Twitter is a walled garden. To @reply someone, you have to go through Twitter.
Facebook is even more of a walled garden. There's a large number of things you can only do with other people on facebook, once you have a facebook account. And, facebook may keep your data forever.
But URL shorteners? I'm all for making things open source and interoperable, but all this does is make a long URL into a short one. What would opening it up accomplish compared to, say, making Facebook work with OpenID and XFN, or have Skype adopt SIP, or have MSN, AIM, and Yahoo messenger adopt Jabber?
Those are some walled gardens that could be torn down -- but really, I don't need an account with TinyURL to make a TinyURL, nor do I need one to follow a TinyURL, nor is there any way TinyURL is locking me in to using it instead of bit.ly or tr.im.
So, I'm all for tearing the walls down around these gardens, but I'd rather not see the "walled garden" metaphor abused until it's no longer useful.
If they have a web browser, but not email, the first site you could send them to would be gmail dot com -- or, if you're security-conscious, h-t-t-p-s colon slash slash mail dot google dot com. It doesn't have to be email, either -- at that point, they'll also have a nice web-based chat client.
Granted, in most cases, it should be reasonable to copy between two computers -- you're on a damn network (the Internet, if nothing else), so you should be able to use some sort of messaging service. You've got even less excuse if they're both your own computer -- I haven't tried in awhile, but surely someone picked up Google Browser Sync.
But sometimes, it is useful -- for example, on the phone, talking someone through installing an IM client so you can do this the normal way. That, or relaying links in other media -- Internet Radio, podcasts, screencasts, etc.
My understanding was, it's similar to the PS2. One PS2 Linux developer compared it to "programming on a giant video card" -- the idea was that you have a very slow traditional CPU, and a couple of very powerful but very specialized coprocessors.
In other words, my assumption was that they threw the cheapest PowerPC CPU they could at it, and that the Cell was something accessed in addition to this weak, but general-purpose CPU.
But you sound like you know what you're talking about, so tell me: Am I entirely wrong about that?
Drop shadows help distinguish the window border, without making the border itself ginormous. By being sort of pseudo-3D, they tap into a part of our brain that's used to dealing with real 3D objects, and make it very easy for us to tell what's part of the window we're looking at, and what's part of another window.
Again, not important, but useful.
A similar example might be the fact that when I get a modal dialog box, that actually intends to block me from doing anything with its parent window until I deal with it, the entire parent window is darkened. Efficient to do with compositing, and does actually give me visual feedback for something I need to know, but not really worth the effort.
On the other hand, people are making good points that dropping 2D support and adding exclusively 3D, since you need 3D anyway, is probably a good thing. I just don't see it as a deal-breaker, certainly if it's the only advantage an OS has -- it's why I didn't use Vista.
our technology and scientific understanding of reality hasn't worked out how to define or measure 'energy',
Oh, bullshit. Real energy can be defined and quantified. It's even a standard unit.
I don't think it's a flaw in science that proponents of "energy" or "qi" can't actually define their terms.
It also seems to be tightly linked to awareness; if you don't believe it should be there, then it tends to obey.
How convenient for it.
Things we can see in the observable world tend to not care whether or not we believe in them. We would like to believe Earth is the center of the Universe, but that doesn't match our observations.
Furthermore, if it's defined as "life force", as cparker15 does, does this mean that if I stop believing in it, I'll die?
more traditional approaches to science have a lot of difficulty accepting that some aspects of reality exist relative to the state of the observer.
Actually, we know a fair amount about bias, and the placebo effect, and so on.
Nor is it entirely untestable. Get a few dozen people who believe Qi exists, and have them run the experiments, properly recorder. Then have impartial observers interpret the data.
When concrete examples of such unknown forces which fall into the same category as 'energy' rise, (and they do from time to time; we've had one or two recently),
Citation needed.
they are often ignored or otherwise awkwardly received and quietly forgotten about.
That isn't generally how science works.
Oh, it's true -- if you aren't able to play the peer-review game, and provide sufficient evidence through repeatable experiments, you'll likely be ignored, as you should be. But again, I could point to the round Earth, relativity, quantum physics, and evolution as ideas that many scientists would've loved to ignore, but couldn't, because that is how the world works.
