With more than two decades of serious computing behind me, I still do not understand what "Administrative privileges" really means in Windows.
If you understand multiuser security, you understand Windows security. It's basically the same as the Unix model, with a few twists:
+ Administrator is not quite as all-powerful as root -- still bound by ACLs for example + ACL permissions apply to not just files but also registry keys + There's a policy layer to control who can perform certain actions (setting the clock, installing device drivers, etc).
The entire computer-as-an-appliance model of how a computer should behave in Windows just doesn't lend itself to the notion of a "privileged account". You don't have a privileged account in your toaster or your microwave, do you?
I strongy agree with this sentiment. Multiuser security wasn't designed for personal computing and really only works on the desktop as a kludge. PC security is never really going to work until we have a system that acknowledges that the "user" is not a trust level. Instead, it should be task-oriented. For example, installing software is a high-trust activity, while (say) running Kazaa should be severely locked down.
the malware will become "more clever" and thus that my machine will be less secure than it is now.
The malware is already pretty damn clever in a Windows system programming sense. It's probably unavoidable regardless of the system put in place.
There is a fairly large corporate installedbase of Windows 2000 users simply because XP was such a non-upgrade. These folks have been waiting so long for Vista that it would be unfair for MS to start cutting support before it's even shipped. Not to mention that it would take them at least a year to plan and execute a deployment.
And FWIW, MS has promised a 10 year support window for their OSes.
If Apple wanted to sell an generic x86 computer, they would have 10 years ago.
Obviously that argument doesn't fly. Apple is not a constant -- they change their strategy to fit the fundementals of the computer industry. There could well be a point in the future when the lines cross and software has a better profit potential than hardware.
The fulltext searching is actually one of the few bright spots in Notes. Of course, it's disabled by default and you need to wade through the mysterious UI to find the magic buttons.
In the mid-90s I worked at a place that was ditching their PS/2s and their "IBM Mentality". So I got to see all sorts of great MCA and other IBM stuff piled around. There was a mysterious 8595 (I think) sitting in the corner doing something critical and only running because there was a paperclip jammed into the powerswitch.
Instead of a real network out in the cubefarm, they had used Apple's LocalTalk standard and IBM's 3270 "twinax" stuff. $500 MCA LocalTalk cards! Plus the closets were full of MCA sound/video/younameit cards, every sort of PS/2 snapin or addon you could imagine, and tons of copies of OS/2 1.3 EE still in shirnkwrap. Too bad I didn't have the collector bug then, it would be fun to have some of this junk now.
Anyway, it was interesting seeing what was left of a "True Blue" shop that had bought everything IBM sold. With customers like this place, you could see why IBM thought they could get away with MCA. And when even the True Blue shops had given up on IBM and started buying Compaq Servers and Windows NT, it was obvious why IBM was in trouble.
Well, one of the interesting things about MCA was that IBM sat on it until they got out their consent decree. But I'm sure you're right and the were still being watched closely by the government.
This was also right during the period when IBM was holding back on higher-end PCs in order to protect their midrange systems. If they could have contractually limited the clones somehow, they probably would have.
The second 'fear' is just nonsense as everyone knows who ever looked at the interface it offers to an OS.
I wasn't implying there was any technical connection between MCA and OS/2, the fear among clonemakers was that IBM would "bundle" them somehow.
While cost/license issues may have played a role, the biggest problem with MCA was control -- IBM could have used the MCA license to basically dictate what sort of machines the PC industry could build, and even what operating system to put on them. Or that was the fear anyway.
I love how I complain about bullshitting IT guys, and I get a bunch of bullshitting in reply.:)
Maybe a more realworld example is in order:
Me: Sometimes it takes about 10 minutes to logon to my machine. Could there be a ActiveDirectory issue? Good IT answer 1: Hmm, I'll look at the Domain Controller tonight Good IT answer 2: Let me check Technet and do some research Good IT answer 3: I'll call the AD expert Actual IT answer: It's probably because you installed Yahoo (runs off)
See the problem here? And even if their explaination had some basis in reality, there was no attempt to fix it.
Now, I'm a technical person. You're a technical person. If you can't level with me, you are a lost cause for J. Random Luser. Nobody is convinced by this stuff.
