Couldn't they have picked better protocols? It seems to be me for reliability and performance that isn't the best of choices.
And what, specifically, do you think is wrong with TCP/IP? It's pretty minimal and simple.
Although TCP/IP is so commonplace I wouldn't want my 15 million dollar satellite to depend on it.
Even if TCP/IP had some technical drawbacks relative to some alternative protocol, software implementation errors and engineering mistakes are likely much bigger risks than some theoretical limitations of the protocol.
In different words, I'd much rather bet $15 million on a proven, debugged, mature TCP/IP implementation than on some implementation nobody has ever used for a protocol nobody has ever heard of.
I have two Macintoshes. They are nice consumer devices, look sleek, and work pretty well most of the time. I recommend them to friends and family who want a decent personal computer that's a little easier to use than Windows. And the fact that they come with partial support for UNIX APIs makes them easier to deal with than Windows.
I do appreciate Apple trying to cooperate with the free software community to some extent. But, frankly, I don't expect too much from it. I think there is really not all that much Apple has to offer, and most of the things that Apple has opened up were really forced by licenses or business realities (e.g., gcc was forced by the license, Rendezvous and QuickTime were forced by business considerations).
Overall, I think it just doesn't matter much what Apple does, either technically or in terms of their business. I don't see Apple winning in a big way: they can't take away much market share from Windows, and, based on my experience with them, I think they are not a major competitor for UNIX or Linux workstations or servers either. Apple doesn't have the power to fight the DMCA or DRM. And technically, I think Darwin and Cocoa are dead ends for practical purposes and it really doesn't matter whether people clone them (GNU Darwin and GNUStep). Open source developers should emulate Macintosh style and simplicity in their; copying lots of low-level nuts and bolts is neither necessary nor sufficient for that.
So, basically, I gave up getting particularly pushed out of shape about Apple or open-source OS X-related projects either way--there just isn't any point to it.
Sure, the QT container format is free. It would be great if QT+some-open-codec became the standard for phones, streaming, and other applications. But QT+some-open-codec is not a standard and there is no open codec that people are willing to agree on, while MPEG4 is a standard.
And MPEG4 does have the advantage that it is well and completely documented; the patents covering it will probably only be valid another decade, if that.
I have yet to see any evidence that basing applications on Qt/Embedded is any more efficient/smaller than basing them on X11. If anything, and environment built around toolkits like FLTK/X11 seems much smaller and more efficient than an environment built around Qt/Embedded, even taking into account the footprint of the X11 server. Also, just in terms of architecture, I see no savings or performance gains from the Qt/Embedded architecture. If Qt/X11 has a lot more overhead than Qt/Embedded, then that's a problem with Qt, not with X11.
The short of it is that I think Opie-based handhelds should switch to Qt/X11 entirely and just get rid of Qt/Embedded. If that turns out to require more memory or be slower, then the problem is with the X11-based Qt implementation, not the use of X11, and the way to fix that is to fix the Qt/X11 implementation.
Several HCI research groups have worked on tokens and user tracking before. Some, for example, have used IR beacons emitted from lapel tags that are received by IR receivers situated near computers in order to determine whether the user is near the computer or not, and as part of that, people have discussed automatic authentication schemes (including placing the cryptographic token into the beacon).
The ZIA paper does describe a technically nice piece of work, and its specific approach may be novel. But the omission of references to prior work related to user tracking and ubiquitous computing approaches really leaves me wondering whether the authors have done their homework and whether this is really the first time that the method has been published. I think the authors would do well to track down more HCI references on beacons, wearable tags, physical user interfaces, and tracking.
Note that, in terms of hardware, you can fairly easily implement such a system these days with a Bluetooth PDA (which you wear on your person) and a disk encryption card. The range for Bluetooth is perhaps a little far, but tinkering with the Bluetooth dongle and some conductive paint should fix that.
What is becoming the "standard" on cell phones? Quicktime 6 or MPEG4? While they have some things in common in terms of bit stream, the two terms refer to very different things.
Quicktime 6 is a container format defined by Apple that might be used with a huge number of proprietary codecs, as well as a software infrastructure implementing multimedia encoding, decoding, and transport using that format; saying that something uses "Quicktime 6" doesn't tell you much about whether you can read it or not; it's like knowing that you can plug into a wall socket without knowing that the voltage is right.
