I got into and argument with Linus, et. al. many, many years ago about pulling most of the drivers (video especially) out of the kernel and creating a stable ABI. This was about GGI then, but I made the suggestion that as things progress, building drivers directly into the kernel would become unmanageable and that the kernel would always be behind the curve. He kept arguing that it would sacrifice too much performance and it would be very difficult to build a usable HAL (he was right on both of these points, BTW). I countered that it was a necessary step. I was basically called a heretic.
The model I suggested is almost exactly the same that Apple chose for OS X and it seems to work fine. Maybe it's time for Linux to follow suit. But then, what the hell do I know?
Firefox is free, open source, and runs on anything. When OS X is all of those, then Microsoft might have something to worry about at least on the desktop.
I guess Citrix should just pack up it's multi-billion dollar business and go home because Microsoft (who co-developed the technology) doesn't want them.
Ok, FUD boy, NT has been multi-user for a long, long time. The fact that they want end users to PAY for licenses to be able to simultaneously run several users at once is a business decision, not a technical limitation. Why in the world would you need full desktop access for several people for a HOME computer?
There are no such limitations on remote management, remote configuration, CLI remoting, file sharing, the remote help system, printer sharing, etc. So, again, why in the hell would you want multiple users on single machine at the same time with full desktop access in a home environment on a home PC?
I am really sick of the naivete of people who have never really used, devoloped for, or managed Windows in a business environment (I am talking 1000's of PC's...not some Mom-n-Pop) harping on the stupidest of FUD.
There are lot's of great reasons to go with open source applications, OS's, and infrastructure but in the end it is just a tool to accomplish a goal, just like Windows.
Ignoring the fact that the US dollar was and is the buoy for all currencies worldwide and we have publicly stated for decades that we are committed to a strong dollar policy so as to not crash others economoies (even in the face of gross undervaluation such as China's yuan), the dollar has dropped in real valuation by more than 50% and is considered pretty much dead-on now. That "the dollar is over-valued" stuff went out about two years ago.
Scale and distribution will nail you every time. Even if used to an extreme amount at the same time, all the alternative sources still can't make up for current consumption let alone future. The only reasonable alternatives found so far that work today are nuclear and geothermal/core tap. The difficult parts are one: getting widespread geothermal to work without doing something really bad, and two: how to store and distribute energy better. Distribution and storage is where most of the waste is right now. Heck if we could figure out how to transmit electricity without loosing 80% or more of it to heat, we would add a couple of hundred years to our reserves.
FB-DIMM is really great if you want to make an extremely stable server and don't care as much about latency or just happen to not already have a memory controller on die.
Oh, and yeah, Intel owns the patents so you would be paying your competitor and be beholden to them to chose your memory type.
And all the memory manufacturers would have to pony up to Intel and also be beholden to them.
I won't go into the technical details about how AMD's design is geared towards low latencies and suffers greatly when latencies get too high while Intel's (old Netburst) design was geared towards high speeds and could tolerate terrible latencies but suffered greatly when speeds choked it. We probably won't see FB-DIMM as Intel is shifting over to the Centrino family of CPU's that are closer to AMD's design idea.
Actually, chips for Apple accounted for less than 2% of the capacity of just one IBM fab. IBM's tech division (which does chip fabbing) accounted for less than 3% of IBM's total revenue. That's a really small piece of IBM's global business. It's kind of like an oil company losing one gas station...not really gonna hurt them that much.
The RF engineering crowd can enlighten us, but wasn't the idea of plasma antennas kinda in the same vein? The idea being that the device could modify itself on the fly to have wireless ubiquity.
Not even the same code lines
on
Buy Vista or Else
·
· Score: 2, Insightful
You do realize that the Win95, etc. core and the NT cores aren't even from the same code lines?
It uses Windows Media Player and all the DRM'd code behind it as the player. It follows almost the exact model of Window Media Center Edition with additional abilities and personal customization. Like I said, check it out. I am also a paying customer of Meedio and if I could get my brand spanking new RF remote to work with it properly, I would strongly recommend it, too.