Until then, one must test on a person by person basis.
The point is that these tests must be repeatable, and must adjust for bias, and be backed up by numbers -- for example, how many people tested had no result at all? Are the positive results you're getting outside the realm of what you'd expect by chance?
Like scientifically proving dreams exist;
What do you mean by "exist"?
We know that people dream. We know a lot about REM sleep. If you mean to suggest that dreams are actually real in some sense other than as nightly hallucinations, I'm very curious as to how you know that.
No? That tells me a lot about you; Like most taboo knowledge,
Taboo knowledge? That's funny...
Science has a lot of taboo knowledge.
I'd start with Robert O. Becker's book, Cross Currents [amazon.com].
At a quick glance, I don't see a lot of credentials on the man, and I do see some mentions ofhow unsubstantiated his recent research is. Moreover, I see "The New Plagues" in the table of contents -- AIDS, Autism, SIDS -- seriously, is he suggesting that AIDS is caused by "electorpollution", and not, say, dirty needles or sex?
Here's the thing: You've already tripped by bullshit-o-meter with your citations of "energy". When you start talking about
things that the power brokers of the world don't want you to know,
you're well into conspiracy-theory land. I'm sure they don't want me to know about the alien implant in your brain, either.
But it does raise an interesting question: Why did this guy write a book targeted at people like you and me, rather than trying to get his theories published in a scientific journal? If this stuff really works, I am sure the scientific community, at least, would be interested.
Ah, but your brain is vulnerable to the Rubber Hose. Software can actually be hardened against this -- for example, by changing the key to something unmemorizable before you travel, and leaving it with a sufficiently trustworthy and paranoid friend.
Wasn't the whole hoopla of JAVA to be cross platform development?
No, the point of Java was "compile once, run anywhere". So that's actually a small part of this:
I thought the whole deal was code in JAVA, and everyone will be able to run your app no matter what platform...
The rest of it is done with cross-platform libraries that abstract away the differences -- but you can do that in any language.
And it does work, mostly. I can, in fact, take little Java programs written and tested exclusively on 32-bit Windows, and run them on 64-bit Linux without any problems.
Games don't fit this bill because they tend to need to squeeze as much performance as possible out of each platform. While I think it's past time for games to be targeting something like Java or.NET, people who actually develop game engines will still occasionally dig into assembly for a specific platform -- preferring to squeeze that last 2% performance rather than develop a more stable, more easily portable game.
what a scam THAT turned out to be!
Oh, bullshit. It worked, and worked well. In fact, aside from the mobile area, where Java is still rampant, the main area Java tried and failed was the web, and that's only because people realized that JavaScript was good enough and didn't take ten seconds to start. And guess what? Web apps do run anywhere, without needing to be "ported", unless they go out of their way to depend on something that only exists on one platform.
So, the Java idea is alive and well, it's just not using Java.
And while people seem to be enamored with Flash lately, it's still quite easy to find Java applets, written back when it was popular. They even have a few advantages -- using real, native threads. And they work just fine on Konqueror on Linux, even when I doubt their authors had anything in mind beyond IE (or maybe Netscape) on maybe Windows 95 or 2k. So while Java applets may be slow to start, and annoying, they're certainly as portable as they promised.
Indeed, referring to the Cell as a "CPU" made me facepalm a bit. Ok, yes, you're going to want to use the Cell if you want any performance out of the PS3, but you're still using a PowerPC CPU.
Ok, to be clear: Say I've got a SIP phone, and you've got a SIP phone, and we're both behind NAT firewalls.
How do we connect to each other? How do we even address each other?
Because Skype solves this rather elegantly, and peer-to-peer, with UDP hole punching -- thus, Skype's own servers only establish connections, they don't have to pay the bandwidth for the actual calls.
I haven't looked at SIP in a long time, but I definitely didn't see any tricks like that. Has that changed? And if so, how does it work?
So you think about them for a while, apply changes and after so many months you...
...suddenly die, before they can get to the notary's office. It may not be perfect, but you're dead, so it's better than nothing, right?
Think of it as a software release.
