I'm an earthlink customer and when I mistype something it sends me to an eartlink page that suggests one or two site URL's one of which is almost always the right one. The page itself has no ads on it.
Set aside for the moment the fact that they flat out REFUSE to sell us what we've been telling them we want and will pay for and have ALREADY gotten to work on our own. This confusing mixed message yes I mean no I mean not yet nonsense is confirming to us that Dell has its head way up yonder the colon. Is it any wonder why Dell and all the lesser companies are dying?
Don't know don't use it but is it possible that Windows is finally butting heads with its own complexity and the fact that it's a big object stitched together by lots of dissimilar design groups? Or maybe it's a problem that they wanted to create a Windows for everyone everywhere on every device from Cell phones to home theaters to PCs to who knows what?
After all they proudly trotted out their new SOUND that they took a few million dollars to develop. If that isn't pitching into the brambles and losing track of what you're doing, I don't know what is.
About one in 10 SMS messages I get are from Sprint (I'm a Sprint customer) advertising a service, ringtone or some other downloadable. So excuse me while I don't feel Verizon's pain.
But the same can be said for people who own any kind of high end phone now. On average people are replacing their phones every 18 months. Moreover the market for unlocked phones is huge so people who have to spend MSRP is already there. So the price point I think is not that important. Afterall a Samsung A900-MM is about $350. Smartphones are in the $400-$600 range now. When phone/PDA combos first came out they were $1000. Seems to me that Apple is smart to go after the MS Windows-for-phones segment of the market.
No what I meant was that it is shortsighted to implement Virtualization on a host system that doesn't truly have that function nor does it typically have the free overhead to do it in the first place. One of the key reason VM or z/OS or AIX play so well in this field is that they scale up to dozens of processors for a single 'host' system image. In fact that's how we run Linux on a mainframe, not as an LPAR but as a guest under VM. Similarly we can spool up t0 4 simultaneous ADSM (Tivoli Storage Manager) hosts in a single system image implemented across multiple processor complexes. If we tried to to do that with a Windows host we'd wind up with a bigger physical server than we could have implemented putting each one in its own hardware. This was the problem we had with Citrix Metaframe. In order to get 30 client 'hosts' we had to build the biggest goddamn Windows server available at the time. It just turned out to be flat out cheaper to do it with a heavy client.
If someone can show me how to run Windows on 24-48 processors for a single system image that I can then carve up into 100 VMs then I'll be impressed.
Well then that's the problem. If you can be a good executive assistant for that pay or a grunt level legal editor then there is no real upside to being a coder. And may I add that many many many people don't actually want to live in West Gopher. The problem with places like Sioux Falls, SD or Graham, NC is that there's nothing there.
Im sure Windows is fine for troop ship management, slow rides up the Thames past Buckingham and the usual lot of do nothing feather in the hat bilge the UK Navy is up to.
That's the problem. There is a shortage of people willing to work 80 hrs a week for $60K with a relocation to West Gopher Hole, South Dakota.
Blah blah blah not enough high schools teach Microsoft coding skills. Blah blah blah not enough Indians coming to America to debug Redmond's code. Blah blah blah we need more wage slaves.
Assuming zero or near zero growth. Over time when capacity hits the ceiling then you'll be forced to deploy new hardware and because you've already hit some arbitrary ceiling the new hardware will be very large and very powerful. Hopefully you won't have to deploy the app on its own hardware or figure out how to move images around from server to server. VM environments can be quite a bit more complex from a capacity planning and load balancing perspective unless each of the apps is really small and trivial. Moreover it's very hard to isolate specific instances for things like VPN tunnels, health checking, port scanning and NIC utilization. It more or less commits you to running different little apps for the same customer on the same box. Now that can work but from the point of view of running a hosting center for 1500 different customers I'm sure there's a lot of empirical evidence to suggest what the real quantitative material benefits are.
No but if I need to actually virtualize something I'll run it on a system that actually runs virtualization well with low overhead that allows me to dial in the specific capacity I need. Just because I can partition something arbitrarily doesn't mean it's running in the right sized sandbox. And if your apps are really trivial enough to do that then why are they being hosted at all.
Port them to AIX? No of course not. That would mean they're big enough and important enough to do that, which clearly they're not.
BTW wait 18 months when you run out of overhead unevenly and you're forced to move specific applications to their own hardware and/or you're forced to create new virtualized servers on the new hardware. You will find that either the hardware you need to deploy is either insanely large or, those applications will wind up on their own servers anyhow.
You can't do virtualized port scanning, health checking. You can't easily manage a virtualized VPN tunnel and you're stuck with more or less keeping all the same customers on the same iron anyway. Plus patch management really isn't much easier and capacity planning is much much harder to do.