MPEG4 generally refers to a specific bitstream based on a specific, standard set of codecs. Apple's Quicktime 6 happens to be able to represent MPEG4, but that's where the relationship ends. The difference between Quicktime 6 and MPEG4 is the difference between being able to encode and decode streams or not.
If phone manufacturers are actually using Quicktime 6, with multiple codecs and all, then that's a major victory for Apple and a major loss for open source and interoperability. If phone manufacturers are actually using MPEG4 but Apple calls it "Quicktime 6" for PR reasons, then that's a major PR victory for Apple, but it is hard to see what that kind of usage of MPEG4 has to do with Apple. In fact, a lot of video-based devices are already using MPEG4.
In fact, the NetworkWorldFusion article suggests that the latter is the case: NTT is switching specifically to MPEG4, not to Quicktime 6. And that's actually good.
Brand names are protected by trademarks, not copyrights. That does matter because copyright law and trademark law are very different. For example, trademarks evaporate if you don't enforce them, while copyrights don't.
And it's a "copyright"; "copywrite" is something a "copywriter" does.
That's the way even many successful startups work: they grow, then they get bought out, and the initial investors make a decent amount of money. Few startups make it on their own in the long run.
What happens to the woman who's husband was abusing her if he can track down where she now lives? [...] Okay, she may be able to avoid being beaten, but she still can't settle down and form a new life ever.
Quite to the contrary: if information on the whereabouts and actions of people is widely and publically available, she knows when her ex-husband is sticking to the restraining order and doesn't have to worry constantly. She will also know that if he violates his restraining order, it will be documented, recorded, and widely known.
Explain to me what you believe it is that guys who beat up on gays are "imagining" that lends to their fear? Once you've done that, tell me how them having more knowledge about who is and is not gay is going to curb that imagining?
I don't know whether gay bashing is due to fear or insecurity or other problems. But I suspect that most gay bashers would not engage in it if there was a high probability that video footage of their actions would end up on the six o'clock news.
People also suspect that many gay bashers are closet homosexuals, so if they found out that a significant fraction of the US population is gay and lives happy and productive lives, one might think that they might not even get that far.
you assumed that the person would either be forced to change their behavior from something that was not doing anybody any harm to something that more conforms to a whim of the moral majority
Yes, some people might choose to stop doing certain things that don't cause anybode else any harm because others disapprove. But, so what? Most people already behave that way, and when they violate majority opinion secretly, they generally feel lousy. Between the three choices of abstaining, doing it secretly, and doing it openly, doing it secretly is not a particularly good choice anyway--people end up feeling guilty and afraid of being found out. Either abstain or stand by your convictions.
On the other hand, there are many people that do things in secrecy that do harm other people: fraud, corruption, cheating on their spouse, weapons development, murder, gay bashing, spousal abuse, etc. Lifting the veil of secrecy on those actions is beneficial.
Sure, abolishing secrecy and most privacy does not fix all problems--it's a tradeoff. However, on balance, I think it's a better tradeoff than what we have now, where people have a certain illusion of privacy and secrecy in their personal lives, but everything is pretty much an open book to the government and companies anyway.
The 800MHz C3 runs like a 266MHz P3 in some benchmarks I ran. I think on the EPIA-M, they did something with the memory that may improve performance beyond the simple increase in clock speed; so, I would expect the 933MHz C3 to come in like a 300-400MHz P3. I'd expect the 900MHz Crusoe to be somewhat faster, but not a whole lot.
But for $160, including processor, the EPIA-M is an amazing deal, and it's fast enough for most day-to-day uses. Dollar for dollar, it's by far the best deal around. Those things should be in every school and business as workstations and desktop machines.
Does this sound a little like the Microsoft/Disney/AOLTW/whoever monopolistic practices to anyone else
No. With MS/Disney/AOLTW, you must pay them money to continue using the software, even if it turns out that none of their promises of support or ease-of-use work out. With Linux-based systems, including Lineo, if you don't want to deal with that vendor anymore for support, you go somewhere else (well, except for the GUI toolkit, which has restrictions for commercial use).
When I ssh to it there's little to tell me it's not a powerful server system.
The fact that most of the applications only run on the main display or via VNC might give you a hint; UNIX and Linux servers use X11, and for good reason.
Fortunately, some people are working on making Opie play nice under X11. Let's hope that Motorola's Lineo will see the light and make this the standard way of delivering Qtopia for Embedix as well.
