I have little doubt that you would be annoyed if the Pledge said "Under Allah". Whether or not the majority happened to believe in something when the country was founded is immaterial; it is the minority that needs protection, not the majority.
Considering that "Under Allah" would undoubtedly annoy many, the most cursory examination would indicate that "Under God" would annoy at least some Muslims. And, since the Pledge is about as official a bit of government communication as could possibly exist, it has to have some chilling effect on those who do not believe in God, an implication that they are second-class citizens.
Additionally; I am unsure of this, but as a condition for naturalization, isn't a new citizen required to cite the Pledge of Allegiance? That would also seem to violate the establishment clause. If this is true, aren't we in essence forcing a new citizen to recognize the existence of God?
The phrase "Under God" was added in the 1950s. It was not a part of the original conception; it is likely that the Founding Fathers would have been horrified at the thought. And It is quite clear just exactly why "under God" was added: it is there to recognize, venerate, and give thanks to a supreme being, one of a particular class of religious thought.
Since "Under Allah" wouldn't work in the Pledge, "under God' doesn't belong either. It really is that simple.
We are not a nation under God, we are a nation under the Constitution, and all religions are welcome here.
That's good to hear. I was specifically criticizing the Windows implementation, which I despise. I haven't used 9.2 yet, so I can't exactly criticize it.:-)
This is from memory without double-checking, so don't treat this as 100% factual without double-checking it.
Mandrake has been having financial trouble for some time, but they've been steadily improving. Basically, they wasted tons of money going after stupid stuff that the venture-capital type management wanted to do (mostly education), instead of spending their money where they actually made revenue.
They declared bankruptcy late last year or early this year. With their debt payments suspended, they're "cash-flow positive" at the moment, and are hoping to get back to full normal operations (ie, paying down debt) by the end of the year.
[end bad memory warning area]
I'm a Silver member of their club and plan to continue my membership. I like their distro very much and feel it's worth supporting. I know that the Club is doing fairly well by them and they're pleased with the results, but the more members the better... it's not that expensive, gives you some nice little perks, and helps make sure a really good distro keeps evolving. I think of it like my local NPR station and like the EFF.. a small contribution by a lot of people can really make a difference.
Your confusion may have arisen because of all the ill-wishing the Slashdot crowd was spouting when they were really hurting for money... "I hope they die, they deserve to", that kind of garbage. Fortunately, crap spewage and reality did not intersect.
disclaimer: I have no connection to Mandrake other than my membership and a couple of installations of their distro.
Personally, I find that "feature" to be EXTREMELY annoying.... if you do it too fast, it's a double click, but if you do it slower than you need to, it's wasted time.
"Click" and "double click" are ok metaphors, but "two slow clicks" is horrid, mostly because "slow" is so ill-defined. A very poor interface mechanism, IMO.
I've been using Mandrake a long time, but I didn't get "serious" about it until the 8.X series.
I recall 8.0 and 8.1 as being a little flaky, but 8.2 as being very solid. I had TONS of problems when I upgraded to 9.0... not because of the distro itself, but because everything ELSE broke. They did some upgrade ( to glibc maybe?) that broke almost all the RPMs I tried from earlier versions. I was able to get around it in many cases by rebuilding source RPMS, but it was still painful. I should have waited longer.
I'm presently on 9.1, and am happy with it. I use the High Performance Liquid theme (you can get it in the Penguin Liberation Front rpms) on KDE. I get an occasional Konqueror segfault, probably about twice a week, but other than that it's been very stable. I'm using it on a P4 1.8, Intel chipset, GF4MX440, Nvidia binary drivers.
My home Linux machine has had occasional lockups of X (machine still up, X frozen), which seem related to my switchbox. I am of the belief that the lockups aren't Mandrake's fault...I'm pretty sure it's the Nvidia drivers. I haven't had the problems with the most recent version, but I also haven't been using the machine very much for the last few weeks, so I'm not sure it's fixed yet. That machine is an Athlon 1900+ on an Asus A7V333, GeForce Ti4200.
On the whole, I'd call it about equivalently stable to XP, maybe a little less. The kernel itself, of course, is almost uncrashable, but X is a little more fragile than I'd like.
What really killed Commodore was their mishandling of the Amiga. They had possibly the finest system engineering team ever assembled, and treated them like dirt; they had neither any idea what genius of that caliber was worth, nor just what that system could have become if left in the same super-capable hands.
I have often said that if Apple had owned the Amiga technology, we'd all be using descendants of that machine instead of the PC. It was that good. Put every other machine in its era absolutely to shame. They sold several million Amigas based almost purely on word of mouth; with a good management team and advertising behind it, it would have taken over practically everything.
C= didn't just shoot themselves in the foot with the Amiga, they shot themselves in the head. Several times. Revenue from the 64 made the Amiga buyout and manufacture possible, but it was the Amiga itself that was the make or break for the company. Break, as it turned out.
The article's lead-in was deliberately inflammatory. The actual quote was:
Paul Bresson, an F.B.I. spokesman, said, "It is legal; it is a concern; and it is something that we are looking into."
This is journalism of the caliber of talk radio; leaving out the important part "it is legal" to highlight the dreaded imminent threat. Probably helped get it chosen as a story, since putting one to the Feds is always popular in the geek crowd, but it's still irresponsible.
You're right about it being ready as a desktop. At my present job, I started with both a Windows and a Mandrake PC. (my choice on Mandrake). The drive in the Windows machine started to fail after a couple of months, and I never bothered fixing it. Ended up just using that machine for parts to fix others. I was perfectly happy with Mandrake, and still am, a year and a half later. KDE has been truly desktop-ready for at least 1.5 years, possibly longer. I don't have an opinion on GNOME, since I don't use it, but I'm very pleased with KDE.
But being "just as easy as windows" isn't good enough. It needs to be easier than Windows to really take off. "Just as good as but free" will definitely sell people on Linux, but "easier, more powerful, and free" will move it by the boatload.
We need to also remember that OS choice is a social phenomenon as much as a technical one. People have been taught for 10+ years that "PCs run Windows". It's going to take some time to overcome that conditioning, no matter how good the alternatives are.
I actually disagree with this. As long as the expectation isn't that the software is free, and that the person copying the software knows it is wrong to do so, then even large-scale copying isn't that large a problem.
If Joe Badguy sends out 10,000 copies of Quicken, then at least some of those people are going to upgrade Quicken by paying for it in later revisions. This may result in MORE revenue, not less. In addition, those 10,000 people A) have not given Microsoft any money, B) are using and learning to like Intuit's product, and C) may recommend it to their friends.