Aside from developers and a tiny group of specialists who need access to a particular app? In the datacenter world this is anathema. No one running a gaggle of boxes would ever seriously consider this and get paid for it. Cheaper and easier by far to throw up one more server and spend the 0.04 FTE (1/25th of a person) it takes to run it.
And if you seriously considering multi image same system partitioning of Windows then you my friend need to re examine what it is you're doing. LPARs are not for Windows code. Go out and by an iSeries midrange or an AIX machine.
Was $99 for installation on 3 machines. In the 4 years I've had it I've seen no need to replace it or upgrade it. So that's $33 per machine/48 months = 69 cents/month. If MS thinks I'm going to pay a 2180% premium they have been smoking too much crack. Well to me it's just another nudge off any and all Microsoft code forever.
Never mind that the Arabic/Persian world completely ignored the Renaissance, the Enlightenment and the Industrial Revolution. I'm sure and my Marxist teachers in college swear on a stack of Howard Zinn's that it's some white imperialist capitalist Zionist nation's fault. It always is.
In North Carolina if a minor screams something out which is offensive, inflammatory or threatening to more than one person it's considered terrorism or making a terroristic threat. For 16 yo and above it can be charged as a felony and while it's not often charged as one, it's not unheard of. So one can extrapolate and guess that making threats online is also terrorism or making a terroristic threat. Free speech be damned we're talking PATRIOT act here. And what better group of people to apply it to than people who can't vote but who are charged as adults in the first place?
NASA is talking about going to the moon in order to open up funding streams for all of the precursor projects which will then be used for some other purpose, ostensibly a revenue generating one. Sorry to crush your moon rocks but we are not going to back to the moon this century. If anyone gets there before the year 2100 it will be India or China, if only to say YAAAAY FOR US !!!!
But manned spaceflight out of the orbit of the earth is in fact dead and over, forever, or if not, for the next 100+ years. No one wants to do it. Governments don't want to pay for it and we don't have the attention span for it. In fact I'd put money on the complete elimination of manned spaceflight by the year 2015. We will have officially spent enough money to piss everyone off by that time.
Phase 2 is the complete elimination of all space science, in space, by 2020. Expect that orbital telescopes, research satellites etc will all be defunded by then. Unless there is a commercial or military purpose for space science, it will be killed.
It was a pretty good run but now it's over. I grew up pouring over every detail, every photograph, every newscast, every lay science paper for everything associated with the Gemini and Apollo programs. I had a family member who worked in both programs. It was magic.
But by the time Skylab was discontinued it was clear that NASA was looking to get into the commercial heavy lifting business, ergo Space Shuttle. But NASA didn't bank on the expense and complexity of Space Shuttle, nor did they anticipate smaller payloads becoming the norm. So the Air Force became NASA's only paying customer. They're the only people who have a need for the capacity of Space Shuttle. So NASA is just treading water until Space Shuttle and ISS are killed off. They hope to have another heavy lifter online by then but if they don't then that's that. End of story. We'll be able to go the ESA or India, Japan, Russia or China for launch capability by then and NASA will have ceased to have a purpose.
The key, the only key to successfully implementing security in Vista or any other MS codebase is not to work from the assumption that everything can be locked down 100% and nothing bad can ever be made to happen. That's just stupid. Feel free to write an airliner fly by wire system and charge consumers a million dollars for each copy.
No, the problem with Vista and XP and.....is that they think they can both build an elegant system which simultaneously checks everything all the time and prevents an unknown thing from occuring.
The approach should be 180 different from that. It should be to assume that problems will occur and simply mitigate the damage or the extent of the damage they can cause. Build it such that even if it's botnet'd that the outbound traffic is blocked and the damage is limited to that one machine. Build it so that buffer overflows only go as far as that one application or subsystem.
Sandbox sandbox sandbox and when you're done, virtualize it. I really thought that when Intel announced the dual core processors we'd finally see some progress. We'd see one of the two cores devoted to all of the security and protection functions from port scanning, to encryption, to firewalling (in both directions) and so on. But instead we got the dancing bears 3D lucite animated we spent 10 million dollars developing the SOUND that the taskbar makes interface.
What a colossal waste of time and effort. Most of the problems associate with Windows security are DIRECTLY traceable to the fact that none of any of the original problems were ever addressed. They were embraced and layered over with yet more code. Sometimes the code is a workaround, sometimes it's nothing more sophisticated than an alert.
"Do you want to execute this program?"
I don't know. Is it bad? Why don't you tell me? Why don't you give it a whirl in a contained environment, let me know if it's bad, and if not write a sig to the system that lets it know the next time I want to run it it checks that sig for verification purposes.
Slashdot Mirror
I'm an earthlink customer and when I mistype something it sends me to an eartlink page that suggests one or two site URL's one of which is almost always the right one. The page itself has no ads on it.