It's integrated into AD and local permissions meaning that an enterprise admin could revoke or grant priviledges on the fly which is a really big deal in an infrastructure environment. Also, it is a key cert style encryption so you can always back up the decrypt key or use a central key server. I have never tried to use it to hide a volume for plausible deniability, but I don't think it was designed with that in mind.
That video is a well debunked fake. It was meant as a joke (the easy to spot earplugs in the video and the still-attached core should be a hint...if the 3 GHz on a Duron without water cooling isn't).
It depends. Cable companies recently won a Supreme Court case to retain their classification as an "information service" which exempts them from some of the regulations of a common carrier (chiefly, that they would have to give access to their lines to competitors). This also means they can be held liable for what their users do legally. Can...depends on who's lawyer is doing the talking. It's also the reason why cable modem companies are a heck of a lot more strict than telco's.
That brings me to DSL companies. They ARE classified as a "telecommunication service" and ergo, a common carrier. This also goes for T1's, etc. and other telco provided services.
Now if the ISP does not handle the physical line provisioning, etc. then they are not a common carrier but may try this to keep their legal shield (common carriers are not legally liable for what their clients transport over it's service with a few caveats). If the ISP is involved with the telco end (and there are cases on record from way back dealing with modems that go both ways on this), then it could be considered a common carrier (but not in all instances).
The big deal here is competeing companies could use this law to show that cable companies are common carriers and try to force them to open up their lines (like CLEC's in the telco biz).
I can't believe no one brought this up. As soon as they show an ability to shape and control all types of traffic and actually make it their business to do so, they lose common carrier status and can be sued for anything and everything. I can't even imagine what damage this would do. The carriers are either insane or greedy. I vote the for the latter.
I would think she would want to decide what type of user you are testing for. The needs of a large or medium sized business user would deviate greatly from those of a home user.
I worked at a lawfirm that used Blackberries for everything. You had instant, two-way access to everything that was in Outlook: calendering, contacts, email, attachments (with editing capabilities), the document management system, research and conflict of interest systems, real-time transcription systems, etc. There are so many add on technologies for Blackberry it's not even funny.
OBTW, most of the representitive part of federal government (The House and Senate, mostly) run off these things.
I agree, it isn't worth it if you don't have it plugged into Exchange or Domino.
Copyright helps protect the owner from several bad things including: keeping someone else from modifying your work in a way you do not want it to be and then attributing it to you (What if someone changed Schindlers List to be favourable to the Nazi's and then stamped Speilbergs name on it? It's copyright law, among others, that protects against this. In most of Europe, the original copyright owner cannot give up his right of "creative control" although in the US you can sell that right and it is usually demanded), knockoffs and forgeries (I am all for sampling, etc. and so are some copyright holders, but full fledged forgeries are out and out stealing. It takes a lot of work and money to make a brand or name, etc. and when someone co-ops that for financial gain, it is theft of real value.)
If you want to go after someone about how screwed up copyright laws are (especially in the US but it is having a viral effect accross the pond), then go after Congressmen and the lawyers egging them on. I like the idea put forth by a Judge in Canada (sorry, can't remember my source to cite) where he proposed limiting the length applied (it used to be 28 years max here) and change copyright to fall under tort rules. Meaning that you could never really criminalize it. Tort law litigation would mean that a plaintiff would actually have to prove they were legally "wronged" and further prove real damages. The only results would either be an order to stop the injurous activity and/or monetary damages. None of this "you will go to jail if you copy that CD" BS. I think that makes a whole lot more sense than the pseudo-criminalization we have now.
""There are three huge problems with linux in business as a workstation and as an infrastructure base. One, lack of a coherent and consistant interface throughout all apps and the OS."
Ever heard of gnome/gtk+ or kde/qt? Crack kills dude."
>> Neither of which enforces a single standard human interface accross every app.