There's the problem - the inexperienced guy or someone you consider of a much lower skill level is the one you have to go to for advice which they cannot supply
I'm not going them for advice. I'm trying to get them to do their jobs. Getting the right IT person involved isn't my issue.
Most IT people I know get treated like crap, and they don't deserve that.
Most IT people are full of shit and do deserve it.
Seriously, when confronted with some problem that they don't know the solution for, the first instinct of the vast majority of "Computer Guys" is to start lying their asses off by spewing psuedo-technical bullshit. They do this even when they know you know they're full of it
Typical conversation with IT:
Me: Problem IT: Must be the poliarity on the flux capacitors, don't know what we can do about that... Me: I've been in IT my entire career, there's no flux capacitors IT: Yup, those flux capacitors, always acting up Me: I used to do your job a long time ago IT: Well, it could also be your registry. Heh, Windows. Me: So... can you look into it? IT: Well I'm not going to have any time today (goes back to napping).
So, treat people with respect, stop acting like the uber-computer-god that you are not, and maybe people will reciprocate.
I'm thinking that most OS searches are caused by technical support issues or problem solving, so it's not necessarily a good thing to have a high score.
Given that, Linux very high relative to the size of it's installed base. Which makese sense because idiots like me can't get anything done in Linux without 1000 google searches.
If someone's going to screw us, we'd rather it be "regular" people on ebay rather than those sneaky bastards at the store.
I agree. In the big picture, all that Ebay activity will be a drop in the bucket in terms of XBox360's installed base. It wasn't enough to make a pricing decision on.
Now that Sony has gone with a high-price strategy, they're somewhat stuck. If they keep the price high, they're going to have to live with low marketshare. If they lower the price to $299 for XMas 2007, people are going to feel ripped-off and they'll face a Sega Saturn Style backlash.
The only way out them really is to push Blu-Ray hard (like say a discounted Sony PS3/HDTV combo.)
The first thing to understand is that there would be no Blu-Ray/HD-DVD war if it wasn't for the Playstation 3. There's very little technical differences between the formats, it's purely a political thing.
Every company in the Blu-Ray camp is expecting that PS3 is going to push the format to victory. That's the only reason they're involved in this silly format war to begin with.
They way the normal settop developer play in is by making higher-end Blu-Ray recorders that have DVR/Time-shift abilities. It's expected that PS3 will dominate the low-end player market.
That having been said, $500 isn't a bad price if you actually want HiDef movie playback. Only problem is that a lot more people want Video Games than HiDef Movies.
I for one actually really like the AMD approach. I periodically check the benchmarks to verify that their numbers correlate at least somewhat with real application-level performance.
Actually, there's been many cases where AMD has bumped their model number due to some internal change without any appreciable gain in performance.
AMD model numbers are based on an arbietariy formulas that change from time-to-time, and not any sort of realworld benchmark performance. Then you have the Turion chips which use a *different* scheme than the Athlons.
Actually I don't see what Intel is doing as being that different from cars. You have "Core 1", "Core 2" for model year/generation, and after that, anything goes. It's not like most car models/sub-models can be decoded by name alone either.
Yes, I think you're right on. It used to be that CPUs followed a simple progression, faster ones came out and the slower ones got cheaper. Then there was a couple minor varations (SX chips). Then the whole Celeron thing started and new chips started appearing on the bottom of the line up. Now it's more of a spread than a progression.
I have an old PC -- Compaq Deskpro XL 5133. Just by the name you can tell exactly what's in it. You couldn't do the same thing with a 3Ghz machine -- Intel has built dozens of different CPUs that run at that speed.
For an extreme example of this, check out the graphic card market. ATI has so many different model numbers out there that it's impossible to understand how their heirarchy works. Frequently the faster boards will have lower numbers attached to them.
Many companies are doing similar -- the goal is to emphasize the company brand name over the individual product names.
For example, Cadillac replaced the Seville and Deville with anonymous letters like STS and DTS. This puts more brand id on "Cadillac" part. And Apple is moving to a generic Mac* naming scheme to emphisize the "Apple" and "Mac" parts over the individual model names.