I am. Release early and often.
Quick fixes mean you didn't think everything through and sometimes backfire horribly.
So does holding things back for months or years, trying to perfect them. See: Windows Vista, Duke Nukem Forever...
One advantage of using the notary's office is that people will gather to hear/see your message. Imagine that uncle Phil who will inevitably read the email just before or after a chicken choke session.
The former makes sense. I'm not sure I care much about the latter.
I'm not sure I like email as a delivery mechanism, but I definitely like the idea of making it electronic. Ideally, I'd make them accessible on a website, but encrypted (or at least password-protected), and then leave the keys at a notary's office. Still probably would want a deadman's switch of some sort, to prevent said notary from grabbing anything while I'm still alive.
trails when moving windows around, the glitchy redraw of background window contents, the 100% CPU use when dragging a window around the screen
I don't really see any of these, unless something's crashed.
the non-compositing desktop
All of these are, however, a side effect of that -- which also has the somewhat nice effect of not needing to store a bitmap of every single window, whether or not it's visible. I suppose RAM has gotten cheap enough that this doesn't matter, but it still kind of bothers me -- a properly designed app, revealed after being minimized or otherwise obscured, should redraw instantly and with minimal CPU. Any app that's done any sort of complex rendering, such that it would take seconds or minutes to recalculate, should cache that rendering -- but other apps could be much more efficient than storing that image.
The obvious example would be a text editor or a terminal window. Which takes less RAM, and which is easier to draw, an 850x525 pixel image, or an 80x24 array of characters? (Bear in mind that you're going to store the 80x24 array anyway.) At what point do we have enough RAM that I will stop caring?
All in all, I was quite surprised to see the push for Win 7 on netbooks, and I'm still surprised by how much people seem to like it. I would think a netbook wouldn't have enough RAM, let alone video hardware, to make compositing make sense -- certainly not enough to make it usable, if it was Vista's compositing.
the poorly designed UI of XP's explorer.
I'll give you that, mostly because I barely use explorer.
it's important to add 3D rendering.... he ability to do hardware-accelerated compositing, to render to a texture so you can have easy window previews, scaling, etc.
I think the problem is, we have a different idea of what "Important" means.
To me, that's a "would be nice" feature. At the end of the day, there are still plenty of places where a simple framebuffer makes sense, and our GUIs don't magically stop working by not being composited.
Contrast this to something moderately important, like proper SSD support -- doing it properly, avoiding hacks, and making sure data doesn't spontaneously evaporate...
Or something really important, like improving the Web as a software platform, thus simultaneously making deployment easier, and helping break Microsoft's stranglehold on the desktop. Perhaps you disagree about whether this should be done, but it's certainly a lot more significant of a change than just rendering to a texture.
More importantly, by using only the 3D pipeline vendors can simplify their hardware and software because they don't need to worry about 2D acceleration anymore,
They'll have to implement 2D in some form -- perhaps not accelerated, so XP would be slower, but it has to be there.
In the end, everybody benefits, even the luddite who just wants to run a bunch of full-screen terminal apps.
Yes, in the end, it's nice for everyone.
But not 'important'. Maybe I'm overreacting, but calling this 'important' just sounds like you're easily impressed.
Most noticeably, it has a user interface which doesn't look like it was designed in the mid 1990s.
In other words, Win 7 is prettier.
You do realize XP is themable, right?
It looks and 'feels' a hell of a lot better, as well as being vastly more customizable.
This might matter -- it depends what you mean.
Overall the UI in Windows 7 looks good and is very responsive.
You're basically telling me it's not, say, Vista, which looked good and wasn't responsive at all. Being responsive is a basic requirement, not something to be impressed about.
Similarly, whereas switching screens under XP frequently causes issues with a video that was playing fine on one screen not transferring to another without restarting playback, in Win 7 this seems to happen seamlessly.
For what it's worth, this works perfectly for me on Linux. Not just "transferring", either -- I can have the video window half on one screen, half on the other.
It is important to add the 3d effects to the UI, no matter what anybody says.
I say, sorry, no, it's not.