Last I heard/read, Sprint broadband was no longer accepting new customers
Sure, that's what makes it so great:-)
Seriously, my point is that the technology works, it's fast, and it's here now. It may be too small for Sprint, but other companies could easily take the same wireless technology and provide comparable service at a comparable price.
as opposed to per machine is a HUGE advantage in my book
You must be talking about what Windows used to be like a couple of years ago, since networks of UNIX workstations have never been managed like that. Come on, people have run UNIX networks with thousands of machines since the 1980's. Do you think they didn't figure out how to deal with those issues long ago?
There are several common ways of setting up such networks, and they are generally much simpler to deal with than anything Microsoft offers even today. Adding a new machine to a UNIX network requires no more than just plugging it into the network and possibly adding it to a list of recognized clients. Users, data, and applications are installed centrally. Applications run transparently over the network, or locally, whichever way you prefer. "The latest patches" or "new applications" aren't even issues--things are just automatically consistent.
Windows has taken some of those ideas and thrown them together into an inconsistent and cumbersome juble. But where networks of UNIX workstations just tick along by themselves, Windows-based networks require constant handholding, fixing, patching, and reinstalling. Microsoft is trying to paper over how messy and dysfunctional their system is with lots of dialog boxes and GUIs, but it just doesn't help: in the end, managing Windows networks is still a lot more work. Oh, of course, you can try and buy lots of expensive third party software to get some of the UNIX-like manageability, but that only makes things even more expensive and complicated.
I used to manage networks of UNIX workstations with dozens of users on the side. If I had to spend more than an hour or two on it per week, that was the rare exception (and then it was usually due to some hardware failure on the server). And I certainly didn't need any expensive or complicated third party software for doing it either.
That's not the point: just because everyone knows your SSN and VISA card number doesn't mean nobody is going to perform an identity theft.
Identity theft is rampant because we rely on weak secrecy rather than strong cryptography for proving identity. If our SSNs, mother's maiden names, and ccard numbers were public, it would dawn on even the most idiotic bank manager that this is not the way to "protect" people's life savings.
Or: just because everyone knows who is married or not (or gay or not) doesn't mean that philanderers (or homophobes) will cease their harassing or criminal attacks.
Well, if everybody is tracked publically, then their illegal harassment and attacks would be as much public knowledge as the legal conduct they are harassing and attacking people for. I think on balance, I'd feel safer that way because it would weed out the sociopaths that harass and attack people much more quickly than we can right now. Right now, many attackers (domestic violence, homophobia, etc.) are well-know, they just can't be locked up for a long time because the legal proof is lacking.
What happens to the woman who's husband was abusing her if he can track down where she now lives?
She can also track him and his movements.
What happens to the poor 68 year old single man who happens to use his credit card to buy a lot of lacey underwear and garters that happen to be his size?
He'd probably find out that there are hundreds of thousands of others like him, or he could make an informed choice not to make the purchase.
How about the head of the PTA or school principal who in her off hours frequents the fetish clubs?
Same deal: either that becomes acceptable for the heads of PTAs or school principals, or they stop doing it.
Of course, this is ludicrous - we can't fire our boss if we don't like his choice of reading material.
Sure we can: it's called changing jobs. People do it all the time.
People tend to be reactionary, prejudiced, and frightened.
People are reactionary, prejudiced, and frightened about things they imagine but don't know. Knowledge and information is the antidote to that.
FWIW, you can do all the things you mention with Windows. [...] Before you go bashing something, please be informed about it.
Before you go criticizing something, perhaps read it more carefully: nowhere did I claim that you couldn't also use equivalent set-user-id mechanism under Windows; they just happen to be rather cumbersome to deploy and manage compared to the UNIX approach.
You see, the value of the original UNIX design is in its minimalism: it makes it easy to use a small set of necessary and sufficient mechanisms. The UNIX designers were as busy removing features from the OS as they were adding new features.
The Windows philosophy is to give you features and more features and more options, and to wrap that up in GUIs. Sorry, but more isn't better, it's usually worse.
whereas I've seen Linux machines that have gone belly-up with a rather pissed off admin trying to fix it. Then again, I spend more time in the Linux labs.
Linux machines don't just go "belly-up", and certainly not from normal usage.