If copying becomes really widespread, it could become a problem... if it becomes so accepted that common sense says "it's stupid to pay for Quicken when you can just copy it", then perhaps Intuit wouldn't be able to make money.
Note, however, that this HASN'T HAPPENED YET. Intuit has made HUGE amounts of money with very little copy protection. They trusted their customers, focused on them, solved their problems, and in exchange they made boatloads of cash.
But instead of focusing on their customers, now they are focusing on "all those deadbeats". Intuit is trying to strong-arm pirates into buying the software they copied. This hurts the usability of the software for people that DID and DO pay them money, in an effort to extort money from people who HAVEN'T. Those people who haven't paid for it probably have reasons for it, and while Intuit can undoubtedly coerce some of them into paying, they inconvenience their existing customers enough that they lose more than they gain. They are thinking purely in terms of "deadbeats" and not at ALL about marketing and word of mouth and network effects.
Trying to force someone into a business relationship is not a long-term route to profitability. (although the Mafia may disagree with me there.:-) )
Microsoft can get away with this stuff because, for most people, there are no alternatives. Most people think they have to use Windows; they probably don't even realize there's any real choice. That's how Microsoft can get noxious ideas like Product Activation through.
Microsoft, being a monopoly, IS in a position to force people to do things they don't want to. Intuit is not, and they may have just learned that lesson.
I'm a lost customer to you; it is unlikely that I would ever buy another Intuit product. I've been a paying customer of yours for years, both at retail and directly from you, but this year I bought TaxCut and am perfectly happy with it. I am still shocked that you think that treating your customers like thieves is a good business practice. I see you have an "unlock code" now for Quicken too, from the front page.
You're supposed to be in the business of SOLVING customers' problems, not CREATING them.
Casual copying probably helps you; it gets people hooked on your product, and it turns even people who didn't pay for it into advertising. If everyone is using a product, some percentage will always buy new versions, and if you provide a good quality, problem-free product, that percentage will most likely climb steadily over time. Mindshare really matters. If I ask five friends "what's the best tax program", and they all tell me Turbotax, that's going to have a much greater impact on me than three Turbotax votes and two TaxCut votes.
As you just discovered from watching your revenue plummet, getting 50% paid customers from three times the installed base nets you more money.
A personal example: I used my parents' copy of TurboTax for a few years back in the early 90s. Starting sometime around 1995, I converted into a paying customer myself. My "theft' of your product *turned me into a customer*. I bought the product for years. This year you lost me, and you probably won't get me back.
It is casual copying that got Microsoft where they are today. That casual copying, in fact, turned them into the one of the most profitable companies in the world. Don't think their use of DRM is helping them; it is causing fragmentation in their customer base. People are now actively resistant to upgrading to the most recent versions of things, and this is providing competitors like Linux an opportunity to expand at their expense. You can't see it as much yet, because Microsoft is a monopoly, but DRM is hurting them. Think of it as cracks in the foundation.
Intuit isn't a monopoly and has competitors to all of its products. Your decision to try to limit casual copying and to treat customers like thieves hurt you badly. I'm a lost customer. I know there are lots of others. And I'm not at all sure you're going to get many of us back.
TaxCut is really quite good, and I imagine I'll be checking out Money soon. And I'll most likely make a dreaded casual copy of Money to do it. Microsoft may not know it, but they want me to do that; any chance of converting me from a "deadbeat" into a paying customer is better than the zero money they're making from me now.
And if a friend asks me "hey, what's the best software to track my checkbook?", and I answer "Microsoft Money", well.... suddenly the "Quicken is best" message is no longer unanimous and maybe Microsoft makes a sale.
The author of this article mentions only three real differences between Linux and Windows; Linux users don't run as root, Linux email programs don't automatically run attachments, and Linux has more than two main programs to read mail.
Now, I have believed for a long time that Linux is less susceptible to virii than Windows. I stated as much in a fairly long post in a Lindows discussion, which basically boiled down to "don't run as root!". I got some great replies to that post which shifted my thinking quite a bit.
Now, I still stick to my original assertion that running as root is bad. The single biggest reason is rootkits; if you are compromised while running as root (or if a process running as root is exploited), it's possible for the exploit software to hide itself from you almost completely. A good rootkit can make itself extremely difficult to detect, and no user-space virus can do that without a further exploit.
That said, it's perfectly possible to have userspace virii. They can't hide themselves as well, but they can still run, and can still propagate. And while they can't take down *the whole machine*, they can certainly wipe out your home directory, and for most users, it's almost the same thing.
We're used to looking at this as sysadmins; as long as the system stays intact, we can restore data. But for home users, most of whom don't back things up, losing the home directory is about the same as losing the machine. They may save a couple of hours on the reinstall, but compared to the weeks or months to recreate data, big deal. We can be all smug about "the system wasn't compromised!", but the user lost everything, and from his or her perspective, that's all that matters.
So the "no running as root" idea is useful for disinfection, but doesn't help at all for data loss, and doesn't stop propagation in most cases. Outbound connections can be made just fine from userspace.
The more secure email programs are good, but I suspect that may be a temporary advantage. There's a natural tendency toward monopoly in system software; over time, groups of people tend to converge on similar software. My office, for instance, has three main Linux users; each of us runs a different distro, but we have all independently chosen Evolution for email. Less security means more ease-of-use, and more ease-of-use means more uptake among the end-user population, so over time, it seems likely that Linux email programs will become less and less secure. The Linux email program advantage is very high, but over time, I believe it will be much less so.
Additionally, there's more and more tendency for everyone to use the same fundamental object libraries, like, say, OpenSSL. This tendency to write against the same core libraries means that whole classes of programs may suddenly become exploitable at once.
It's easy for us to get complacent, but our armor here is pretty thin. The lack of a Linux monopoly really may be much of the reason we don't see many virii.
After all, how many email virii do you know that spread via Eudora on Windows?
Hey, out of interest, if you encode with the LAME presets (--alt-preset standard, extreme, and insane), can you still hear the MP3ness on a rig that good?
I listen on (good) headphones and I can sometimes tell with --aps, but I just can't hear the difference with --ape. (or, presumably, with --api, but I haven't bothered checking.)
All networks are fundamentally shared bandwidth. It just depends on where the bandwidth is shared.
The DSL circuit itself isn't shared, and the full b/w allocation gets to the router, but starting there and going upstream, ALL connections are shared. T1, cable, DSL -- they are all oversubscribed; the ISP sells more bandwidth than they buy. As far as I know, this is universal, except possibly for backbone providers, which probably have to overprovision in case of failure.