Some of the 2D stuff that you get from these hardware-accelerated compositing window managers, like drop shadows and zooming, is actually useful. Most of the 3D stuff is complete eye candy fluff.
Not that I'm complaining -- I do think it's a step in the right direction, though I wonder if it's the right approach. (If I put SVG stuff on my KDE4 desktop -- even as a wallpaper -- and zoom in, what happens?) But at the same time, you shouldn't need a 3D desktop to use a word processor.
But it's funny how my wife's old Macbook with the ancient GMA 950 chip runs OS-X liquid smooth.
Try Compiz on just about any card. It's the main reason I'm not down on this stuff in general -- because it can be done right. But again, requiring it, or overplaying its importance, is a mistake. At the end of the day, the GUI, the mouse, the web browser and web apps, are all innovations that burn more CPU than they ought to, but pay off immensely. 3D effects in a 2D UI, so far, give you about two minutes of "Ooh, Shiny", and then it's back to work, with very little difference.
What is the real thing? How can it be measured and tested? Couldn't it be used for evil, as well as good?
Cue the conceited, ill-informed rationalizations.
Oh, I'm not saying that this particular "Qi" is a good thing. I'm saying that there's no proof that "real" Qi exists, nor that it would be beneficial.
I'm also not aware of any evidence that this "EM pollution" is harmful, but at least we sort of know what it is, and can measure its effect, so I trust it a hell of a lot more than I'd trust "real" Qi, if you managed to produce any.
First off, trying to encrypt all the disks in all of those systems (some of which are HUGE) would be a massive undertaking.
But would be worth doing -- at the very least, you could start doing it for new systems.
Then there's the issue of trying to find an encryption system that's compatible across all these systems,
Why does it need to be the same system on all of them? LUKS works on Linux, Truecrypt on Windows... I'm not sure what there is for Solaris, though.
the additional overhead needed to do the encryption/decryption,
Negligible, if you're using hard disks. It might start to matter if you're using fast SSDs, or some ridiculous RAID, but plain vanilla hard disks are slow, and CPUs are fast.
the process of storing the encryption keys for all these systems.
That would be the real challenge -- maybe. At least on Linux, you could netboot them -- not secure, granted, but it means that if anyone runs off with a box without at least sniffing the network first, they'd have no keys. It also means you'd be able to update said keys without even walking to the box.
And that assumes the eagles can be trusted. Even if the eagle grabs Frodo, what will Frodo be able to do if the Eagle decides it wants the Precious for itself, and drops Frodo from a relatively high place, then flies down to grab the ring from his corpse?
On the other hand, how fast are the dragons, really? Especially if Gandalf alone is enough to stop them, with that glowy white light -- Gandalf on eagleback, problem solved. (Of course, this assumes he could do that before he fought the Balrog and was reborn...)
The truly paranoid make sure it never gets written to disk unencrypted in the first place, and then use thermite anyway once a disk needs to be disposed of, because it's not only paranoid, it's fun!
In fact, the truly paranoid will store a bit of thermite above the drive, ready to be triggered by some sort of alarm system. Several days of wiping it the way you describe may work, but all it takes is a bit of theft and you no longer have the chance.
And of course, even if the thermite fails to trigger, it's a lot faster to wipe only wherever the key is stored. Even if that fails, they still need a passphrase.
You know, we would really appreciate a way to solve this without just banning us from the service.
Actually, that's the purpose of just banning you from the service -- getting you pissed off enough to call this person, who, if she gets enough calls, would presumably let Facebook do whatever they want.
I don't agree with this -- I think Facebook really should just start actually deleting people's information when they make that request. But there have been times when I really think a website should've taken similar steps -- for example, one of the bigger ways for Google to help net neutrality would be to confirm which ISPs don't care about net neutrality, and show a message to those users (without actually blocking search) that encourages them to switch ISPs and/or write their congressmen.
Slashdot Mirror
I'm sorry...
Twitter is a walled garden. To @reply someone, you have to go through Twitter.
Facebook is even more of a walled garden. There's a large number of things you can only do with other people on facebook, once you have a facebook account. And, facebook may keep your data forever.