Intel had the problem that "Pentium" had such high brand recognition that it was difficult to move away from it, and after a while having products like "Pentium D" got very silly & confusing. They could create a new product brand like "Stupendium", but then they're starting at zero and they would just create the same problem again in the future. Instead they put that money behind "Intel" by picking a rather generic product names.
OK -- Apple jumped on Intel's back just like all the old RISC companies did (DEC, SGI, HP). Furthermore, they're trying to accomodate Windows somehow just like DEC, SGI, HP did. I don't see how Apple is a real counter-example here -- they're following the same patterns.
The old Apple systems had been soundly beat on Photoshop/$ for years, and their pro market has become less and less relevant to Apple as the years passed.
However, in the consumer market, the extreme cheapness of hardware has actually helped Apple because they're able to sell high margin PCs as a luxury good.
Quite frankly, you are blaming the symptom and not the disease. The market for UNIX Workstations had pretty much dried up to nothing, and SGI was already going in the crappper. Going with Wintel was a last-ditch attempt to stay in the market.
Look at all the success that Nvidia (former SGI people) has had with Wintel graphics. SGI's problem was just poor execution.
This is an interesting example, because MS Word has a feature like Page Styles called Sections. I've always been in the habit of creating a new section for cover pages and the like, and was unaware of this 'first page' feature. MS must have figured out through user testing that a new section was too much of a conceptual leap for a cover page and added the feature somewhere in there.
I wouldn't assume that the factory uses that much thermal paste just because the repair manual shows it that way. They're assembling millions of Core Duo laptops, presumably they have some idea what they're doing.
However, you might want to avoid a referb/repair MacBook just in case there's an Apple tech out there without a clue.
Slashdot Mirror
With more than two decades of serious computing behind me, I still do not understand what "Administrative privileges" really means in Windows.
If you understand multiuser security, you understand Windows security. It's basically the same as the Unix model, with a few twists:
+ Administrator is not quite as all-powerful as root -- still bound by ACLs for example
+ ACL permissions apply to not just files but also registry keys
+ There's a policy layer to control who can perform certain actions (setting the clock, installing device drivers, etc).
The entire computer-as-an-appliance model of how a computer should behave in Windows just doesn't lend itself to the notion of a "privileged account". You don't have a privileged account in your toaster or your microwave, do you?
I strongy agree with this sentiment. Multiuser security wasn't designed for personal computing and really only works on the desktop as a kludge. PC security is never really going to work until we have a system that acknowledges that the "user" is not a trust level. Instead, it should be task-oriented. For example, installing software is a high-trust activity, while (say) running Kazaa should be severely locked down.
the malware will become "more clever" and thus that my machine will be less secure than it is now.
The malware is already pretty damn clever in a Windows system programming sense. It's probably unavoidable regardless of the system put in place.
There is a fairly large corporate installedbase of Windows 2000 users simply because XP was such a non-upgrade. These folks have been waiting so long for Vista that it would be unfair for MS to start cutting support before it's even shipped. Not to mention that it would take them at least a year to plan and execute a deployment.
And FWIW, MS has promised a 10 year support window for their OSes.
If Apple wanted to sell an generic x86 computer, they would have 10 years ago.
Obviously that argument doesn't fly. Apple is not a constant -- they change their strategy to fit the fundementals of the computer industry. There could well be a point in the future when the lines cross and software has a better profit potential than hardware.
Searching sucks. I mean, just sucks.
The fulltext searching is actually one of the few bright spots in Notes. Of course, it's disabled by default and you need to wade through the mysterious UI to find the magic buttons.
Otherwise, in complete agreement.
In the mid-90s I worked at a place that was ditching their PS/2s and their "IBM Mentality". So I got to see all sorts of great MCA and other IBM stuff piled around. There was a mysterious 8595 (I think) sitting in the corner doing something critical and only running because there was a paperclip jammed into the powerswitch.
Instead of a real network out in the cubefarm, they had used Apple's LocalTalk standard and IBM's 3270 "twinax" stuff. $500 MCA LocalTalk cards! Plus the closets were full of MCA sound/video/younameit cards, every sort of PS/2 snapin or addon you could imagine, and tons of copies of OS/2 1.3 EE still in shirnkwrap. Too bad I didn't have the collector bug then, it would be fun to have some of this junk now.