In any case, a common way of dealing with this is to not worry too much about students doing stuff as root at all--you just have the machine reboot on logout and restore the default installation with "rsync".
That's easy to deal with: either you adopt the Windows model and disallow remote logins for users different from the console user, or you set reasonable limits on the number of processes and amount of memory per user.
Keep in mind that Linux, out of the box, is configured for single-user desktop use. You do have to do a little bit of configuration for a multi-user environment.
Group Policy kicks ass. [...] The point is, Linux has nothing to compare with this.
Sure it does. By default, regular Linux users can perform no system management functions. You give them access to system management functions through setuid and setgid programs. You can control access to those on a per-user or per-group basis using standard UNIX protection mechanisms.
If you like something more general, you can use the "sudo" program, which allows detailed policies to be specified of who can do what as who and when, and it also logs the actions.
The FACT is no one has a better way to administrate and trouble-shoot end-user desktops than Microsoft right now.
As usual, Microsoft has an in-your-face solution that screams at you "I let you edit policies; here is a point-and-click interface--isn't it easy?". Trouble is, in real life, the options it gives you are rarely the options that are needed, and extending and managing those policies is a chore.
The UNIX/Linux solution is simple, elegant, powerful, and has proven itself for more than 20 years in large, multi-user environments.
So, the "FACT" is, "Windows Policy Editor" is indeed like a lot of Windows: flashy but not all that useful in practice.
The amazing thing about Vivato's stuff is not that it delivers broadband over a few miles distance, but that it does so within the regulatory confines of 802.11b and without manual aiming of antennas. If you just want to deliver broadband over long distances, there are lots of ways of doing that.
and replacement of outdated wireless technology [Sprint broadband]
Well, Sprint Broadband works, it delivers >3Mbps, it's fairly easy to install, and it costs $50/mo. And I doubt it's a money losing venture, otherwise they'd have discontinued the service entirely rather than just not taking new signups.
If companies will compete with Sprint broadband using Vivato technology, that would be great. But with the Vivato APs being released in 2003, I think it's at least another year away until you are going to see viable commercial broadband services based on it springing up.
Over the same area, it should actually reduce interference compared to trying to cover the same area with regular access points.
Think of it this way. With a normal access point, it's like lighting a stage with diffuse lighting: there ends up being light everywhere. This access point is intended to be like a bunch of spotlights on a dark stage: only the areas where it is aimed are actually lit up; the rest of the stage remains in darkness.
The problem with "TIA" is not that it peeks into everybody's finances and personal life--the problem is that it gives the government so much more information than private citizens. That kind of information monopoly means that the government can then use blackmail to manipulate people. And don't think it doesn't happen--this kind of thing has a long history, both in the US and other countries.
If information like taxes, license plates and vehicle registration, purchase patterns, driving records, medical treatments, etc., were universally and publically available, I think we would have fewer problems than we have now. Most people would realize that their deep, dark secrets are not so deep and dark--that there are many other people with similar issues. It would keep politicians and regular folks more honest and polite--because nothing would be really anonymous anymore. And blackmail would be pretty much impossible--how can you blackmail someone if everybody can find out almost anything anyway? And, finally, people could negotiate their salaries sensibly--right now, chances are you don't know how much you are being paid relative to your co-workers--how can you ever get an efficient labor market if the prices are not known?
Of course, public access does not mean that things need to be as easy as Googling someone. I think Brin has captured a good balance between privacy and publically accessible information in his book "Kiln People" (it's incidental to the story): basically, you can find out, but the data is not aggregated in a single place, so if you do want to find out, it still costs you some time and money.
For commercial CD replication (which is what was probably involved here), the speed of the CD burners matters.
However, that is no excuse for lying about the actual numbers. "156 CD burners" are "156 CD burners", no ifs or buts. Since most people use high-speed burners already, that's what they will assume a "CD burner" refers to, and inflating the numbers is seriously misleading.
What would be legitimate would be a number like "156 CD burners, capable of producing 25000 counterfeit CDs a day".
Slashdot Mirror
And what, specifically, do you think is wrong with TCP/IP? It's pretty minimal and simple.
Although TCP/IP is so commonplace I wouldn't want my 15 million dollar satellite to depend on it.
Even if TCP/IP had some technical drawbacks relative to some alternative protocol, software implementation errors and engineering mistakes are likely much bigger risks than some theoretical limitations of the protocol.