But URL shorteners? I'm all for making things open source and interoperable, but all this does is make a long URL into a short one. What would opening it up accomplish compared to, say, making Facebook work with OpenID and XFN, or have Skype adopt SIP, or have MSN, AIM, and Yahoo messenger adopt Jabber?
Those are some walled gardens that could be torn down -- but really, I don't need an account with TinyURL to make a TinyURL, nor do I need one to follow a TinyURL, nor is there any way TinyURL is locking me in to using it instead of bit.ly or tr.im.
So, I'm all for tearing the walls down around these gardens, but I'd rather not see the "walled garden" metaphor abused until it's no longer useful.
So you use HTML email and send it as a link.
If they have a web browser, but not email, the first site you could send them to would be gmail dot com -- or, if you're security-conscious, h-t-t-p-s colon slash slash mail dot google dot com. It doesn't have to be email, either -- at that point, they'll also have a nice web-based chat client.
This.
Granted, in most cases, it should be reasonable to copy between two computers -- you're on a damn network (the Internet, if nothing else), so you should be able to use some sort of messaging service. You've got even less excuse if they're both your own computer -- I haven't tried in awhile, but surely someone picked up Google Browser Sync.
But sometimes, it is useful -- for example, on the phone, talking someone through installing an IM client so you can do this the normal way. That, or relaying links in other media -- Internet Radio, podcasts, screencasts, etc.
My understanding was, it's similar to the PS2. One PS2 Linux developer compared it to "programming on a giant video card" -- the idea was that you have a very slow traditional CPU, and a couple of very powerful but very specialized coprocessors.
In other words, my assumption was that they threw the cheapest PowerPC CPU they could at it, and that the Cell was something accessed in addition to this weak, but general-purpose CPU.
But you sound like you know what you're talking about, so tell me: Am I entirely wrong about that?
Drop shadows help distinguish the window border, without making the border itself ginormous. By being sort of pseudo-3D, they tap into a part of our brain that's used to dealing with real 3D objects, and make it very easy for us to tell what's part of the window we're looking at, and what's part of another window.
Again, not important, but useful.
A similar example might be the fact that when I get a modal dialog box, that actually intends to block me from doing anything with its parent window until I deal with it, the entire parent window is darkened. Efficient to do with compositing, and does actually give me visual feedback for something I need to know, but not really worth the effort.
On the other hand, people are making good points that dropping 2D support and adding exclusively 3D, since you need 3D anyway, is probably a good thing. I just don't see it as a deal-breaker, certainly if it's the only advantage an OS has -- it's why I didn't use Vista.
our technology and scientific understanding of reality hasn't worked out how to define or measure 'energy',
Oh, bullshit. Real energy can be defined and quantified. It's even a standard unit.
I don't think it's a flaw in science that proponents of "energy" or "qi" can't actually define their terms.
It also seems to be tightly linked to awareness; if you don't believe it should be there, then it tends to obey.
How convenient for it.
Things we can see in the observable world tend to not care whether or not we believe in them. We would like to believe Earth is the center of the Universe, but that doesn't match our observations.
Furthermore, if it's defined as "life force", as cparker15 does, does this mean that if I stop believing in it, I'll die?
more traditional approaches to science have a lot of difficulty accepting that some aspects of reality exist relative to the state of the observer.
Actually, we know a fair amount about bias, and the placebo effect, and so on.
Nor is it entirely untestable. Get a few dozen people who believe Qi exists, and have them run the experiments, properly recorder. Then have impartial observers interpret the data.
When concrete examples of such unknown forces which fall into the same category as 'energy' rise, (and they do from time to time; we've had one or two recently),
Citation needed.
they are often ignored or otherwise awkwardly received and quietly forgotten about.
That isn't generally how science works.
Oh, it's true -- if you aren't able to play the peer-review game, and provide sufficient evidence through repeatable experiments, you'll likely be ignored, as you should be. But again, I could point to the round Earth, relativity, quantum physics, and evolution as ideas that many scientists would've loved to ignore, but couldn't, because that is how the world works.
Until then, one must test on a person by person basis.
The point is that these tests must be repeatable, and must adjust for bias, and be backed up by numbers -- for example, how many people tested had no result at all? Are the positive results you're getting outside the realm of what you'd expect by chance?