Anyway, it was interesting seeing what was left of a "True Blue" shop that had bought everything IBM sold. With customers like this place, you could see why IBM thought they could get away with MCA. And when even the True Blue shops had given up on IBM and started buying Compaq Servers and Windows NT, it was obvious why IBM was in trouble.
Well, one of the interesting things about MCA was that IBM sat on it until they got out their consent decree. But I'm sure you're right and the were still being watched closely by the government.
This was also right during the period when IBM was holding back on higher-end PCs in order to protect their midrange systems. If they could have contractually limited the clones somehow, they probably would have.
The second 'fear' is just nonsense as everyone knows who ever looked at the interface it offers to an OS.
I wasn't implying there was any technical connection between MCA and OS/2, the fear among clonemakers was that IBM would "bundle" them somehow.
While cost/license issues may have played a role, the biggest problem with MCA was control -- IBM could have used the MCA license to basically dictate what sort of machines the PC industry could build, and even what operating system to put on them. Or that was the fear anyway.
I love how I complain about bullshitting IT guys, and I get a bunch of bullshitting in reply. :)
Maybe a more realworld example is in order:
Me: Sometimes it takes about 10 minutes to logon to my machine. Could there be a ActiveDirectory issue?
Good IT answer 1: Hmm, I'll look at the Domain Controller tonight
Good IT answer 2: Let me check Technet and do some research
Good IT answer 3: I'll call the AD expert
Actual IT answer: It's probably because you installed Yahoo (runs off)
See the problem here? And even if their explaination had some basis in reality, there was no attempt to fix it.
Now, I'm a technical person. You're a technical person. If you can't level with me, you are a lost cause for J. Random Luser. Nobody is convinced by this stuff.
There's the problem - the inexperienced guy or someone you consider of a much lower skill level is the one you have to go to for advice which they cannot supply
I'm not going them for advice. I'm trying to get them to do their jobs. Getting the right IT person involved isn't my issue.
Most IT people I know get treated like crap, and they don't deserve that.
... can you look into it?
Most IT people are full of shit and do deserve it.
Seriously, when confronted with some problem that they don't know the solution for, the first instinct of the vast majority of "Computer Guys" is to start lying their asses off by spewing psuedo-technical bullshit. They do this even when they know you know they're full of it
Typical conversation with IT:
Me: Problem
IT: Must be the poliarity on the flux capacitors, don't know what we can do about that...
Me: I've been in IT my entire career, there's no flux capacitors
IT: Yup, those flux capacitors, always acting up
Me: I used to do your job a long time ago
IT: Well, it could also be your registry. Heh, Windows.
Me: So
IT: Well I'm not going to have any time today (goes back to napping).
So, treat people with respect, stop acting like the uber-computer-god that you are not, and maybe people will reciprocate.
I'm thinking that most OS searches are caused by technical support issues or problem solving, so it's not necessarily a good thing to have a high score.
Given that, Linux very high relative to the size of it's installed base. Which makese sense because idiots like me can't get anything done in Linux without 1000 google searches.
If someone's going to screw us, we'd rather it be "regular" people on ebay rather than those sneaky bastards at the store.
I agree. In the big picture, all that Ebay activity will be a drop in the bucket in terms of XBox360's installed base. It wasn't enough to make a pricing decision on.
Now that Sony has gone with a high-price strategy, they're somewhat stuck. If they keep the price high, they're going to have to live with low marketshare. If they lower the price to $299 for XMas 2007, people are going to feel ripped-off and they'll face a Sega Saturn Style backlash.
The only way out them really is to push Blu-Ray hard (like say a discounted Sony PS3/HDTV combo.)
The first thing to understand is that there would be no Blu-Ray/HD-DVD war if it wasn't for the Playstation 3. There's very little technical differences between the formats, it's purely a political thing.
Every company in the Blu-Ray camp is expecting that PS3 is going to push the format to victory. That's the only reason they're involved in this silly format war to begin with.
They way the normal settop developer play in is by making higher-end Blu-Ray recorders that have DVR/Time-shift abilities. It's expected that PS3 will dominate the low-end player market.