""Two, most drivers need to get the hell out of the kernel... There is no longer any compelling technical reason to keep all the drivers as kernel modules anymore."
Wrong. Modules are not "in the kernel", that's why they are called "modules" and are loaded from outside the kernel. Companies can write proprietary modules that will compile and run with any kernel at any time they like. They are not required to give up their intellectual property when creating driver modules. Ever heard of Nvidia? They make self-installing video driver modules available for downloading/distribution."
>> Now who's smoking crack? Read Linus' own words on instability of the ABI or vendors cursing every kernel change. Most drivers live directly in the kernel tree and most of the modules have to be rewritten every single minor revision. What I was proposing (and has been proposed many, many times) is pull all but the most essential drivers out of kernel space into user space. Yes, you give up some performance, but a modern computer has lots to give. Read up on OS X's model. It is exactly what Apples engineers came up with.
""Three, the big one for busines, a lack of enterprise thinking towards infrastructure. Yes, you can do single sign on (painfully). Yes, authentication/authorization is robust and do have some interoperability. And, yes, there is LDAP, X400, DAP, etc. available for a directory structure. The problem is that no one entity has brought it all together in a single package that works from the desktop to the router to NOC to the SAN to, well, you get the picture."
WTF? Most companies with 20 to 200 employees don't need anything more exotic than what is already offered in common distributions. Large companies with 200+ employees have enough resources to customize GNU/Linux to do whatever they need. Your problem description is very over rated."
>> Now that is just naive. It is patently obvious you have no clue what you are talking about and have never either been a part of nor led an infrastructure team. My organization is just 15,100 individuals and the last thing we would be burning cash on is trying to shoehorn a distro into our business needs on the desktop. We use BSD's and to a lesser extent, Linux's, in our appliance devices and we do maintane some custom distros for that reason, but nothing near the complexity of a desktop system or trying to string together all the pieces necessary to compete with a full Novell or Microsoft infrastructure.
Take the example of a meeting I just got out of. A multi-departmental shareholders meeting with over 12,000 people in 70+ countries all in different time zones. It took 72 hours to set it up so everyone could be available. It all coordinated automagically via Exchange calendering. This is not an uncommon occurance although 100 people is more the norm for project or department meetings. Until you have actually worked for a large (think 10,000 or more employees) organization that has a high need for communication (which would be about any company), you wont get how essential this type of communication is. Also, remember it's not just people you are scheduling, but satellite or circuit time, rooms, video teleconferencing equipment, laptops, etc.
That's why application layer firewalls are such a big deal. They are supposed to understand what data and more importantly what kind of data specific apps from both sides are supposed to be generating and filter at that level. They can be a royal pain to design and keep running. With custom apps, however, it's a whole different story.
One of the great concepts I learned at my last training was the idea of compartmentalization. This is after many other layers of defense, but the idea is even if you are compromised, they get nothing of real value.
There also many, many other ways to add protection: gambits, exclusion, audit-traps, active detection, etc.
Like the author, one of my big concerns is with poorly written apps. Anytime anything is specifically designed to release data but has restrictions on who gets what and how, you have a security concern. I have seen companies go to such extremes as pushing data to the screen as one big jpeg to keep from anything grabbing said data (of course, there are tools that can grab visual data from graphics just like humans now). My biggest concern is the original and still number one concern of all types of security: corruption from within. Be it sabotage, espionage, or unintentional damage (virus) the equipment and personnel inside an organisation is always more dangerous than those outside it.
And that's when manageing your information through compartmentalization and multilayer defense, etc. come in. You may get in systems, but you are not getting anything of real value and you aren't getting out again.
Slashdot Mirror
I got into and argument with Linus, et. al. many, many years ago about pulling most of the drivers (video especially) out of the kernel and creating a stable ABI. This was about GGI then, but I made the suggestion that as things progress, building drivers directly into the kernel would become unmanageable and that the kernel would always be behind the curve. He kept arguing that it would sacrifice too much performance and it would be very difficult to build a usable HAL (he was right on both of these points, BTW). I countered that it was a necessary step. I was basically called a heretic.