Like scientifically proving dreams exist;
What do you mean by "exist"?
We know that people dream. We know a lot about REM sleep. If you mean to suggest that dreams are actually real in some sense other than as nightly hallucinations, I'm very curious as to how you know that.
No? That tells me a lot about you; Like most taboo knowledge,
Taboo knowledge? That's funny...
Science has a lot of taboo knowledge.
I'd start with Robert O. Becker's book, Cross Currents [amazon.com].
At a quick glance, I don't see a lot of credentials on the man, and I do see some mentions ofhow unsubstantiated his recent research is. Moreover, I see "The New Plagues" in the table of contents -- AIDS, Autism, SIDS -- seriously, is he suggesting that AIDS is caused by "electorpollution", and not, say, dirty needles or sex?
Here's the thing: You've already tripped by bullshit-o-meter with your citations of "energy". When you start talking about
things that the power brokers of the world don't want you to know,
you're well into conspiracy-theory land. I'm sure they don't want me to know about the alien implant in your brain, either.
But it does raise an interesting question: Why did this guy write a book targeted at people like you and me, rather than trying to get his theories published in a scientific journal? If this stuff really works, I am sure the scientific community, at least, would be interested.
After all, the "power brokers of the world" hav
Ah, but your brain is vulnerable to the Rubber Hose. Software can actually be hardened against this -- for example, by changing the key to something unmemorizable before you travel, and leaving it with a sufficiently trustworthy and paranoid friend.
It works!
Cool, thanks! Now I can get all righteously indignant at the ubiquity of Skype.
Wasn't the whole hoopla of JAVA to be cross platform development?
No, the point of Java was "compile once, run anywhere". So that's actually a small part of this:
I thought the whole deal was code in JAVA, and everyone will be able to run your app no matter what platform...
The rest of it is done with cross-platform libraries that abstract away the differences -- but you can do that in any language.
And it does work, mostly. I can, in fact, take little Java programs written and tested exclusively on 32-bit Windows, and run them on 64-bit Linux without any problems.
Games don't fit this bill because they tend to need to squeeze as much performance as possible out of each platform. While I think it's past time for games to be targeting something like Java or .NET, people who actually develop game engines will still occasionally dig into assembly for a specific platform -- preferring to squeeze that last 2% performance rather than develop a more stable, more easily portable game.
what a scam THAT turned out to be!
Oh, bullshit. It worked, and worked well. In fact, aside from the mobile area, where Java is still rampant, the main area Java tried and failed was the web, and that's only because people realized that JavaScript was good enough and didn't take ten seconds to start. And guess what? Web apps do run anywhere, without needing to be "ported", unless they go out of their way to depend on something that only exists on one platform.
So, the Java idea is alive and well, it's just not using Java.
And while people seem to be enamored with Flash lately, it's still quite easy to find Java applets, written back when it was popular. They even have a few advantages -- using real, native threads. And they work just fine on Konqueror on Linux, even when I doubt their authors had anything in mind beyond IE (or maybe Netscape) on maybe Windows 95 or 2k. So while Java applets may be slow to start, and annoying, they're certainly as portable as they promised.
And by the way, it's spelled Java, and not JAVA.
Indeed, referring to the Cell as a "CPU" made me facepalm a bit. Ok, yes, you're going to want to use the Cell if you want any performance out of the PS3, but you're still using a PowerPC CPU.
Why is Verizon paying for a stupid employee's action?
Mostly because they haven't fired said stupid employee.
can go through firewalls just fine.
Ok, to be clear: Say I've got a SIP phone, and you've got a SIP phone, and we're both behind NAT firewalls.
How do we connect to each other? How do we even address each other?
Because Skype solves this rather elegantly, and peer-to-peer, with UDP hole punching -- thus, Skype's own servers only establish connections, they don't have to pay the bandwidth for the actual calls.
I haven't looked at SIP in a long time, but I definitely didn't see any tricks like that. Has that changed? And if so, how does it work?
So you think about them for a while, apply changes and after so many months you...
...suddenly die, before they can get to the notary's office. It may not be perfect, but you're dead, so it's better than nothing, right?