That having been said, $500 isn't a bad price if you actually want HiDef movie playback. Only problem is that a lot more people want Video Games than HiDef Movies.
Not for an SGI customer, no, not at all.
The inferno/flame is now an IBM (Lenovo?) PC
FWIW, IBM still makes their PC workstations -- that part was not sold to Levano.
I for one actually really like the AMD approach. I periodically check the benchmarks to verify that their numbers correlate at least somewhat with real application-level performance.
Actually, there's been many cases where AMD has bumped their model number due to some internal change without any appreciable gain in performance.
AMD model numbers are based on an arbietariy formulas that change from time-to-time, and not any sort of realworld benchmark performance. Then you have the Turion chips which use a *different* scheme than the Athlons.
Actually I don't see what Intel is doing as being that different from cars. You have "Core 1", "Core 2" for model year/generation, and after that, anything goes. It's not like most car models/sub-models can be decoded by name alone either.
Yes, I think you're right on. It used to be that CPUs followed a simple progression, faster ones came out and the slower ones got cheaper. Then there was a couple minor varations (SX chips). Then the whole Celeron thing started and new chips started appearing on the bottom of the line up. Now it's more of a spread than a progression.
I have an old PC -- Compaq Deskpro XL 5133. Just by the name you can tell exactly what's in it. You couldn't do the same thing with a 3Ghz machine -- Intel has built dozens of different CPUs that run at that speed.
For an extreme example of this, check out the graphic card market. ATI has so many different model numbers out there that it's impossible to understand how their heirarchy works. Frequently the faster boards will have lower numbers attached to them.
Yes, SGI failed because their monitors required a $20 adapter cable to work with a PC. Thanks for your input.
Many companies are doing similar -- the goal is to emphasize the company brand name over the individual product names.
For example, Cadillac replaced the Seville and Deville with anonymous letters like STS and DTS. This puts more brand id on "Cadillac" part. And Apple is moving to a generic Mac* naming scheme to emphisize the "Apple" and "Mac" parts over the individual model names.
Intel had the problem that "Pentium" had such high brand recognition that it was difficult to move away from it, and after a while having products like "Pentium D" got very silly & confusing. They could create a new product brand like "Stupendium", but then they're starting at zero and they would just create the same problem again in the future. Instead they put that money behind "Intel" by picking a rather generic product names.
I think (and thought at the time) they should have focused on a cheaper version of their products and tried to be an Apple alternative.
Actually they did this, showing off an Indy running Photoshop at MacWorld one year. I don't know how serious they were about it.
OK -- Apple jumped on Intel's back just like all the old RISC companies did (DEC, SGI, HP). Furthermore, they're trying to accomodate Windows somehow just like DEC, SGI, HP did. I don't see how Apple is a real counter-example here -- they're following the same patterns.
The old Apple systems had been soundly beat on Photoshop/$ for years, and their pro market has become less and less relevant to Apple as the years passed.
However, in the consumer market, the extreme cheapness of hardware has actually helped Apple because they're able to sell high margin PCs as a luxury good.
Quite frankly, you are blaming the symptom and not the disease. The market for UNIX Workstations had pretty much dried up to nothing, and SGI was already going in the crappper. Going with Wintel was a last-ditch attempt to stay in the market.
Look at all the success that Nvidia (former SGI people) has had with Wintel graphics. SGI's problem was just poor execution.
This is an interesting example, because MS Word has a feature like Page Styles called Sections. I've always been in the habit of creating a new section for cover pages and the like, and was unaware of this 'first page' feature. MS must have figured out through user testing that a new section was too much of a conceptual leap for a cover page and added the feature somewhere in there.
I wouldn't assume that the factory uses that much thermal paste just because the repair manual shows it that way. They're assembling millions of Core Duo laptops, presumably they have some idea what they're doing.
However, you might want to avoid a referb/repair MacBook just in case there's an Apple tech out there without a clue.
Nowhere in copyright law does it state that a copy has to cause harm to be infringing.
This is untrue. Economic and other damages are taken into consideration when determining Fair Use.
(Standard IANAL disclaimer goes here.)