In different words, I'd much rather bet $15 million on a proven, debugged, mature TCP/IP implementation than on some implementation nobody has ever used for a protocol nobody has ever heard of.
I do appreciate Apple trying to cooperate with the free software community to some extent. But, frankly, I don't expect too much from it. I think there is really not all that much Apple has to offer, and most of the things that Apple has opened up were really forced by licenses or business realities (e.g., gcc was forced by the license, Rendezvous and QuickTime were forced by business considerations).
Overall, I think it just doesn't matter much what Apple does, either technically or in terms of their business. I don't see Apple winning in a big way: they can't take away much market share from Windows, and, based on my experience with them, I think they are not a major competitor for UNIX or Linux workstations or servers either. Apple doesn't have the power to fight the DMCA or DRM. And technically, I think Darwin and Cocoa are dead ends for practical purposes and it really doesn't matter whether people clone them (GNU Darwin and GNUStep). Open source developers should emulate Macintosh style and simplicity in their; copying lots of low-level nuts and bolts is neither necessary nor sufficient for that.
So, basically, I gave up getting particularly pushed out of shape about Apple or open-source OS X-related projects either way--there just isn't any point to it.
And MPEG4 does have the advantage that it is well and completely documented; the patents covering it will probably only be valid another decade, if that.
The short of it is that I think Opie-based handhelds should switch to Qt/X11 entirely and just get rid of Qt/Embedded. If that turns out to require more memory or be slower, then the problem is with the X11-based Qt implementation, not the use of X11, and the way to fix that is to fix the Qt/X11 implementation.
The ZIA paper does describe a technically nice piece of work, and its specific approach may be novel. But the omission of references to prior work related to user tracking and ubiquitous computing approaches really leaves me wondering whether the authors have done their homework and whether this is really the first time that the method has been published. I think the authors would do well to track down more HCI references on beacons, wearable tags, physical user interfaces, and tracking.
Note that, in terms of hardware, you can fairly easily implement such a system these days with a Bluetooth PDA (which you wear on your person) and a disk encryption card. The range for Bluetooth is perhaps a little far, but tinkering with the Bluetooth dongle and some conductive paint should fix that.
Quicktime 6 is a container format defined by Apple that might be used with a huge number of proprietary codecs, as well as a software infrastructure implementing multimedia encoding, decoding, and transport using that format; saying that something uses "Quicktime 6" doesn't tell you much about whether you can read it or not; it's like knowing that you can plug into a wall socket without knowing that the voltage is right.
MPEG4 generally refers to a specific bitstream based on a specific, standard set of codecs. Apple's Quicktime 6 happens to be able to represent MPEG4, but that's where the relationship ends. The difference between Quicktime 6 and MPEG4 is the difference between being able to encode and decode streams or not.
If phone manufacturers are actually using Quicktime 6, with multiple codecs and all, then that's a major victory for Apple and a major loss for open source and interoperability. If phone manufacturers are actually using MPEG4 but Apple calls it "Quicktime 6" for PR reasons, then that's a major PR victory for Apple, but it is hard to see what that kind of usage of MPEG4 has to do with Apple. In fact, a lot of video-based devices are already using MPEG4.
In fact, the NetworkWorldFusion article suggests that the latter is the case: NTT is switching specifically to MPEG4, not to Quicktime 6. And that's actually good.
And it's a "copyright"; "copywrite" is something a "copywriter" does.
That's the way even many successful startups work: they grow, then they get bought out, and the initial investors make a decent amount of money. Few startups make it on their own in the long run.
Quite to the contrary: if information on the whereabouts and actions of people is widely and publically available, she knows when her ex-husband is sticking to the restraining order and doesn't have to worry constantly. She will also know that if he violates his restraining order, it will be documented, recorded, and widely known.
Explain to me what you believe it is that guys who beat up on gays are "imagining" that lends to their fear? Once you've done that, tell me how them having more knowledge about who is and is not gay is going to curb that imagining?
I don't know whether gay bashing is due to fear or insecurity or other problems. But I suspect that most gay bashers would not engage in it if there was a high probability that video footage of their actions would end up on the six o'clock news.