The amount of oversubscription matters more than where it is. If your neighborhood is trying to run 20 megabits/second over a 2 megabit/second pipe, things are going to suck. But if a DSL provider is trying to aggregate 500 "unshared" connections through a T3 (45 megabits), it's also going to suck. Distance isn't completely unrelated, though; if your neighborhood is oversubscribed, EVERYTHING will suck. If the connection to just, say, AT&T is overtaxed, then only sites behind that connection will be bad.
That said, in general, cablemodem is oversubscribed worse than is DSL, and both are far worse than T1-class (and higher) connections. That's why they're cheaper.
Many (most?) cable/DSL companies are in the business of selling you something with the expectation that you will not actually use it.. In many cases, if you actually DO have the temerity to use what you were told you were getting, you damage their network and business plan and they will get rid of you. In general, cable providers are much worse than DSL providers in this way.
The fundamental problem is, of course, that their actual bandwidth needs are outstripping their flawed projections, and instead of buying more bandwidth (which costs a lot of money), they're trying to force usage back into their models. This is a cap on usage, whether or not they will admit it.
Also, the AUPs on many of these networks are horrible; if you read them, for the most part you are not allowed to run a server of any kind. Period. Some of them even say you can't use VPN software. At least one provider was going so far as to block protocols and ports used for VPN, unless you upgraded to a "business class" account. (ie, if you actually want to USE your circuit, well, that costs extra.) Moral: read your AUPs carefully.
I'm currently a Speakeasy customer and very happy with them. Their AUP is quite liberal; it's basically "don't do illegal things with our network, don't DOS people, don't spam, and don't abuse our staff." Their fundamental business model is sustainable; it's built on the assumption that you will use your bandwidth. As long as you're using it legally, I don't think they care how much traffic you send. The network is good, with no visible packet loss and reasonable latency, although I wish the latency was a trifle better. (I got spoiled on PacBell's ATM network.) It's still lightyears ahead of the Comcast connection my sister has.
When I was in California, PacBell was very similar. I was one of the very first people to sign up for DSL when the price got reasonable, so their AUP may have changed, but as of several years ago it was very similar to Speakeasy's. At the time, their network was incredibly good, with super-low latency and great connectivity almost everywhere. I've been gone three years, so I don't know if it's still as good now, but when I left they were stellar.
Many cable companies are, in my opinion, involved in a bait-and-switch with their networks. You get higher burst download speeds, but usually the latency is much worse, and they can terminate you for daring to actually use the service you're paying for. You'll pay more, but I believe DSL from a good provider is a better bet.
Well, I thought about mentioning Windows accelerators in the VLB timeframe, but it wasn't directly related. AGP was invented, more than anything else, for games, and as far as I know, most games don't use the Windows graphics-acceleration functions. Back in the day, they were all written in DOS, and nowadays, they all use DirectX. Admittedly, some Windows-based strategy games use standard Windows rendering functions, but those games aren't the type that require fast graphics anyway.
You say that texture loading is gawdawful slow, but I disagree. It's not real-time, but you can load a 128MB texture buffer in about a second with PCI. Standard AGP is only twice that fast, or about a half-second... it's just not that much of a difference. Games that do a lot of texture swapping are going to show a hit, but as far as I know, EVERY major game out there is designed around the idea of fitting textures into memory and only swapping them out occasionally.
I suppose I should also cop to the fact that the HD is sometimes on the same bus as a PCI video card, so the available bandwidth might be cut in half (have to copy from the HD to the PCI card), whereas AGP is always on a separate bus. That would probably double the load time again, up to two seconds. But I think that's probably worst-case, assuming a decent PCI implementation. Pretty acceptable, IMO.
Conclusion: AGP, for most uses, is overrated. PCI-X, for the current games and video apps, isn't really necessary. But it might allow a major shift (finally) in how texturing works. And, as you reminded me, the ability to read from video memory at full speed might be really useful.
This prompts the thought that the GPUs are just monstrously powerful, far far faster than general-purpose CPUs. If the PC had a way of sending a program to the GPU and then reading the results back at full speed, it might be possible to jack those distributed.net results WAY up.:-)
Re:I wouldn't buy the Athlon anyway
on
Is Prescott 64-bit?
·
· Score: 4, Informative
Athlon chipsets sucked rocks for a long time, and were really unstable. But VIA finally got their act together, I think with the KT133A.
AFAIK, other than stomping on occasional driver bugs, Athlon chips have been pretty excellent ever since. I have an Athlon 1900+ on an ASUS A7V333 that's rock solid, and a new Athlon 2500+ on an Nforce2 board that's not quite as solid, but which is still pretty good.
I'd like to see some improvements on the NForce2 chip stability. It's not all the way there yet, in my opinion. But the VIA chipsets are extremely solid.
It will if it's based on Linux. NAT is part of the routing process. You can't do NAT without routing, but doing NAT in *no way* stops regular routing from working.
In other words, if I say "translate 192.168.0.10 to address 1.2.3.4", and then I get packets at the external interface headed to 192.168.0.10, they will be routed. Further, the replies will come back without being NATted, because the kernel is doing stateful inspection, and it will know that the reply to a "normal" connection shouldn't be tampered with.
I use this functionality all the time, so I know exactly what I'm talking about.
Oops, I made a math mistake here. AGP 8x should allow an absolute maximum of about 275 frames/second of 1600x1200x32 bit color. It would only do a thousand frames if it was in 8 bit color. Forgot to divide by 4. Sorry.
What he's saying, and perhaps isn't doing a very good job, is that NAT and firewalling are separate entities.
Most consumer-level NAT devices use 192.168.0.0 or 1.0/24 as their NAT network. What mindstrm is trying to say is that your provider could easily add a route for 192.168.0.0/24 aimed at your DSL router. A machine doing NAT-only translation will happily forward packets addressed directly to its "private" network. It will do NAT if the rules are triggered, but if they aren't, it *still routes the packets*, so Joe Evil Employee can access your machines freely. He can ping, nmap, and hack remotely any machine he likes in your "private" network. To him, you look exactly the same as any other un-firewalled machine in the world. There's nothing magical about 192.168.0.0; it's set aside for private use, and is used in probably hundreds of thousands of installations, but if the ISP routes to your modem, he's temporarily saying "ok, you are the real 192.168.0.0 as far as I'm concerned"... and he's in.
In real life, of course, that mostly won't work, because NAT devices *also* do firewalling, and will refuse to route the inbound packets. What mindstrm is trying to point out is that it is really the firewalling that gives you the security, not the NAT.