Think of it as a software release.
I am. Release early and often.
Quick fixes mean you didn't think everything through and sometimes backfire horribly.
So does holding things back for months or years, trying to perfect them. See: Windows Vista, Duke Nukem Forever...
One advantage of using the notary's office is that people will gather to hear/see your message. Imagine that uncle Phil who will inevitably read the email just before or after a chicken choke session.
The former makes sense. I'm not sure I care much about the latter.
I'm not sure I like email as a delivery mechanism, but I definitely like the idea of making it electronic. Ideally, I'd make them accessible on a website, but encrypted (or at least password-protected), and then leave the keys at a notary's office. Still probably would want a deadman's switch of some sort, to prevent said notary from grabbing anything while I'm still alive.
trails when moving windows around, the glitchy redraw of background window contents, the 100% CPU use when dragging a window around the screen
I don't really see any of these, unless something's crashed.
the non-compositing desktop
All of these are, however, a side effect of that -- which also has the somewhat nice effect of not needing to store a bitmap of every single window, whether or not it's visible. I suppose RAM has gotten cheap enough that this doesn't matter, but it still kind of bothers me -- a properly designed app, revealed after being minimized or otherwise obscured, should redraw instantly and with minimal CPU. Any app that's done any sort of complex rendering, such that it would take seconds or minutes to recalculate, should cache that rendering -- but other apps could be much more efficient than storing that image.
The obvious example would be a text editor or a terminal window. Which takes less RAM, and which is easier to draw, an 850x525 pixel image, or an 80x24 array of characters? (Bear in mind that you're going to store the 80x24 array anyway.) At what point do we have enough RAM that I will stop caring?
All in all, I was quite surprised to see the push for Win 7 on netbooks, and I'm still surprised by how much people seem to like it. I would think a netbook wouldn't have enough RAM, let alone video hardware, to make compositing make sense -- certainly not enough to make it usable, if it was Vista's compositing.
the poorly designed UI of XP's explorer.
I'll give you that, mostly because I barely use explorer.
it's important to add 3D rendering.... he ability to do hardware-accelerated compositing, to render to a texture so you can have easy window previews, scaling, etc.
I think the problem is, we have a different idea of what "Important" means.
To me, that's a "would be nice" feature. At the end of the day, there are still plenty of places where a simple framebuffer makes sense, and our GUIs don't magically stop working by not being composited.
Contrast this to something moderately important, like proper SSD support -- doing it properly, avoiding hacks, and making sure data doesn't spontaneously evaporate...
Or something really important, like improving the Web as a software platform, thus simultaneously making deployment easier, and helping break Microsoft's stranglehold on the desktop. Perhaps you disagree about whether this should be done, but it's certainly a lot more significant of a change than just rendering to a texture.
More importantly, by using only the 3D pipeline vendors can simplify their hardware and software because they don't need to worry about 2D acceleration anymore,
They'll have to implement 2D in some form -- perhaps not accelerated, so XP would be slower, but it has to be there.
In the end, everybody benefits, even the luddite who just wants to run a bunch of full-screen terminal apps.
Yes, in the end, it's nice for everyone.
But not 'important'. Maybe I'm overreacting, but calling this 'important' just sounds like you're easily impressed.
Most noticeably, it has a user interface which doesn't look like it was designed in the mid 1990s.
In other words, Win 7 is prettier.
You do realize XP is themable, right?
It looks and 'feels' a hell of a lot better, as well as being vastly more customizable.
This might matter -- it depends what you mean.
Overall the UI in Windows 7 looks good and is very responsive.
You're basically telling me it's not, say, Vista, which looked good and wasn't responsive at all. Being responsive is a basic requirement, not something to be impressed about.
Similarly, whereas switching screens under XP frequently causes issues with a video that was playing fine on one screen not transferring to another without restarting playback, in Win 7 this seems to happen seamlessly.
For what it's worth, this works perfectly for me on Linux. Not just "transferring", either -- I can have the video window half on one screen, half on the other.
It is important to add the 3d effects to the UI, no matter what anybody says.
I say, sorry, no, it's not.
