I mean can start an application running on machine A and displaying on A's display, then access that same instance of the application from machine B's display, then display C if I want etc., all the while never exiting the application running on machine A.

xrpa seems interesting but appears to be VNC for X, not actual X. This is what the GGGGP was referring to as "bitmap tossing" which is pretty useless over slow connections while windows RDP is not.

Which is exactly what I was talking about. Windows has provided this capability for over a decade and linux has not except with VNC which is unacceptable.

The X server that the application will connect to is based on the value of the DISPLAY environment variable at the time it is loaded. Backgrounding and then foregrounding won't result in it sending it's stuff to a new X server. In fact, once it's attached to an X server, it can't be detached and reattached to another X server. This is a limitation of X -- the very limitation that I was talking about.

In addition to this, the screen session inherits and keeps the DISPLAY variable set during the X forwarding session. Therefore anything started in that screen session, regardless of where that screen has been reattached to will actually run through the original X tunnel and back on the original SSH client machine.

I.e. if I SSH -X to a host and start a screen session that you attach to. Even if I detach my screen, as long as my SSH session is open, you could attach to that same screen and use it to start applications on my desktop.

The problem with things like XRDP is exactly the same as the problem I've been explaining. If I start an application against one X server, then network transparency is from there on useless to me. I can't simply move it to another server. With windows, this has always been possible.

Which one of those lets me start an application on one machine and then continue using it on another machine like Windows has been able to do for well over a decade?

With all the praise that X gets for its network transparency, it's mostly unusable except on the highest bandwidth links because its synchronous calls and uber-chattiness make it unusable without adding a wrapper around it, like NX does. That's the irony. Pretty much any protocol can be called "network transparent" as long as that includes writing a suitably complicated wrapper for it. In this regard, X is no more network transparent than any other display technology.

Alright. Now when we can have the same desktop span multiple cards we'll have caught up to windows of 12 years ago.

The Beatles' use of drugs is epic and is regarded by the band members and others as a huge contributing factor to their work.

There seems to be a lot of evidence that the brain's ability to be logical and to be creative are actually competing functions. You can have both, but an expansion of one is often a contraction of the other. That drugs seem to diminish rational thinking and promote creativity is not exactly controversial.

For jobs that require a lot of both creativity and rational thought (e.g. software development) people report getting a bit stoned during the creative aspects, then doing the rational thought parts sober.

By mentioning "Atlas Shrugged" in your post, you have invalidated your invalidation. Original post stands.