NAT does protect some against people coming from the "rest of world", those who don't have access to all the routers between their machine and yours. But firewalling provides security that's both a lot better and a lot more flexible.
If NAT goes away, it won't cause any appreciable security loss.
Actually, AGP was really developed to solve a problem that went away on its own; 2-D graphics.
The big bottleneck on PC graphics for years and years was the bus speed. When you are doing 2-D graphics, in essence you have to copy your graphic data out, frame by frame, to the display memory. The system bus was always the bottleneck here. To animate a 320x200 screen at 30 frames per second, you have to push out about 2 megabytes per second. 640x480 is four times that; 1024x768 takes about 24 megs per second. These are all in 1-byte pixels, or 256 colors; to do this in 32-bit color you'd have to push 4 times as much again, or about 95 megabytes/second for that 1024x768 screen. The numbers go up really, really fast as you get to higher resolutions.
So the big thing with PC graphics, for a long time, was increasing the bus speed. The original 8-bit ISA bus can push 4.77 megabytes per second, which wasn't able to animate even 320x200 because that same bus also had to do all memory access and other I/O. The 16-bit ISA slots could do 16 megabytes/second (8mhz x 16 bits.): with the system overhead, you could definitely do 320x200, and you could probably do 640x400 with very, very clever programming, but it would be iffy.
It was about then that graphics really started getting important, and VESA Local Bus was invented to extend the spec; I believe that was a 32-bit bus running at 33, 40, or 50 megahertz, depending on the CPU it was attached to. (VLB was a very simple design that was "close" to the processor electrically, and ran off its front-side bus.) As per the calculations above, you could put out a lot of data to a VLBus card, enough to render a 1024x768 screen. PCI was invented at about the same time, and while it wasn't as fast as a 50Mhz VLB card, it was flexible enough that it eventually supplanted VLBus, which died quietly.
PCI can do 133 megabytes/second (33 mhz x 32 bits wide). so 1024x768 is about the hard limit there. I'm not sure if they were pulling main memory off onto its own bus yet (as they do with modern machines), so if the video card was still competing with main memory, true 1024x768 at 30 frames per second would have been very difficult. With a modern machine, it would be no problem, as long as the computer wasn't trying to hit any of the other PCI cards too hard.
Well, Intel could see the writing on the wall, and came up with AGP. AGP 1x runs at twice the speed of PCI and hence has twice the bandwidth; the later 2x, 4x, and 8x specs doubled the speed each time. An AGP8x board can shovel out about 2.1 gigabytes/second, which is enough to comfortably animate 1600x1200 at several hundred frames per second. (perhaps as much as a thousand with, again, clever programming.)
But while they were busily solving the bandwidth problem, it went away. All this speed isn't really being used anymore. Over the last few years, PCs have switched away from using 2-D graphics to using 3-D graphics for most games. And 3-D is represented very differently; it is sent to the graphics card as a series of textures and triangles, and rendered on the graphic card itself. What this means is that the necessary bus speed DROPPED by a great deal. You can run most modern games very nicely on an NVidia PCI card. As long as your textures fit in the RAM of the card, you probably won't even be able to tell it's PCI instead of AGP.
The other thing that AGP promised was texturing out of main RAM, but the AGP bus (not to mention the RAM in the average PC) is nowhere near fast enough to do that.
So it's that last thing that PCI Express may be good for. Assuming that it can still do texturing out of main RAM, it's possible that that idea might finally start working. But that's pretty much the only reason it would be very interesting, at the moment; in a mostly-3d world, graphics are fine on the regular old PCI bus.
One really interesting snippet from the lwn.net writeup:
We are informed that participants in the Linux industry have attempted to influence participants in the markets in which we sell our products to reduce or eliminate the amount of our products and services that they purchase. They have been somewhat successful in those efforts and will likely continue.
I found that more interesting than anything else mentioned so far.
(lwn.net, btw, rocks; one of the best sources for Linux news. They need subscribers. [hint hint] And I'm not affiliated, beyond having a subscription.)
I'm confused about why it takes such a major act of governmental will to override the FCC. They aren't elected, and should be subservient to Congress. Congress is the only body that can make law; why is the FCC being granted that power so strongly?
In essence, the FCC, part of the executive branch, is being given equal status to Congress. To override the FCC, Congress has to pass a new law (which the President has threatened to veto). Congress would then have to override that veto.... requiring a supermajority to regain *their own lawmaking power*.
Also, remember that the plane is a narrow metal tube, which is likely to cause significant increase in radiation levels. A good chunk of the radiation, rather than dissipating harmlessly in every direction, is going to be bouncing around the inside of that plane multiple times before it finds an exit or is absorbed.
Multiply that by fifty passengers with electronic devices, and you could have a very definite problem.
That kind of problem probably doesn't happen very often (ie, two or three passengers are sitting in such a way that their signals reinforce each other at a key point in the aircraft), which would probably explain why the reports of problems aren't all that numerous.
Another poster was pointing out that "only 100 flights reported problems out of all the flights worldwide".... we don't actually know that. It sounded like it was 100 incidents on flights based in Australia, which would be a much smaller sample. They only have like 20 million people in the whole country.
On the whole, airplanes have a very good record, which is mostly due to their overwhelming focus on safety. It would not surprise me in the least to hear that the FAA had banned electronics on airplanes from *just one* proven case of a life-threatening incident from running a Game Boy on an aircraft. This would really suck.
I was imagining that the airlines could start some sort of testing agency, despite the fact that they don't have much money, because not being allowed to use portable electronics would make their product even less appealing than it is now. But I just can't see how it would work properly.... even a small change to a product could invalidate the testing. A Thinkpad model 2621 might be fine, but a 2641 with a wireless antenna in the screen might not be. (just making up numbers.) How on earth are security personnel, even with a list, going to be able to figure this out in a accurate, speedy way? I don't think I could do it, and I worked as a computer tech for a couple of years. And I think people would lie like mad to be allowed to use their gadgets. ("oh, this is exactly the same as a 2621, it's fine.")
Are we willing, as a society, to take on a slightly higher level of risk to avoid boredom in-flight? Personally, I am. Even if I add 50% to the very low likelihood of crashing, it's still an extremely small number, and well within my personal levels of acceptable risk.
In essence, we are blaming users for things that aren't their fault.
The article talks about the need to install anti-virus software, and keep up on patches, and to read the fine print in click-through licenses to prevent spyware from being installed. All of these things need to be done to operate a computer safely, true.
But why the hell are they required? We are giving users HORRIBLE software that is prone to constant infection. Some companies are taking advantage of click-through licensing to hijack people's computers. And we're blaming USERS for not doing the right things?
