Indeed; I've always wondered why we don't see more attacks like this. Killing people can be extremely cost-effective; you can shut down tens of millions of dollars worth of economy for a few thousand dollars.
Arguably, the only way to prevent that is precisely these sorts of draconian measures, designed to provide every single opportunity to catch them. That would effectively destroy the US, not just in spirit but literally: we achieve what we have achieved by encouraging people to invent and invest.
So I wonder why they haven't done it yet. They've got plenty of opportunity.
I've always been leery of net neutrality legislation, not because I'm opposed to the concept but because I don't expect Congress to define it correctly. I'd actually rather see it as an RFP amending the IP standard. And there are perhaps things to be accomplished by violating neutrality that would make life better rather than worse.
But the nightmare scenario has always been there: since the number of ISPs available to most consumers are limited, that monopoly power could be used to force choices on consumers. The market could be used to reward innovative ideas that require breaking net neutrality, but monopolies break markets.
I've never really understood what the telcos expect to get from Google on this. When Google starts getting a thousand extortion bills from a thousand separate carriers, there's no way they can track which ones are valid. (Am I going to start Bob's ISP and send Google a bill for it?) I expect Google to toss them all into the trash.
And if they find that consumers are unable to reach them, I sure hope their lawyers can convince the courts that this is antitrust behavior. I trust the courts very slightly more than I trust Congress.
It phones home more than once because hard drives can be cloned. If the thing only poked its head up once, pirates would run a single legal box until WGA verified itself, then make copies of that disk and sell them. So it has to check every time if the brain wakes up in a "new body".
There's far less cause for it to phone home if it wakes up in the same old body. There's some complexity going on if you replace a hardware component; defining "a computer" is tricky.
Yeah, there are voluntary evacuations as near as 5 miles from my house, but my house is pretty far above the water table. I, and more importantly my tomatoes, are safe (though one of them fell over with the rain pounding on it.) But I think the rain has drowned my watermelon.
If you've never had tomatoes grown in a garden, picked fully ripe and carried 30 feet rather than 3,000 miles, you're really missing out. Store-bought tomatoes are bred for durability, not flavor, and they're generally picked under-ripe. They redden, but not really ripen, in shipping. Garden tomatoes come in far more varieties, and taste far better.
Farm stand tomatoes are also pretty good, though they tend to be hardier, less flavorful varieties. But if you happen to find a farmer's market with, say, some Brandywines, you can get an inkling of how good a real tomato is.
So, BadAnalogyGuy (945258), if you want to try a real tomato, drop by my place. You'll have to come to Maryland, because you can't ship a real tomato. And you probably shouldn't, because once you do, you'll never be happy with store-bought red baseballs instead.
Lettuce is lettuce; get it from the store. Zucchini ships fine. But if you want a tomato, a real tomato, a garden is the only way to go. They're seasonal, and temperamental, and prone to a truly stunning variety of bugs, slugs, and blights, but taste one and you'll believe it's worth it.
Hey, big time. I'm a programmer, and like economics, programming is a very young science. There's a lot to know, but not so much that you have to go to school for years to get the important parts. (As opposed to, say, quantum physics, where there's simply no place for a "talented amateur" to make progress. But try telling that to the thousands of guys who seem to believe that Einstein was pretty smart but not as smart as they are.)
I hire programmers all the time who weren't college-educated, and sometimes people who don't even speak the language we work in. I'll train 'em in the language if they can already code.
Physicists are experts in building and applying mathematical models. They observe a complex system, try to build a model that matches it, and then report the results. EBay bidding is just another system for them to model. In some ways it's more complex than the ones they usually build, because the fundamental units are thinking human beings, but humans often behave rather more predictably than we'd like to think, at least in aggregate.
Ideally this work should be done by an economist, but increasingly economists are thinking of themselves more like physicists. The recently popular book Freakonomics actually shows things moving the other direction: an economist applying the scientific method to areas outside of pure economics, in areas like baby-naming and crime.
But since economics as a science is a relatively young science, there's a lot of work that can be done by people who aren't experts in the field but who do bring expertise from their own. Like physicists and their mathematical models.
Ya know, I'd really like to see that. Snipers who outbid me really are willing to pay more than I am, so they might as well have it. Nonetheless, they piss me off, because it means I have to wait the full length of the auction to be out-bid, preventing me from going off and bidding on some other item.