The model I suggested is almost exactly the same that Apple chose for OS X and it seems to work fine. Maybe it's time for Linux to follow suit. But then, what the hell do I know?
His 22 years were apparantly spent at E-Systems which was at the time a CIA shell company that ran Air America (among other freaky stuff).
Don't forget the Alienware Aurora m7700 with a dual core AMD X2 chip.
Firefox is free, open source, and runs on anything. When OS X is all of those, then Microsoft might have something to worry about at least on the desktop.
I guess Citrix should just pack up it's multi-billion dollar business and go home because Microsoft (who co-developed the technology) doesn't want them.
Ok, FUD boy, NT has been multi-user for a long, long time. The fact that they want end users to PAY for licenses to be able to simultaneously run several users at once is a business decision, not a technical limitation. Why in the world would you need full desktop access for several people for a HOME computer?
There are no such limitations on remote management, remote configuration, CLI remoting, file sharing, the remote help system, printer sharing, etc. So, again, why in the hell would you want multiple users on single machine at the same time with full desktop access in a home environment on a home PC?
I am really sick of the naivete of people who have never really used, devoloped for, or managed Windows in a business environment (I am talking 1000's of PC's...not some Mom-n-Pop) harping on the stupidest of FUD.
There are lot's of great reasons to go with open source applications, OS's, and infrastructure but in the end it is just a tool to accomplish a goal, just like Windows.
Zealotry serves no one.
Ignoring the fact that the US dollar was and is the buoy for all currencies worldwide and we have publicly stated for decades that we are committed to a strong dollar policy so as to not crash others economoies (even in the face of gross undervaluation such as China's yuan), the dollar has dropped in real valuation by more than 50% and is considered pretty much dead-on now. That "the dollar is over-valued" stuff went out about two years ago.
When you have heard a DTS:ES movie with all full range discrete channels going, then you'll know why people want 7 channels.
Now, for Dolby Digital then yeah, 5.1 is probably fine with it's matrixed and limited range surround channels.
Scale and distribution will nail you every time. Even if used to an extreme amount at the same time, all the alternative sources still can't make up for current consumption let alone future. The only reasonable alternatives found so far that work today are nuclear and geothermal/core tap. The difficult parts are one: getting widespread geothermal to work without doing something really bad, and two: how to store and distribute energy better. Distribution and storage is where most of the waste is right now. Heck if we could figure out how to transmit electricity without loosing 80% or more of it to heat, we would add a couple of hundred years to our reserves.
FB-DIMM is really great if you want to make an extremely stable server and don't care as much about latency or just happen to not already have a memory controller on die.
Oh, and yeah, Intel owns the patents so you would be paying your competitor and be beholden to them to chose your memory type.
And all the memory manufacturers would have to pony up to Intel and also be beholden to them.
I won't go into the technical details about how AMD's design is geared towards low latencies and suffers greatly when latencies get too high while Intel's (old Netburst) design was geared towards high speeds and could tolerate terrible latencies but suffered greatly when speeds choked it. We probably won't see FB-DIMM as Intel is shifting over to the Centrino family of CPU's that are closer to AMD's design idea.
Actually, chips for Apple accounted for less than 2% of the capacity of just one IBM fab. IBM's tech division (which does chip fabbing) accounted for less than 3% of IBM's total revenue. That's a really small piece of IBM's global business. It's kind of like an oil company losing one gas station...not really gonna hurt them that much.
The RF engineering crowd can enlighten us, but wasn't the idea of plasma antennas kinda in the same vein? The idea being that the device could modify itself on the fly to have wireless ubiquity.
You do realize that the Win95, etc. core and the NT cores aren't even from the same code lines?
Nice FUD, though...
