Does NT share code with Accent or Mach, or does it just share concepts with it?
From what I understand, it shares code. NT definitely shares Accent code. But you are getting the order of who shared with whom backwards. If we had a family tree:
Accent -> Mach -> XNU Accent -> NT
However, that is a bit of an obstacle to Apple "giving back" code, especially kernel code, as it would involve Apple doing work to port that code to the BSD in question.
I think that is a mostly unreasonable expectation. Darwin is (mostly) open source, and well documented. If a BSD team wanted to take some code from Darwin I think they have to do the porting effort. Anyway, my main point to GP was that the kernels aren't much alike. You seem to be arguing my 98% dissimilar is too high, and maybe it should be 90% dissimilar. Even if true, I think that addresses GP's objection that Apple isn't contributing kernel code back....
You got the spirit of how this works. I've never done non permitter on AIX, though I have used it as an end user, so I can't comment. But on Solaris, exactly. The big culprit though is x86 hardware.
You had asked for people who had done it. I don't see the cost savings either.
Your DMZ style servers are shattered. That's why they are imaged and easy to restore. Your servers aren't your security layer. Where you want security you use much more secure OSes. For example a mainframe, i-Series. Solaris 10+ using Trusted Solaris. I did it with VMS but that was years ago. I've used hardened Linuxes, but it is still risky since x86 hardware doesn't handle security well. You wouldn't use Windows or a typical Linux for your secure boxes.
Webservers are hacked because they are running way too many services too casually. For example applications which tie into advertising are a notorious vector for attack. You just don't do that on boxes you care about.
And yes you have a lot of monitoring. And the real question is whether you want only perimeter defense. I can't see using this strategy for a company that doesn't already want multiple permitters. It would be too expensive. The kinds of companies this works for are ones that already have to have multiple levels where one more is no big deal. So for example they separate out DBA roles so a DBA can't just alter data by himself.
The company can demand you return their property. They can't however do an inspection to determine if you have. What happens is, it shifts the burdon of proof. The company has to prove by preponderance of the evidence that you do have their property so as to get a court order requiring you to return it.... If the company says they want to erase your laptop, you say you already deleted their stuff, they can't do much.
That's one of the reasons companies might want DRMed data and use application with much more DRM support if they want to move to this sort of remote model.
IT will confiscate your laptop as it holds company secrets.
That's called theft, he can call the police and get it back. IT has to request the secrets be disposed of properly they can't sieze the laptop anymore than they could break into his house and steal files he kept at home.
I worked with companies that had untrusted client machines working with servers. The servers just have to be hardened. Instead of a perimeter defense model, typical corporate security, you move to a interior defense model where client facing machines are individually hardened and secured. Someone else mentioned a typical DMZ, that's a good analogy. Just imagine a huge DMZ, a small secure area and a small fully external area. What you do today inside the DMZ is what you have to do over the bulk of your network.
It will end with recentralization, a push back towards higher IT budgets. As less and less of a company's crucial systems are under its control the fragmented IT maintenance costs skyrocket to keep systems in sinc with one another. Suddenly centralization becomes a source of obvious savings....
KDE is not a full GUI. Qt is the GUI toolkit; KDE is the applications built with the toolkit.
How does the existence of a toolkit (and QT is actually more of a framework at this point) disprove that KDE is a GUI?
The term "desktop" has become synonymous with "window manager".
BS. Both Windows and Mac use the term window manager properly in all their documentation. Microsoft window manager is called DWM (desktop window manager) and was advertised as a replacement of the GDI window manager from pre-Vista because of the Aero upgrade. They used Window manager to mean the software responsible for things like buffering, placement,... of windows, they used it correctly. Apple advertises constantly the advantage of Quartz Compositor (the window manager of Aqua) and its features on their website. Their end users are familiar with terms like Quartz Extreme (window manager acceleration), Core Image... Apple doesn't confuse their window manager with their desktop (whatever desktop means). The default window manager for KDE is KWin, and KDE can be used with a different window manager, and KWin can be used by itself or even with Gnome.