Set aside for the moment the fact that they flat out REFUSE to sell us what we've been telling them we want and will pay for and have ALREADY gotten to work on our own. This confusing mixed message yes I mean no I mean not yet nonsense is confirming to us that Dell has its head way up yonder the colon. Is it any wonder why Dell and all the lesser companies are dying?
Don't know don't use it but is it possible that Windows is finally butting heads with its own complexity and the fact that it's a big object stitched together by lots of dissimilar design groups? Or maybe it's a problem that they wanted to create a Windows for everyone everywhere on every device from Cell phones to home theaters to PCs to who knows what?
After all they proudly trotted out their new SOUND that they took a few million dollars to develop. If that isn't pitching into the brambles and losing track of what you're doing, I don't know what is.
Tune in to Fox News and hear the truth. Science and Reality are evil plots to turn your kids into muslim lesbian liberal queer feminists.
thanks that's good to know
About one in 10 SMS messages I get are from Sprint (I'm a Sprint customer) advertising a service, ringtone or some other downloadable. So excuse me while I don't feel Verizon's pain.
But the same can be said for people who own any kind of high end phone now. On average people are replacing their phones every 18 months. Moreover the market for unlocked phones is huge so people who have to spend MSRP is already there. So the price point I think is not that important. Afterall a Samsung A900-MM is about $350. Smartphones are in the $400-$600 range now. When phone/PDA combos first came out they were $1000. Seems to me that Apple is smart to go after the MS Windows-for-phones segment of the market.
who spend $2700 for an Apple monitor?
No what I meant was that it is shortsighted to implement Virtualization on a host system that doesn't truly have that function nor does it typically have the free overhead to do it in the first place. One of the key reason VM or z/OS or AIX play so well in this field is that they scale up to dozens of processors for a single 'host' system image. In fact that's how we run Linux on a mainframe, not as an LPAR but as a guest under VM. Similarly we can spool up t0 4 simultaneous ADSM (Tivoli Storage Manager) hosts in a single system image implemented across multiple processor complexes. If we tried to to do that with a Windows host we'd wind up with a bigger physical server than we could have implemented putting each one in its own hardware. This was the problem we had with Citrix Metaframe. In order to get 30 client 'hosts' we had to build the biggest goddamn Windows server available at the time. It just turned out to be flat out cheaper to do it with a heavy client.
If someone can show me how to run Windows on 24-48 processors for a single system image that I can then carve up into 100 VMs then I'll be impressed.
Well then that's the problem. If you can be a good executive assistant for that pay or a grunt level legal editor then there is no real upside to being a coder. And may I add that many many many people don't actually want to live in West Gopher. The problem with places like Sioux Falls, SD or Graham, NC is that there's nothing there.
My bad, sorry.
Im sure Windows is fine for troop ship management, slow rides up the Thames past Buckingham and the usual lot of do nothing feather in the hat bilge the UK Navy is up to.
That's the problem. There is a shortage of people willing to work 80 hrs a week for $60K with a relocation to West Gopher Hole, South Dakota.
Blah blah blah not enough high schools teach Microsoft coding skills. Blah blah blah not enough Indians coming to America to debug Redmond's code. Blah blah blah we need more wage slaves.
Assuming zero or near zero growth. Over time when capacity hits the ceiling then you'll be forced to deploy new hardware and because you've already hit some arbitrary ceiling the new hardware will be very large and very powerful. Hopefully you won't have to deploy the app on its own hardware or figure out how to move images around from server to server. VM environments can be quite a bit more complex from a capacity planning and load balancing perspective unless each of the apps is really small and trivial. Moreover it's very hard to isolate specific instances for things like VPN tunnels, health checking, port scanning and NIC utilization. It more or less commits you to running different little apps for the same customer on the same box. Now that can work but from the point of view of running a hosting center for 1500 different customers I'm sure there's a lot of empirical evidence to suggest what the real quantitative material benefits are.
No but if I need to actually virtualize something I'll run it on a system that actually runs virtualization well with low overhead that allows me to dial in the specific capacity I need. Just because I can partition something arbitrarily doesn't mean it's running in the right sized sandbox. And if your apps are really trivial enough to do that then why are they being hosted at all.
Port them to AIX? No of course not. That would mean they're big enough and important enough to do that, which clearly they're not.
BTW wait 18 months when you run out of overhead unevenly and you're forced to move specific applications to their own hardware and/or you're forced to create new virtualized servers on the new hardware. You will find that either the hardware you need to deploy is either insanely large or, those applications will wind up on their own servers anyhow.
You can't do virtualized port scanning, health checking. You can't easily manage a virtualized VPN tunnel and you're stuck with more or less keeping all the same customers on the same iron anyway. Plus patch management really isn't much easier and capacity planning is much much harder to do.
With the operating system you have, not the one you wish you had.