People also suspect that many gay bashers are closet homosexuals, so if they found out that a significant fraction of the US population is gay and lives happy and productive lives, one might think that they might not even get that far.
you assumed that the person would either be forced to change their behavior from something that was not doing anybody any harm to something that more conforms to a whim of the moral majority
Yes, some people might choose to stop doing certain things that don't cause anybode else any harm because others disapprove. But, so what? Most people already behave that way, and when they violate majority opinion secretly, they generally feel lousy. Between the three choices of abstaining, doing it secretly, and doing it openly, doing it secretly is not a particularly good choice anyway--people end up feeling guilty and afraid of being found out. Either abstain or stand by your convictions.
On the other hand, there are many people that do things in secrecy that do harm other people: fraud, corruption, cheating on their spouse, weapons development, murder, gay bashing, spousal abuse, etc. Lifting the veil of secrecy on those actions is beneficial.
Sure, abolishing secrecy and most privacy does not fix all problems--it's a tradeoff. However, on balance, I think it's a better tradeoff than what we have now, where people have a certain illusion of privacy and secrecy in their personal lives, but everything is pretty much an open book to the government and companies anyway.
But for $160, including processor, the EPIA-M is an amazing deal, and it's fast enough for most day-to-day uses. Dollar for dollar, it's by far the best deal around. Those things should be in every school and business as workstations and desktop machines.
No. With MS/Disney/AOLTW, you must pay them money to continue using the software, even if it turns out that none of their promises of support or ease-of-use work out. With Linux-based systems, including Lineo, if you don't want to deal with that vendor anymore for support, you go somewhere else (well, except for the GUI toolkit, which has restrictions for commercial use).
The fact that most of the applications only run on the main display or via VNC might give you a hint; UNIX and Linux servers use X11, and for good reason.
Fortunately, some people are working on making Opie play nice under X11. Let's hope that Motorola's Lineo will see the light and make this the standard way of delivering Qtopia for Embedix as well.
Sure, that's what makes it so great :-)
Seriously, my point is that the technology works, it's fast, and it's here now. It may be too small for Sprint, but other companies could easily take the same wireless technology and provide comparable service at a comparable price.
You must be talking about what Windows used to be like a couple of years ago, since networks of UNIX workstations have never been managed like that. Come on, people have run UNIX networks with thousands of machines since the 1980's. Do you think they didn't figure out how to deal with those issues long ago?
There are several common ways of setting up such networks, and they are generally much simpler to deal with than anything Microsoft offers even today. Adding a new machine to a UNIX network requires no more than just plugging it into the network and possibly adding it to a list of recognized clients. Users, data, and applications are installed centrally. Applications run transparently over the network, or locally, whichever way you prefer. "The latest patches" or "new applications" aren't even issues--things are just automatically consistent.
Windows has taken some of those ideas and thrown them together into an inconsistent and cumbersome juble. But where networks of UNIX workstations just tick along by themselves, Windows-based networks require constant handholding, fixing, patching, and reinstalling. Microsoft is trying to paper over how messy and dysfunctional their system is with lots of dialog boxes and GUIs, but it just doesn't help: in the end, managing Windows networks is still a lot more work. Oh, of course, you can try and buy lots of expensive third party software to get some of the UNIX-like manageability, but that only makes things even more expensive and complicated.
I used to manage networks of UNIX workstations with dozens of users on the side. If I had to spend more than an hour or two on it per week, that was the rare exception (and then it was usually due to some hardware failure on the server). And I certainly didn't need any expensive or complicated third party software for doing it either.
Identity theft is rampant because we rely on weak secrecy rather than strong cryptography for proving identity. If our SSNs, mother's maiden names, and ccard numbers were public, it would dawn on even the most idiotic bank manager that this is not the way to "protect" people's life savings.
Or: just because everyone knows who is married or not (or gay or not) doesn't mean that philanderers (or homophobes) will cease their harassing or criminal attacks.
Well, if everybody is tracked publically, then their illegal harassment and attacks would be as much public knowledge as the legal conduct they are harassing and attacking people for. I think on balance, I'd feel safer that way because it would weed out the sociopaths that harass and attack people much more quickly than we can right now. Right now, many attackers (domestic violence, homophobia, etc.) are well-know, they just can't be locked up for a long time because the legal proof is lacking.
She can also track him and his movements.
What happens to the poor 68 year old single man who happens to use his credit card to buy a lot of lacey underwear and garters that happen to be his size?
He'd probably find out that there are hundreds of thousands of others like him, or he could make an informed choice not to make the purchase.