I'm not being pedantic you are using the word incorrectly. The word means exactly what it has meant for several decades. The general public doesn't use the term window manager at all. You have a 5 digit user number which means you are old enough to use the term properly.
As for your final comment about the direction of Gnome 3 and using a tablet / phone interface I agree. But I think it is healthy for Linux GUIs to be very different from one another and aim for different niches. Linux has the ability to offer real choice, and I think diverging offers real choice.
Finally on QT vs. GTK. Yeah I've always felt QT is a much more logical framework than GTK.
Or they could go back to the original purpose. Inexpensive systems designed for portability. Essentially what smart phones are being used today except without the phone part.
Windows has phenomenal driver support and quite often driver bypasses that deal with buggy hardware. Linux however had drivers that layer beautifully and that can be terrific when the problem is broken hardware.
Windows is pretty good but not outstanding on usability. End users frequently can't figure out how to do what they want. Right now iOS is probably the usability king.
That being said Windows is generally better than most Linux GUIs. You used the term window manager, you are using it wrong. The window manager for Windows is DWM and while DWM is very cool it is not particular user friendly or unfriendly.
KDE used to be the dominant desktop. Then a few things went badly and it was clearly in 2nd place falling. Things were going so well for Gnome over KDE that KDE was starting to get grouped with XFCE, LXDE... in the "also rans". Now it is looking like Gnome dropped the ball so badly with Ubuntu and with Gnome 3, and KDE 4's advantages are showing through that KDE is going to get a chance to be the dominant desktop again.
I think you mean when KDE apps run under another GUI. You aren't using the word "window manager" properly. You can actually other window managers with the KDE GUI and get different behaviors.
Anyway... KDE is a full GUI, the apps inherit the behavior of their native GUI.
Windows power users have always been the most hostile to Linux switching. They have learned how to do stuff in Windows but don't have broad computer knowledge. Switching takes that away from them. Up or down in terms of skill set and things work better.
Dell has gone back and forth. When Dell started in Texas it was an grey market IBM dealer. Later it became the high end of the grey box manufacturers. Dell's server stuff was always pretty good but it allowed the consumer line to be driven down to margins where service was frankly terrible.
It is hard to know what Dell is anymore. I'm not sure Dell even knows.
Re:next we'll hear that Dell is in trouble...
on
Dell Ditches Netbooks
·
· Score: 4, Insightful
I agree. A $100-200 device, 7 in with a keyboard running Android makes sense. An 11 inch windows "netbook" for $350 doesn't when you can buy a laptop for not much more.
Well Oracle was another company urging people to shift to thin clients that discovered it had a windows culture. The big 3 with the failures were Sun, Oracle and IBM. It was egg on the face for all of them.
I don't really think of Sun as a failure. Ultimately of all the workstation makers: Dec, SGI, Sun, IBM, NeXT, Symbolics, Xerox, Apollo... they were one of the most successful, certainly one of the most influential. Ultimately I think they failed because their niches disappeared. No one buys $5k-50k desktop systems. On the server side they never had the value add of the minis / mainframes nor the low cost of the Windows / Linux solutions. Sun existed to fill narrowing niches that existed in the 80s and 90s that just don't exist today.
Once you are there, the second half will be easier: that such a comitee is there not only to make sure IT meets corporate goals but that IT adds up as much value as it can by allowing it to stand what can and can't be done and where are the new opportunities that technology opens for the company, that they are a key element to stablish those corporate goals.
That's exactly what a steering committee does.
Just for an example: IT, and I mean here just plain old IT, nothing fashionable involving innovation, is the most achievable vector for cost cuts through mere automation: you just intelligently increase IT budget by a 10% and you gain a 10% savings all along the company (of course, numbers are made up here). But then, what do you see? "those are hard days, so we all have to make sacrifices: all departments get their budget cut by 10%"
I cover why this makes sense to not IT from a 5 year planning perspective in my book. And I also deal with the longer term costs of this. I agree with you 100% that this is a problem.
Anyway you seem to understand better than most that IT policy and corporate policy are not identical, you and I are having a different conversation.