It uses Windows Media Player and all the DRM'd code behind it as the player. It follows almost the exact model of Window Media Center Edition with additional abilities and personal customization. Like I said, check it out. I am also a paying customer of Meedio and if I could get my brand spanking new RF remote to work with it properly, I would strongly recommend it, too.
Build your own. MediaPortal is great and coming along fast. OpenSource MCE.
http://www.team-mediaportal.com/
It's integrated into AD and local permissions meaning that an enterprise admin could revoke or grant priviledges on the fly which is a really big deal in an infrastructure environment. Also, it is a key cert style encryption so you can always back up the decrypt key or use a central key server. I have never tried to use it to hide a volume for plausible deniability, but I don't think it was designed with that in mind.
That video is a well debunked fake. It was meant as a joke (the easy to spot earplugs in the video and the still-attached core should be a hint...if the 3 GHz on a Duron without water cooling isn't).
Windows has the Encrypted Filesystem built into NTFS.
p pro/deploy/cryptfs.mspx
http://www.microsoft.com/technet/prodtechnol/winx
It depends. Cable companies recently won a Supreme Court case to retain their classification as an "information service" which exempts them from some of the regulations of a common carrier (chiefly, that they would have to give access to their lines to competitors). This also means they can be held liable for what their users do legally. Can...depends on who's lawyer is doing the talking. It's also the reason why cable modem companies are a heck of a lot more strict than telco's.
That brings me to DSL companies. They ARE classified as a "telecommunication service" and ergo, a common carrier. This also goes for T1's, etc. and other telco provided services.
Now if the ISP does not handle the physical line provisioning, etc. then they are not a common carrier but may try this to keep their legal shield (common carriers are not legally liable for what their clients transport over it's service with a few caveats). If the ISP is involved with the telco end (and there are cases on record from way back dealing with modems that go both ways on this), then it could be considered a common carrier (but not in all instances).
The big deal here is competeing companies could use this law to show that cable companies are common carriers and try to force them to open up their lines (like CLEC's in the telco biz).
I can't believe no one brought this up. As soon as they show an ability to shape and control all types of traffic and actually make it their business to do so, they lose common carrier status and can be sued for anything and everything. I can't even imagine what damage this would do. The carriers are either insane or greedy. I vote the for the latter.
I would think she would want to decide what type of user you are testing for. The needs of a large or medium sized business user would deviate greatly from those of a home user.
I worked at a lawfirm that used Blackberries for everything. You had instant, two-way access to everything that was in Outlook: calendering, contacts, email, attachments (with editing capabilities), the document management system, research and conflict of interest systems, real-time transcription systems, etc. There are so many add on technologies for Blackberry it's not even funny.
OBTW, most of the representitive part of federal government (The House and Senate, mostly) run off these things.
I agree, it isn't worth it if you don't have it plugged into Exchange or Domino.
Copyright helps protect the owner from several bad things including: keeping someone else from modifying your work in a way you do not want it to be and then attributing it to you (What if someone changed Schindlers List to be favourable to the Nazi's and then stamped Speilbergs name on it? It's copyright law, among others, that protects against this. In most of Europe, the original copyright owner cannot give up his right of "creative control" although in the US you can sell that right and it is usually demanded), knockoffs and forgeries (I am all for sampling, etc. and so are some copyright holders, but full fledged forgeries are out and out stealing. It takes a lot of work and money to make a brand or name, etc. and when someone co-ops that for financial gain, it is theft of real value.)
If you want to go after someone about how screwed up copyright laws are (especially in the US but it is having a viral effect accross the pond), then go after Congressmen and the lawyers egging them on. I like the idea put forth by a Judge in Canada (sorry, can't remember my source to cite) where he proposed limiting the length applied (it used to be 28 years max here) and change copyright to fall under tort rules. Meaning that you could never really criminalize it. Tort law litigation would mean that a plaintiff would actually have to prove they were legally "wronged" and further prove real damages. The only results would either be an order to stop the injurous activity and/or monetary damages. None of this "you will go to jail if you copy that CD" BS. I think that makes a whole lot more sense than the pseudo-criminalization we have now.