That would be like making cars that exploded if you ran them at exactly 62mph for more than 12 continuous minutes, with brake systems on the outside of the car where anyone could walk by, flip a switch, and disable them, as well as aftermarket accessories that forced cars to drive on particular roads at particular times.... and blaming the drivers when cars blow up, can't brake, or cause traffic jams on certain roads.
People mostly just want to do email and read the web. We should be providing them software that does this with absolute security.
Slashdot Mirror
I have little doubt that you would be annoyed if the Pledge said "Under Allah". Whether or not the majority happened to believe in something when the country was founded is immaterial; it is the minority that needs protection, not the majority.
Considering that "Under Allah" would undoubtedly annoy many, the most cursory examination would indicate that "Under God" would annoy at least some Muslims. And, since the Pledge is about as official a bit of government communication as could possibly exist, it has to have some chilling effect on those who do not believe in God, an implication that they are second-class citizens.
Additionally; I am unsure of this, but as a condition for naturalization, isn't a new citizen required to cite the Pledge of Allegiance? That would also seem to violate the establishment clause. If this is true, aren't we in essence forcing a new citizen to recognize the existence of God?
The phrase "Under God" was added in the 1950s. It was not a part of the original conception; it is likely that the Founding Fathers would have been horrified at the thought. And It is quite clear just exactly why "under God" was added: it is there to recognize, venerate, and give thanks to a supreme being, one of a particular class of religious thought.
Since "Under Allah" wouldn't work in the Pledge, "under God' doesn't belong either. It really is that simple.
We are not a nation under God, we are a nation under the Constitution, and all religions are welcome here.
That's good to hear. I was specifically criticizing the Windows implementation, which I despise. I haven't used 9.2 yet, so I can't exactly criticize it. :-)
This is from memory without double-checking, so don't treat this as 100% factual without double-checking it.
Mandrake has been having financial trouble for some time, but they've been steadily improving. Basically, they wasted tons of money going after stupid stuff that the venture-capital type management wanted to do (mostly education), instead of spending their money where they actually made revenue.
They declared bankruptcy late last year or early this year. With their debt payments suspended, they're "cash-flow positive" at the moment, and are hoping to get back to full normal operations (ie, paying down debt) by the end of the year.
[end bad memory warning area]
I'm a Silver member of their club and plan to continue my membership. I like their distro very much and feel it's worth supporting. I know that the Club is doing fairly well by them and they're pleased with the results, but the more members the better... it's not that expensive, gives you some nice little perks, and helps make sure a really good distro keeps evolving. I think of it like my local NPR station and like the EFF.. a small contribution by a lot of people can really make a difference.
Your confusion may have arisen because of all the ill-wishing the Slashdot crowd was spouting when they were really hurting for money... "I hope they die, they deserve to", that kind of garbage. Fortunately, crap spewage and reality did not intersect.
disclaimer: I have no connection to Mandrake other than my membership and a couple of installations of their distro.
Personally, I find that "feature" to be EXTREMELY annoying.... if you do it too fast, it's a double click, but if you do it slower than you need to, it's wasted time.
"Click" and "double click" are ok metaphors, but "two slow clicks" is horrid, mostly because "slow" is so ill-defined. A very poor interface mechanism, IMO.
I've been using Mandrake a long time, but I didn't get "serious" about it until the 8.X series.
I recall 8.0 and 8.1 as being a little flaky, but 8.2 as being very solid. I had TONS of problems when I upgraded to 9.0... not because of the distro itself, but because everything ELSE broke. They did some upgrade ( to glibc maybe?) that broke almost all the RPMs I tried from earlier versions. I was able to get around it in many cases by rebuilding source RPMS, but it was still painful. I should have waited longer.
I'm presently on 9.1, and am happy with it. I use the High Performance Liquid theme (you can get it in the Penguin Liberation Front rpms) on KDE. I get an occasional Konqueror segfault, probably about twice a week, but other than that it's been very stable. I'm using it on a P4 1.8, Intel chipset, GF4MX440, Nvidia binary drivers.
My home Linux machine has had occasional lockups of X (machine still up, X frozen), which seem related to my switchbox. I am of the belief that the lockups aren't Mandrake's fault...I'm pretty sure it's the Nvidia drivers. I haven't had the problems with the most recent version, but I also haven't been using the machine very much for the last few weeks, so I'm not sure it's fixed yet. That machine is an Athlon 1900+ on an Asus A7V333, GeForce Ti4200.
On the whole, I'd call it about equivalently stable to XP, maybe a little less. The kernel itself, of course, is almost uncrashable, but X is a little more fragile than I'd like.
What really killed Commodore was their mishandling of the Amiga. They had possibly the finest system engineering team ever assembled, and treated them like dirt; they had neither any idea what genius of that caliber was worth, nor just what that system could have become if left in the same super-capable hands.
I have often said that if Apple had owned the Amiga technology, we'd all be using descendants of that machine instead of the PC. It was that good. Put every other machine in its era absolutely to shame. They sold several million Amigas based almost purely on word of mouth; with a good management team and advertising behind it, it would have taken over practically everything.
C= didn't just shoot themselves in the foot with the Amiga, they shot themselves in the head. Several times. Revenue from the 64 made the Amiga buyout and manufacture possible, but it was the Amiga itself that was the make or break for the company. Break, as it turned out.
This is journalism of the caliber of talk radio; leaving out the important part "it is legal" to highlight the dreaded imminent threat. Probably helped get it chosen as a story, since putting one to the Feds is always popular in the geek crowd, but it's still irresponsible.
But being "just as easy as windows" isn't good enough. It needs to be easier than Windows to really take off. "Just as good as but free" will definitely sell people on Linux, but "easier, more powerful, and free" will move it by the boatload.
We need to also remember that OS choice is a social phenomenon as much as a technical one. People have been taught for 10+ years that "PCs run Windows". It's going to take some time to overcome that conditioning, no matter how good the alternatives are.
If Joe Badguy sends out 10,000 copies of Quicken, then at least some of those people are going to upgrade Quicken by paying for it in later revisions. This may result in MORE revenue, not less. In addition, those 10,000 people A) have not given Microsoft any money, B) are using and learning to like Intuit's product, and C) may recommend it to their friends.
If copying becomes really widespread, it could become a problem... if it becomes so accepted that common sense says "it's stupid to pay for Quicken when you can just copy it", then perhaps Intuit wouldn't be able to make money.