How about the head of the PTA or school principal who in her off hours frequents the fetish clubs?
Same deal: either that becomes acceptable for the heads of PTAs or school principals, or they stop doing it.
Of course, this is ludicrous - we can't fire our boss if we don't like his choice of reading material.
Sure we can: it's called changing jobs. People do it all the time.
People tend to be reactionary, prejudiced, and frightened.
People are reactionary, prejudiced, and frightened about things they imagine but don't know. Knowledge and information is the antidote to that.
Before you go criticizing something, perhaps read it more carefully: nowhere did I claim that you couldn't also use equivalent set-user-id mechanism under Windows; they just happen to be rather cumbersome to deploy and manage compared to the UNIX approach.
You see, the value of the original UNIX design is in its minimalism: it makes it easy to use a small set of necessary and sufficient mechanisms. The UNIX designers were as busy removing features from the OS as they were adding new features.
The Windows philosophy is to give you features and more features and more options, and to wrap that up in GUIs. Sorry, but more isn't better, it's usually worse.
Linux machines don't just go "belly-up", and certainly not from normal usage.
In any case, a common way of dealing with this is to not worry too much about students doing stuff as root at all--you just have the machine reboot on logout and restore the default installation with "rsync".
Keep in mind that Linux, out of the box, is configured for single-user desktop use. You do have to do a little bit of configuration for a multi-user environment.
Sure it does. By default, regular Linux users can perform no system management functions. You give them access to system management functions through setuid and setgid programs. You can control access to those on a per-user or per-group basis using standard UNIX protection mechanisms.
If you like something more general, you can use the "sudo" program, which allows detailed policies to be specified of who can do what as who and when, and it also logs the actions.
The FACT is no one has a better way to administrate and trouble-shoot end-user desktops than Microsoft right now.
As usual, Microsoft has an in-your-face solution that screams at you "I let you edit policies; here is a point-and-click interface--isn't it easy?". Trouble is, in real life, the options it gives you are rarely the options that are needed, and extending and managing those policies is a chore.
The UNIX/Linux solution is simple, elegant, powerful, and has proven itself for more than 20 years in large, multi-user environments.
So, the "FACT" is, "Windows Policy Editor" is indeed like a lot of Windows: flashy but not all that useful in practice.
The amazing thing about Vivato's stuff is not that it delivers broadband over a few miles distance, but that it does so within the regulatory confines of 802.11b and without manual aiming of antennas. If you just want to deliver broadband over long distances, there are lots of ways of doing that.
Well, Sprint Broadband works, it delivers >3Mbps, it's fairly easy to install, and it costs $50/mo. And I doubt it's a money losing venture, otherwise they'd have discontinued the service entirely rather than just not taking new signups.
If companies will compete with Sprint broadband using Vivato technology, that would be great. But with the Vivato APs being released in 2003, I think it's at least another year away until you are going to see viable commercial broadband services based on it springing up.
Think of it this way. With a normal access point, it's like lighting a stage with diffuse lighting: there ends up being light everywhere. This access point is intended to be like a bunch of spotlights on a dark stage: only the areas where it is aimed are actually lit up; the rest of the stage remains in darkness.
If information like taxes, license plates and vehicle registration, purchase patterns, driving records, medical treatments, etc., were universally and publically available, I think we would have fewer problems than we have now. Most people would realize that their deep, dark secrets are not so deep and dark--that there are many other people with similar issues. It would keep politicians and regular folks more honest and polite--because nothing would be really anonymous anymore. And blackmail would be pretty much impossible--how can you blackmail someone if everybody can find out almost anything anyway? And, finally, people could negotiate their salaries sensibly--right now, chances are you don't know how much you are being paid relative to your co-workers--how can you ever get an efficient labor market if the prices are not known?
Of course, public access does not mean that things need to be as easy as Googling someone. I think Brin has captured a good balance between privacy and publically accessible information in his book "Kiln People" (it's incidental to the story): basically, you can find out, but the data is not aggregated in a single place, so if you do want to find out, it still costs you some time and money.
However, that is no excuse for lying about the actual numbers. "156 CD burners" are "156 CD burners", no ifs or buts. Since most people use high-speed burners already, that's what they will assume a "CD burner" refers to, and inflating the numbers is seriously misleading.
What would be legitimate would be a number like "156 CD burners, capable of producing 25000 counterfeit CDs a day".