-Donald Balmer Rumsfeld Gates III
Aside from developers and a tiny group of specialists who need access to a particular app? In the datacenter world this is anathema. No one running a gaggle of boxes would ever seriously consider this and get paid for it. Cheaper and easier by far to throw up one more server and spend the 0.04 FTE (1/25th of a person) it takes to run it.
And if you seriously considering multi image same system partitioning of Windows then you my friend need to re examine what it is you're doing. LPARs are not for Windows code. Go out and by an iSeries midrange or an AIX machine.
Was $99 for installation on 3 machines. In the 4 years I've had it I've seen no need to replace it or upgrade it. So that's $33 per machine /48 months = 69 cents/month. If MS thinks I'm going to pay a 2180% premium they have been smoking too much crack. Well to me it's just another nudge off any and all Microsoft code forever.
Sayonara Redmond Dudes.
If you is estupido you winds up in Iranq.
Never mind that the Arabic/Persian world completely ignored the Renaissance, the Enlightenment and the Industrial Revolution. I'm sure and my Marxist teachers in college swear on a stack of Howard Zinn's that it's some white imperialist capitalist Zionist nation's fault. It always is.
In North Carolina if a minor screams something out which is offensive, inflammatory or threatening to more than one person it's considered terrorism or making a terroristic threat. For 16 yo and above it can be charged as a felony and while it's not often charged as one, it's not unheard of. So one can extrapolate and guess that making threats online is also terrorism or making a terroristic threat. Free speech be damned we're talking PATRIOT act here. And what better group of people to apply it to than people who can't vote but who are charged as adults in the first place?
NASA is talking about going to the moon in order to open up funding streams for all of the precursor projects which will then be used for some other purpose, ostensibly a revenue generating one. Sorry to crush your moon rocks but we are not going to back to the moon this century. If anyone gets there before the year 2100 it will be India or China, if only to say YAAAAY FOR US !!!!
But manned spaceflight out of the orbit of the earth is in fact dead and over, forever, or if not, for the next 100+ years. No one wants to do it. Governments don't want to pay for it and we don't have the attention span for it. In fact I'd put money on the complete elimination of manned spaceflight by the year 2015. We will have officially spent enough money to piss everyone off by that time.
Phase 2 is the complete elimination of all space science, in space, by 2020. Expect that orbital telescopes, research satellites etc will all be defunded by then. Unless there is a commercial or military purpose for space science, it will be killed.
It was a pretty good run but now it's over. I grew up pouring over every detail, every photograph, every newscast, every lay science paper for everything associated with the Gemini and Apollo programs. I had a family member who worked in both programs. It was magic.
But by the time Skylab was discontinued it was clear that NASA was looking to get into the commercial heavy lifting business, ergo Space Shuttle. But NASA didn't bank on the expense and complexity of Space Shuttle, nor did they anticipate smaller payloads becoming the norm. So the Air Force became NASA's only paying customer. They're the only people who have a need for the capacity of Space Shuttle. So NASA is just treading water until Space Shuttle and ISS are killed off. They hope to have another heavy lifter online by then but if they don't then that's that. End of story. We'll be able to go the ESA or India, Japan, Russia or China for launch capability by then and NASA will have ceased to have a purpose.
Politicians attack those things that they can have zero actual effect upon per maximum soccermommy goodness.
Next up, a bill to regulate the TIDES.
The key, the only key to successfully implementing security in Vista or any other MS codebase is not to work from the assumption that everything can be locked down 100% and nothing bad can ever be made to happen. That's just stupid. Feel free to write an airliner fly by wire system and charge consumers a million dollars for each copy.
No, the problem with Vista and XP and.....is that they think they can both build an elegant system which simultaneously checks everything all the time and prevents an unknown thing from occuring.
The approach should be 180 different from that. It should be to assume that problems will occur and simply mitigate the damage or the extent of the damage they can cause. Build it such that even if it's botnet'd that the outbound traffic is blocked and the damage is limited to that one machine. Build it so that buffer overflows only go as far as that one application or subsystem.
Sandbox sandbox sandbox and when you're done, virtualize it. I really thought that when Intel announced the dual core processors we'd finally see some progress. We'd see one of the two cores devoted to all of the security and protection functions from port scanning, to encryption, to firewalling (in both directions) and so on. But instead we got the dancing bears 3D lucite animated we spent 10 million dollars developing the SOUND that the taskbar makes interface.
What a colossal waste of time and effort. Most of the problems associate with Windows security are DIRECTLY traceable to the fact that none of any of the original problems were ever addressed. They were embraced and layered over with yet more code. Sometimes the code is a workaround, sometimes it's nothing more sophisticated than an alert.
"Do you want to execute this program?"
I don't know. Is it bad? Why don't you tell me? Why don't you give it a whirl in a contained environment, let me know if it's bad, and if not write a sig to the system that lets it know the next time I want to run it it checks that sig for verification purposes.