Note, however, that this HASN'T HAPPENED YET. Intuit has made HUGE amounts of money with very little copy protection. They trusted their customers, focused on them, solved their problems, and in exchange they made boatloads of cash.
But instead of focusing on their customers, now they are focusing on "all those deadbeats". Intuit is trying to strong-arm pirates into buying the software they copied. This hurts the usability of the software for people that DID and DO pay them money, in an effort to extort money from people who HAVEN'T. Those people who haven't paid for it probably have reasons for it, and while Intuit can undoubtedly coerce some of them into paying, they inconvenience their existing customers enough that they lose more than they gain. They are thinking purely in terms of "deadbeats" and not at ALL about marketing and word of mouth and network effects.
Trying to force someone into a business relationship is not a long-term route to profitability. (although the Mafia may disagree with me there. :-) )
Microsoft can get away with this stuff because, for most people, there are no alternatives. Most people think they have to use Windows; they probably don't even realize there's any real choice. That's how Microsoft can get noxious ideas like Product Activation through.
Microsoft, being a monopoly, IS in a position to force people to do things they don't want to. Intuit is not, and they may have just learned that lesson.
this is what I sent them:
I'm a lost customer to you; it is unlikely that I would ever buy another Intuit product. I've been a paying customer of yours for years, both at retail and directly from you, but this year I bought TaxCut and am perfectly happy with it. I am still shocked that you think that treating your customers like thieves is a good business practice. I see you have an "unlock code" now for Quicken too, from the front page.
You're supposed to be in the business of SOLVING customers' problems, not CREATING them.
Casual copying probably helps you; it gets people hooked on your product, and it turns even people who didn't pay for it into advertising. If everyone is using a product, some percentage will always buy new versions, and if you provide a good quality, problem-free product, that percentage will most likely climb steadily over time. Mindshare really matters. If I ask five friends "what's the best tax program", and they all tell me Turbotax, that's going to have a much greater impact on me than three Turbotax votes and two TaxCut votes.
As you just discovered from watching your revenue plummet, getting 50% paid customers from three times the installed base nets you more money.
A personal example: I used my parents' copy of TurboTax for a few years back in the early 90s. Starting sometime around 1995, I converted into a paying customer myself. My "theft' of your product *turned me into a customer*. I bought the product for years. This year you lost me, and you probably won't get me back.
It is casual copying that got Microsoft where they are today. That casual copying, in fact, turned them into the one of the most profitable companies in the world. Don't think their use of DRM is helping them; it is causing fragmentation in their customer base. People are now actively resistant to upgrading to the most recent versions of things, and this is providing competitors like Linux an opportunity to expand at their expense. You can't see it as much yet, because Microsoft is a monopoly, but DRM is hurting them. Think of it as cracks in the foundation.
Intuit isn't a monopoly and has competitors to all of its products. Your decision to try to limit casual copying and to treat customers like thieves hurt you badly. I'm a lost customer. I know there are lots of others. And I'm not at all sure you're going to get many of us back.
TaxCut is really quite good, and I imagine I'll be checking out Money soon. And I'll most likely make a dreaded casual copy of Money to do it. Microsoft may not know it, but they want me to do that; any chance of converting me from a "deadbeat" into a paying customer is better than the zero money they're making from me now.
And if a friend asks me "hey, what's the best software to track my checkbook?", and I answer "Microsoft Money", well.... suddenly the "Quicken is best" message is no longer unanimous and maybe Microsoft makes a sale.
The author of this article mentions only three real differences between Linux and Windows; Linux users don't run as root, Linux email programs don't automatically run attachments, and Linux has more than two main programs to read mail.
Now, I have believed for a long time that Linux is less susceptible to virii than Windows. I stated as much in a fairly long post in a Lindows discussion, which basically boiled down to "don't run as root!". I got some great replies to that post which shifted my thinking quite a bit.
Now, I still stick to my original assertion that running as root is bad. The single biggest reason is rootkits; if you are compromised while running as root (or if a process running as root is exploited), it's possible for the exploit software to hide itself from you almost completely. A good rootkit can make itself extremely difficult to detect, and no user-space virus can do that without a further exploit.
That said, it's perfectly possible to have userspace virii. They can't hide themselves as well, but they can still run, and can still propagate. And while they can't take down *the whole machine*, they can certainly wipe out your home directory, and for most users, it's almost the same thing.
We're used to looking at this as sysadmins; as long as the system stays intact, we can restore data. But for home users, most of whom don't back things up, losing the home directory is about the same as losing the machine. They may save a couple of hours on the reinstall, but compared to the weeks or months to recreate data, big deal. We can be all smug about "the system wasn't compromised!", but the user lost everything, and from his or her perspective, that's all that matters.
So the "no running as root" idea is useful for disinfection, but doesn't help at all for data loss, and doesn't stop propagation in most cases. Outbound connections can be made just fine from userspace.
The more secure email programs are good, but I suspect that may be a temporary advantage. There's a natural tendency toward monopoly in system software; over time, groups of people tend to converge on similar software. My office, for instance, has three main Linux users; each of us runs a different distro, but we have all independently chosen Evolution for email. Less security means more ease-of-use, and more ease-of-use means more uptake among the end-user population, so over time, it seems likely that Linux email programs will become less and less secure. The Linux email program advantage is very high, but over time, I believe it will be much less so.
Additionally, there's more and more tendency for everyone to use the same fundamental object libraries, like, say, OpenSSL. This tendency to write against the same core libraries means that whole classes of programs may suddenly become exploitable at once.
It's easy for us to get complacent, but our armor here is pretty thin. The lack of a Linux monopoly really may be much of the reason we don't see many virii.
After all, how many email virii do you know that spread via Eudora on Windows?
As far as I know, it's an asteroid until it hits atmosphere, a meteor until it hits the ground, and a meteorite after that.
In other words, there has *never* been an asteroid strike on earth.
</pedant>
(and I can just imagine a pair of dinosaurs arguing about what to call the really, really big rock in the sky. Be a great Far Side.
Hey, out of interest, if you encode with the LAME presets (--alt-preset standard, extreme, and insane), can you still hear the MP3ness on a rig that good?
I listen on (good) headphones and I can sometimes tell with --aps, but I just can't hear the difference with --ape. (or, presumably, with --api, but I haven't bothered checking.)
Just curious.
All networks are fundamentally shared bandwidth. It just depends on where the bandwidth is shared.
The DSL circuit itself isn't shared, and the full b/w allocation gets to the router, but starting there and going upstream, ALL connections are shared. T1, cable, DSL -- they are all oversubscribed; the ISP sells more bandwidth than they buy. As far as I know, this is universal, except possibly for backbone providers, which probably have to overprovision in case of failure.