""There are three huge problems with linux in business as a workstation and as an infrastructure base. One, lack of a coherent and consistant interface throughout all apps and the OS."
Ever heard of gnome/gtk+ or kde/qt? Crack kills dude."
>> Neither of which enforces a single standard human interface accross every app.
""Two, most drivers need to get the hell out of the kernel... There is no longer any compelling technical reason to keep all the drivers as kernel modules anymore."
Wrong. Modules are not "in the kernel", that's why they are called "modules" and are loaded from outside the kernel. Companies can write proprietary modules that will compile and run with any kernel at any time they like. They are not required to give up their intellectual property when creating driver modules. Ever heard of Nvidia? They make self-installing video driver modules available for downloading/distribution."
>> Now who's smoking crack? Read Linus' own words on instability of the ABI or vendors cursing every kernel change. Most drivers live directly in the kernel tree and most of the modules have to be rewritten every single minor revision. What I was proposing (and has been proposed many, many times) is pull all but the most essential drivers out of kernel space into user space. Yes, you give up some performance, but a modern computer has lots to give. Read up on OS X's model. It is exactly what Apples engineers came up with.
""Three, the big one for busines, a lack of enterprise thinking towards infrastructure. Yes, you can do single sign on (painfully). Yes, authentication/authorization is robust and do have some interoperability. And, yes, there is LDAP, X400, DAP, etc. available for a directory structure. The problem is that no one entity has brought it all together in a single package that works from the desktop to the router to NOC to the SAN to, well, you get the picture."
WTF? Most companies with 20 to 200 employees don't need anything more exotic than what is already offered in common distributions. Large companies with 200+ employees have enough resources to customize GNU/Linux to do whatever they need. Your problem description is very over rated."
>> Now that is just naive. It is patently obvious you have no clue what you are talking about and have never either been a part of nor led an infrastructure team. My organization is just 15,100 individuals and the last thing we would be burning cash on is trying to shoehorn a distro into our business needs on the desktop. We use BSD's and to a lesser extent, Linux's, in our appliance devices and we do maintane some custom distros for that reason, but nothing near the complexity of a desktop system or trying to string together all the pieces necessary to compete with a full Novell or Microsoft infrastructure.
Take the example of a meeting I just got out of. A multi-departmental shareholders meeting with over 12,000 people in 70+ countries all in different time zones. It took 72 hours to set it up so everyone could be available. It all coordinated automagically via Exchange calendering. This is not an uncommon occurance although 100 people is more the norm for project or department meetings. Until you have actually worked for a large (think 10,000 or more employees) organization that has a high need for communication (which would be about any company), you wont get how essential this type of communication is. Also, remember it's not just people you are scheduling, but satellite or circuit time, rooms, video teleconferencing equipment, laptops, etc.
That's why application layer firewalls are such a big deal. They are supposed to understand what data and more importantly what kind of data specific apps from both sides are supposed to be generating and filter at that level. They can be a royal pain to design and keep running. With custom apps, however, it's a whole different story.
One of the great concepts I learned at my last training was the idea of compartmentalization. This is after many other layers of defense, but the idea is even if you are compromised, they get nothing of real value.
There also many, many other ways to add protection: gambits, exclusion, audit-traps, active detection, etc.
Like the author, one of my big concerns is with poorly written apps. Anytime anything is specifically designed to release data but has restrictions on who gets what and how, you have a security concern. I have seen companies go to such extremes as pushing data to the screen as one big jpeg to keep from anything grabbing said data (of course, there are tools that can grab visual data from graphics just like humans now). My biggest concern is the original and still number one concern of all types of security: corruption from within. Be it sabotage, espionage, or unintentional damage (virus) the equipment and personnel inside an organisation is always more dangerous than those outside it.
And that's when manageing your information through compartmentalization and multilayer defense, etc. come in. You may get in systems, but you are not getting anything of real value and you aren't getting out again.