Slashdot Mirror
From what I understand, it shares code. NT definitely shares Accent code. But you are getting the order of who shared with whom backwards. If we had a family tree:
Accent -> Mach -> XNU
Accent -> NT
I think that is a mostly unreasonable expectation. Darwin is (mostly) open source, and well documented. If a BSD team wanted to take some code from Darwin I think they have to do the porting effort. Anyway, my main point to GP was that the kernels aren't much alike. You seem to be arguing my 98% dissimilar is too high, and maybe it should be 90% dissimilar. Even if true, I think that addresses GP's objection that Apple isn't contributing kernel code back....
Follow the chain. And then prove it, don't just assert it.
You got the spirit of how this works. I've never done non permitter on AIX, though I have used it as an end user, so I can't comment. But on Solaris, exactly. The big culprit though is x86 hardware.
You had asked for people who had done it. I don't see the cost savings either.
Your DMZ style servers are shattered. That's why they are imaged and easy to restore. Your servers aren't your security layer.
Where you want security you use much more secure OSes. For example a mainframe, i-Series. Solaris 10+ using Trusted Solaris. I did it with VMS but that was years ago. I've used hardened Linuxes, but it is still risky since x86 hardware doesn't handle security well. You wouldn't use Windows or a typical Linux for your secure boxes.
Webservers are hacked because they are running way too many services too casually. For example applications which tie into advertising are a notorious vector for attack. You just don't do that on boxes you care about.
And yes you have a lot of monitoring. And the real question is whether you want only perimeter defense. I can't see using this strategy for a company that doesn't already want multiple permitters. It would be too expensive. The kinds of companies this works for are ones that already have to have multiple levels where one more is no big deal. So for example they separate out DBA roles so a DBA can't just alter data by himself.
http://en.wikipedia.org/wiki/KWin (KDE window manager)
http://en.wikipedia.org/wiki/Metacity (Gnome 1,2 window manager), http://en.wikipedia.org/wiki/Mutter_(window_manager) (Gnome 3 window manager). Those are window managers.
And Motif is a widget toolket same as QT or GTK. MWM the Motif Window Manager is a window manager written in Motif.
The company can demand you return their property. They can't however do an inspection to determine if you have. What happens is, it shifts the burdon of proof. The company has to prove by preponderance of the evidence that you do have their property so as to get a court order requiring you to return it.... If the company says they want to erase your laptop, you say you already deleted their stuff, they can't do much.
That's one of the reasons companies might want DRMed data and use application with much more DRM support if they want to move to this sort of remote model.
That's called theft, he can call the police and get it back. IT has to request the secrets be disposed of properly they can't sieze the laptop anymore than they could break into his house and steal files he kept at home.
I worked with companies that had untrusted client machines working with servers. The servers just have to be hardened. Instead of a perimeter defense model, typical corporate security, you move to a interior defense model where client facing machines are individually hardened and secured. Someone else mentioned a typical DMZ, that's a good analogy. Just imagine a huge DMZ, a small secure area and a small fully external area. What you do today inside the DMZ is what you have to do over the bulk of your network.
It will end with recentralization, a push back towards higher IT budgets. As less and less of a company's crucial systems are under its control the fragmented IT maintenance costs skyrocket to keep systems in sinc with one another. Suddenly centralization becomes a source of obvious savings....
Depends how you think of Palm.
How does the existence of a toolkit (and QT is actually more of a framework at this point) disprove that KDE is a GUI?
BS. Both Windows and Mac use the term window manager properly in all their documentation. Microsoft window manager is called DWM (desktop window manager) and was advertised as a replacement of the GDI window manager from pre-Vista because of the Aero upgrade. They used Window manager to mean the software responsible for things like buffering, placement, ... of windows, they used it correctly. Apple advertises constantly the advantage of Quartz Compositor (the window manager of Aqua) and its features on their website. Their end users are familiar with terms like Quartz Extreme (window manager acceleration), Core Image... Apple doesn't confuse their window manager with their desktop (whatever desktop means). The default window manager for KDE is KWin, and KDE can be used with a different window manager, and KWin can be used by itself or even with Gnome.