The amount of oversubscription matters more than where it is. If your neighborhood is trying to run 20 megabits/second over a 2 megabit/second pipe, things are going to suck. But if a DSL provider is trying to aggregate 500 "unshared" connections through a T3 (45 megabits), it's also going to suck. Distance isn't completely unrelated, though; if your neighborhood is oversubscribed, EVERYTHING will suck. If the connection to just, say, AT&T is overtaxed, then only sites behind that connection will be bad.
That said, in general, cablemodem is oversubscribed worse than is DSL, and both are far worse than T1-class (and higher) connections. That's why they're cheaper.
The fundamental problem is, of course, that their actual bandwidth needs are outstripping their flawed projections, and instead of buying more bandwidth (which costs a lot of money), they're trying to force usage back into their models. This is a cap on usage, whether or not they will admit it.
Also, the AUPs on many of these networks are horrible; if you read them, for the most part you are not allowed to run a server of any kind. Period. Some of them even say you can't use VPN software. At least one provider was going so far as to block protocols and ports used for VPN, unless you upgraded to a "business class" account. (ie, if you actually want to USE your circuit, well, that costs extra.) Moral: read your AUPs carefully.
I'm currently a Speakeasy customer and very happy with them. Their AUP is quite liberal; it's basically "don't do illegal things with our network, don't DOS people, don't spam, and don't abuse our staff." Their fundamental business model is sustainable; it's built on the assumption that you will use your bandwidth. As long as you're using it legally, I don't think they care how much traffic you send. The network is good, with no visible packet loss and reasonable latency, although I wish the latency was a trifle better. (I got spoiled on PacBell's ATM network.) It's still lightyears ahead of the Comcast connection my sister has.
When I was in California, PacBell was very similar. I was one of the very first people to sign up for DSL when the price got reasonable, so their AUP may have changed, but as of several years ago it was very similar to Speakeasy's. At the time, their network was incredibly good, with super-low latency and great connectivity almost everywhere. I've been gone three years, so I don't know if it's still as good now, but when I left they were stellar.
Many cable companies are, in my opinion, involved in a bait-and-switch with their networks. You get higher burst download speeds, but usually the latency is much worse, and they can terminate you for daring to actually use the service you're paying for. You'll pay more, but I believe DSL from a good provider is a better bet.
Well, I thought about mentioning Windows accelerators in the VLB timeframe, but it wasn't directly related. AGP was invented, more than anything else, for games, and as far as I know, most games don't use the Windows graphics-acceleration functions. Back in the day, they were all written in DOS, and nowadays, they all use DirectX. Admittedly, some Windows-based strategy games use standard Windows rendering functions, but those games aren't the type that require fast graphics anyway.
:-)
You say that texture loading is gawdawful slow, but I disagree. It's not real-time, but you can load a 128MB texture buffer in about a second with PCI. Standard AGP is only twice that fast, or about a half-second... it's just not that much of a difference. Games that do a lot of texture swapping are going to show a hit, but as far as I know, EVERY major game out there is designed around the idea of fitting textures into memory and only swapping them out occasionally.
I suppose I should also cop to the fact that the HD is sometimes on the same bus as a PCI video card, so the available bandwidth might be cut in half (have to copy from the HD to the PCI card), whereas AGP is always on a separate bus. That would probably double the load time again, up to two seconds. But I think that's probably worst-case, assuming a decent PCI implementation. Pretty acceptable, IMO.
Conclusion: AGP, for most uses, is overrated. PCI-X, for the current games and video apps, isn't really necessary. But it might allow a major shift (finally) in how texturing works. And, as you reminded me, the ability to read from video memory at full speed might be really useful.
This prompts the thought that the GPUs are just monstrously powerful, far far faster than general-purpose CPUs. If the PC had a way of sending a program to the GPU and then reading the results back at full speed, it might be possible to jack those distributed.net results WAY up.
Athlon chipsets sucked rocks for a long time, and were really unstable. But VIA finally got their act together, I think with the KT133A.
AFAIK, other than stomping on occasional driver bugs, Athlon chips have been pretty excellent ever since. I have an Athlon 1900+ on an ASUS A7V333 that's rock solid, and a new Athlon 2500+ on an Nforce2 board that's not quite as solid, but which is still pretty good.
I'd like to see some improvements on the NForce2 chip stability. It's not all the way there yet, in my opinion. But the VIA chipsets are extremely solid.
It will if it's based on Linux. NAT is part of the routing process. You can't do NAT without routing, but doing NAT in *no way* stops regular routing from working.
In other words, if I say "translate 192.168.0.10 to address 1.2.3.4", and then I get packets at the external interface headed to 192.168.0.10, they will be routed. Further, the replies will come back without being NATted, because the kernel is doing stateful inspection, and it will know that the reply to a "normal" connection shouldn't be tampered with.
I use this functionality all the time, so I know exactly what I'm talking about.
Oops, I made a math mistake here. AGP 8x should allow an absolute maximum of about 275 frames/second of 1600x1200x32 bit color. It would only do a thousand frames if it was in 8 bit color. Forgot to divide by 4. Sorry.
What he's saying, and perhaps isn't doing a very good job, is that NAT and firewalling are separate entities.
Most consumer-level NAT devices use 192.168.0.0 or 1.0/24 as their NAT network. What mindstrm is trying to say is that your provider could easily add a route for 192.168.0.0/24 aimed at your DSL router. A machine doing NAT-only translation will happily forward packets addressed directly to its "private" network. It will do NAT if the rules are triggered, but if they aren't, it *still routes the packets*, so Joe Evil Employee can access your machines freely. He can ping, nmap, and hack remotely any machine he likes in your "private" network. To him, you look exactly the same as any other un-firewalled machine in the world. There's nothing magical about 192.168.0.0; it's set aside for private use, and is used in probably hundreds of thousands of installations, but if the ISP routes to your modem, he's temporarily saying "ok, you are the real 192.168.0.0 as far as I'm concerned"... and he's in.
In real life, of course, that mostly won't work, because NAT devices *also* do firewalling, and will refuse to route the inbound packets. What mindstrm is trying to point out is that it is really the firewalling that gives you the security, not the NAT.
NAT does protect some against people coming from the "rest of world", those who don't have access to all the routers between their machine and yours. But firewalling provides security that's both a lot better and a lot more flexible.
If NAT goes away, it won't cause any appreciable security loss.
Actually, AGP was really developed to solve a problem that went away on its own; 2-D graphics.