Some of the 2D stuff that you get from these hardware-accelerated compositing window managers, like drop shadows and zooming, is actually useful. Most of the 3D stuff is complete eye candy fluff.
Not that I'm complaining -- I do think it's a step in the right direction, though I wonder if it's the right approach. (If I put SVG stuff on my KDE4 desktop -- even as a wallpaper -- and zoom in, what happens?) But at the same time, you shouldn't need a 3D desktop to use a word processor.
But it's funny how my wife's old Macbook with the ancient GMA 950 chip runs OS-X liquid smooth.
Try Compiz on just about any card. It's the main reason I'm not down on this stuff in general -- because it can be done right. But again, requiring it, or overplaying its importance, is a mistake. At the end of the day, the GUI, the mouse, the web browser and web apps, are all innovations that burn more CPU than they ought to, but pay off immensely. 3D effects in a 2D UI, so far, give you about two minutes of "Ooh, Shiny", and then it's back to work, with very little difference.
makes a sick mockery of the real thing.
Ok, so tell me...
What is the real thing? How can it be measured and tested? Couldn't it be used for evil, as well as good?
Cue the conceited, ill-informed rationalizations.
Oh, I'm not saying that this particular "Qi" is a good thing. I'm saying that there's no proof that "real" Qi exists, nor that it would be beneficial.
I'm also not aware of any evidence that this "EM pollution" is harmful, but at least we sort of know what it is, and can measure its effect, so I trust it a hell of a lot more than I'd trust "real" Qi, if you managed to produce any.
First off, trying to encrypt all the disks in all of those systems (some of which are HUGE) would be a massive undertaking.
But would be worth doing -- at the very least, you could start doing it for new systems.
Then there's the issue of trying to find an encryption system that's compatible across all these systems,
Why does it need to be the same system on all of them? LUKS works on Linux, Truecrypt on Windows... I'm not sure what there is for Solaris, though.
the additional overhead needed to do the encryption/decryption,
Negligible, if you're using hard disks. It might start to matter if you're using fast SSDs, or some ridiculous RAID, but plain vanilla hard disks are slow, and CPUs are fast.
the process of storing the encryption keys for all these systems.
That would be the real challenge -- maybe. At least on Linux, you could netboot them -- not secure, granted, but it means that if anyone runs off with a box without at least sniffing the network first, they'd have no keys. It also means you'd be able to update said keys without even walking to the box.
Disk i/o is going to be the limiting factor long before the speed of decryption comes into play.
I used to think that. Now I have an SSD in my laptop, and I'm not sure anymore.
And that assumes the eagles can be trusted. Even if the eagle grabs Frodo, what will Frodo be able to do if the Eagle decides it wants the Precious for itself, and drops Frodo from a relatively high place, then flies down to grab the ring from his corpse?
On the other hand, how fast are the dragons, really? Especially if Gandalf alone is enough to stop them, with that glowy white light -- Gandalf on eagleback, problem solved. (Of course, this assumes he could do that before he fought the Balrog and was reborn...)
The truly paranoid make sure it never gets written to disk unencrypted in the first place, and then use thermite anyway once a disk needs to be disposed of, because it's not only paranoid, it's fun!
In fact, the truly paranoid will store a bit of thermite above the drive, ready to be triggered by some sort of alarm system. Several days of wiping it the way you describe may work, but all it takes is a bit of theft and you no longer have the chance.
And of course, even if the thermite fails to trigger, it's a lot faster to wipe only wherever the key is stored. Even if that fails, they still need a passphrase.
You know, we would really appreciate a way to solve this without just banning us from the service.
Actually, that's the purpose of just banning you from the service -- getting you pissed off enough to call this person, who, if she gets enough calls, would presumably let Facebook do whatever they want.
I don't agree with this -- I think Facebook really should just start actually deleting people's information when they make that request. But there have been times when I really think a website should've taken similar steps -- for example, one of the bigger ways for Google to help net neutrality would be to confirm which ISPs don't care about net neutrality, and show a message to those users (without actually blocking search) that encourages them to switch ISPs and/or write their congressmen.
That may be how things work.
Is that how they should work?