I'm not being pedantic you are using the word incorrectly. The word means exactly what it has meant for several decades. The general public doesn't use the term window manager at all. You have a 5 digit user number which means you are old enough to use the term properly.
As for your final comment about the direction of Gnome 3 and using a tablet / phone interface I agree. But I think it is healthy for Linux GUIs to be very different from one another and aim for different niches. Linux has the ability to offer real choice, and I think diverging offers real choice.
Finally on QT vs. GTK. Yeah I've always felt QT is a much more logical framework than GTK.
Or they could go back to the original purpose. Inexpensive systems designed for portability. Essentially what smart phones are being used today except without the phone part.
I'd say this.
Windows has phenomenal driver support and quite often driver bypasses that deal with buggy hardware.
Linux however had drivers that layer beautifully and that can be terrific when the problem is broken hardware.
It depends why it failed to mount.
You have a 4 digit /. number and haven't had problems with USB on Linux? Seriously during the 2.0 kernel days you didn't have those sorts of problems?
Windows is pretty good but not outstanding on usability. End users frequently can't figure out how to do what they want. Right now iOS is probably the usability king.
That being said Windows is generally better than most Linux GUIs. You used the term window manager, you are using it wrong. The window manager for Windows is DWM and while DWM is very cool it is not particular user friendly or unfriendly.
KDE used to be the dominant desktop. Then a few things went badly and it was clearly in 2nd place falling. Things were going so well for Gnome over KDE that KDE was starting to get grouped with XFCE, LXDE ... in the "also rans". Now it is looking like Gnome dropped the ball so badly with Ubuntu and with Gnome 3, and KDE 4's advantages are showing through that KDE is going to get a chance to be the dominant desktop again.
I think you mean when KDE apps run under another GUI. You aren't using the word "window manager" properly. You can actually other window managers with the KDE GUI and get different behaviors.
Anyway... KDE is a full GUI, the apps inherit the behavior of their native GUI.
I assume you meant Gnome 2. For Ubuntu just use Mint's version:
sudo gedit /etc/apt/sources.list
At the end of the file add this line:
deb http://packages.linuxmint.com/ lisa main upstream import
Save your file and close. Then, run this sequence of commands:
sudo apt-get update
sudo apt-get install linuxmint-keyring
sudo apt-get update
sudo apt-get install mint-meta-mate
and you are done.
Windows power users have always been the most hostile to Linux switching. They have learned how to do stuff in Windows but don't have broad computer knowledge. Switching takes that away from them. Up or down in terms of skill set and things work better.
Dell has gone back and forth. When Dell started in Texas it was an grey market IBM dealer. Later it became the high end of the grey box manufacturers. Dell's server stuff was always pretty good but it allowed the consumer line to be driven down to margins where service was frankly terrible.
It is hard to know what Dell is anymore. I'm not sure Dell even knows.
I agree. A $100-200 device, 7 in with a keyboard running Android makes sense. An 11 inch windows "netbook" for $350 doesn't when you can buy a laptop for not much more.
Well Oracle was another company urging people to shift to thin clients that discovered it had a windows culture. The big 3 with the failures were Sun, Oracle and IBM. It was egg on the face for all of them.
I don't really think of Sun as a failure. Ultimately of all the workstation makers: Dec, SGI, Sun, IBM, NeXT, Symbolics, Xerox, Apollo... they were one of the most successful, certainly one of the most influential. Ultimately I think they failed because their niches disappeared. No one buys $5k-50k desktop systems. On the server side they never had the value add of the minis / mainframes nor the low cost of the Windows / Linux solutions. Sun existed to fill narrowing niches that existed in the 80s and 90s that just don't exist today.
How is this related to my post?
That's exactly what a steering committee does.
I cover why this makes sense to not IT from a 5 year planning perspective in my book. And I also deal with the longer term costs of this. I agree with you 100% that this is a problem.
Anyway you seem to understand better than most that IT policy and corporate policy are not identical, you and I are having a different conversation.
Upper management and IT management are not necessarily the same thing, and don't necessarily have the same objectives.