The big bottleneck on PC graphics for years and years was the bus speed. When you are doing 2-D graphics, in essence you have to copy your graphic data out, frame by frame, to the display memory. The system bus was always the bottleneck here. To animate a 320x200 screen at 30 frames per second, you have to push out about 2 megabytes per second. 640x480 is four times that; 1024x768 takes about 24 megs per second. These are all in 1-byte pixels, or 256 colors; to do this in 32-bit color you'd have to push 4 times as much again, or about 95 megabytes/second for that 1024x768 screen. The numbers go up really, really fast as you get to higher resolutions.
So the big thing with PC graphics, for a long time, was increasing the bus speed. The original 8-bit ISA bus can push 4.77 megabytes per second, which wasn't able to animate even 320x200 because that same bus also had to do all memory access and other I/O. The 16-bit ISA slots could do 16 megabytes/second (8mhz x 16 bits.): with the system overhead, you could definitely do 320x200, and you could probably do 640x400 with very, very clever programming, but it would be iffy.
It was about then that graphics really started getting important, and VESA Local Bus was invented to extend the spec; I believe that was a 32-bit bus running at 33, 40, or 50 megahertz, depending on the CPU it was attached to. (VLB was a very simple design that was "close" to the processor electrically, and ran off its front-side bus.) As per the calculations above, you could put out a lot of data to a VLBus card, enough to render a 1024x768 screen. PCI was invented at about the same time, and while it wasn't as fast as a 50Mhz VLB card, it was flexible enough that it eventually supplanted VLBus, which died quietly.
PCI can do 133 megabytes/second (33 mhz x 32 bits wide). so 1024x768 is about the hard limit there. I'm not sure if they were pulling main memory off onto its own bus yet (as they do with modern machines), so if the video card was still competing with main memory, true 1024x768 at 30 frames per second would have been very difficult. With a modern machine, it would be no problem, as long as the computer wasn't trying to hit any of the other PCI cards too hard.
Well, Intel could see the writing on the wall, and came up with AGP. AGP 1x runs at twice the speed of PCI and hence has twice the bandwidth; the later 2x, 4x, and 8x specs doubled the speed each time. An AGP8x board can shovel out about 2.1 gigabytes/second, which is enough to comfortably animate 1600x1200 at several hundred frames per second. (perhaps as much as a thousand with, again, clever programming.)
But while they were busily solving the bandwidth problem, it went away. All this speed isn't really being used anymore. Over the last few years, PCs have switched away from using 2-D graphics to using 3-D graphics for most games. And 3-D is represented very differently; it is sent to the graphics card as a series of textures and triangles, and rendered on the graphic card itself. What this means is that the necessary bus speed DROPPED by a great deal. You can run most modern games very nicely on an NVidia PCI card. As long as your textures fit in the RAM of the card, you probably won't even be able to tell it's PCI instead of AGP.
The other thing that AGP promised was texturing out of main RAM, but the AGP bus (not to mention the RAM in the average PC) is nowhere near fast enough to do that.
So it's that last thing that PCI Express may be good for. Assuming that it can still do texturing out of main RAM, it's possible that that idea might finally start working. But that's pretty much the only reason it would be very interesting, at the moment; in a mostly-3d world, graphics are fine on the regular old PCI bus.
But I bet the demo coders will love it.
I found that more interesting than anything else mentioned so far.
(lwn.net, btw, rocks; one of the best sources for Linux news. They need subscribers. [hint hint] And I'm not affiliated, beyond having a subscription.)
I'm confused about why it takes such a major act of governmental will to override the FCC. They aren't elected, and should be subservient to Congress. Congress is the only body that can make law; why is the FCC being granted that power so strongly?
In essence, the FCC, part of the executive branch, is being given equal status to Congress. To override the FCC, Congress has to pass a new law (which the President has threatened to veto). Congress would then have to override that veto.... requiring a supermajority to regain *their own lawmaking power*.
Something is REALLY messed up here.
Also, remember that the plane is a narrow metal tube, which is likely to cause significant increase in radiation levels. A good chunk of the radiation, rather than dissipating harmlessly in every direction, is going to be bouncing around the inside of that plane multiple times before it finds an exit or is absorbed.
Multiply that by fifty passengers with electronic devices, and you could have a very definite problem.
That kind of problem probably doesn't happen very often (ie, two or three passengers are sitting in such a way that their signals reinforce each other at a key point in the aircraft), which would probably explain why the reports of problems aren't all that numerous.
Another poster was pointing out that "only 100 flights reported problems out of all the flights worldwide".... we don't actually know that. It sounded like it was 100 incidents on flights based in Australia, which would be a much smaller sample. They only have like 20 million people in the whole country.
On the whole, airplanes have a very good record, which is mostly due to their overwhelming focus on safety. It would not surprise me in the least to hear that the FAA had banned electronics on airplanes from *just one* proven case of a life-threatening incident from running a Game Boy on an aircraft. This would really suck.
I was imagining that the airlines could start some sort of testing agency, despite the fact that they don't have much money, because not being allowed to use portable electronics would make their product even less appealing than it is now. But I just can't see how it would work properly.... even a small change to a product could invalidate the testing. A Thinkpad model 2621 might be fine, but a 2641 with a wireless antenna in the screen might not be. (just making up numbers.) How on earth are security personnel, even with a list, going to be able to figure this out in a accurate, speedy way? I don't think I could do it, and I worked as a computer tech for a couple of years. And I think people would lie like mad to be allowed to use their gadgets. ("oh, this is exactly the same as a 2621, it's fine.")
Are we willing, as a society, to take on a slightly higher level of risk to avoid boredom in-flight? Personally, I am. Even if I add 50% to the very low likelihood of crashing, it's still an extremely small number, and well within my personal levels of acceptable risk.
In essence, we are blaming users for things that aren't their fault.
The article talks about the need to install anti-virus software, and keep up on patches, and to read the fine print in click-through licenses to prevent spyware from being installed. All of these things need to be done to operate a computer safely, true.
But why the hell are they required? We are giving users HORRIBLE software that is prone to constant infection. Some companies are taking advantage of click-through licensing to hijack people's computers. And we're blaming USERS for not doing the right things?
That would be like making cars that exploded if you ran them at exactly 62mph for more than 12 continuous minutes, with brake systems on the outside of the car where anyone could walk by, flip a switch, and disable them, as well as aftermarket accessories that forced cars to drive on particular roads at particular times.... and blaming the drivers when cars blow up, can't brake, or cause traffic jams on certain roads.
People mostly just want to do email and read the web. We should be providing them software that does this with absolute security.
We are blaming users for faulty software.