Sniping works when people don't actually know what something is worth. Every time you make a bid, you literally raise the price of the item: you declare that it's worth at least this much to you. And if nobody knows what the price is, they know that it's worth a little bit more when you raise the price.
By sniping, they lower the final price by hiding the information of what they're willing to bid. The auto-bidding feature is supposed to do that for you, but information still leaks out: it tells you the value of the second-higest bid, which is probably close to that of the highest bid. It's a good proxy of the actual price, and you're raising that.
Personally, I don't get too bugged about the snipers. I know what it's worth to me and let the auto-bidders handle it. If I get sniped, it means they were willing to pay more than I was. The seller gets screwed a bit, because the sniper was hiding their willingness to pay more (and therefore probably more than they ended up paying.) But the only thing being done to me as a bidder is that I got my hopes up.
So I don't get my hopes up, and I don't bid on eBay for anything I really, really want. I use eBay not for its auction, but for its flea-market: the ability to get stuff I can't get elsewhere. I usually just pay the "buy it now" price.
In this case I'm talking about an actual monetary cost. There are some spam-fighting schemes that call for purchased "stamps" on every email you send. If your email doesn't have one, I won't accept it. Spam works only because the mail is zero-cost. Even if it cost $.00001 per email, it wouldn't be cost effective any more.
There are a number of problems with the plan. One is that you can't right now usefully charge $.00001 for anything; the overhead will kill you. There are various schemes to work around that, but none are ready for prime time.
There's also the mass-email problem: I can't maintain a joke-of-the-day email list with a million people. Switching to RSS for that would help.
There's also the fact that much spam comes from email zombies: hacked computers that send spam. If those zombies can't get access to your account, then they can send all the spam they like; the lack of a stamp means it'll be dropped. But if they do hack somebody's account sufficiently to buy stamps with it, suddenly that's real money you're losing when you get infected.
So this is all aside from bandwidth costs. There are issues associated with that, too, but as a sister post explains, those are largely under control.
The nice thing about email newsletters is that they look just like your other communcations; you use one tool to manage them.
But email is a two-way communication; RSS is really primarily one-way. That makes for a technological difference: with RSS, because it's fetch, you know you're not getting spam. Email is push, and so it's hard to distinguish newsletters from spam. And it's one more site to give your email address to, meaning one more opportunity for spammers to steal/buy it.
Getting newsletters out of the email loop will make it easier to support some anti-spam technologies. Newsletters are one of the downfalls of pay-to-send schemes, because a free newsletter emailed to a million people at $.00001 turns into real money.
I like integrating RSS into the email stream. Some email apps already support RSS, and I would like to see them show up in just a single queue of "stuff to read".
I have no idea what the long-term effects are. And God knows the dry ice is more fun to play with. But if you want fog right in a particular place at a particular time, nothing beats a fog machine.
I do theatrical work, not concert work. Concerts usually just crank up the fog and let it go, the more fog the better. In a play there are scenes with fog and scenes without fog, and it's really nice to be able to control that with a guy in the light booth rather than dumping pellets into a bucket. Not to mention the ability to pump fog wherever you want it.
For stage shows, fog machines are far more controllable and produce better results than dropping dry ice in water. They use "fog juice" rather than dry ice.
Though sometimes you'll use dry ice to cool the resulting fog. The hot fog gives you a smoky, atmospheric effect. If you want ground-hugging fog, you've got to cool it down, and dry ice is a pretty good way to chill it quickly.
Possibly just because there's money to be made. That's why they have an Xbox: if they can turn a profit on it, they will.
But in the case of the iPod, they want to compete against it for the same reason Apple created it: to introduce people to the Macintosh. Apple doesn't turn much in the way of profit on iTMS; it's just there to drive people to the iPod. The iPod does turn a profit, a pretty good one, but more importantly it gets people used to the idea that Apple products just work.
The iPod has astonishing market share despite the presence of cheaper, better-equipped alternatives. If people jump from Windows to Mac on the strength of that, it jepoardizes Microsoft's market share. At this point they depend vigorously on being the default OS choice. Erode their market share a little, and you open the door to eroding it a lot, as people no longer have to buy a Wintel box just to keep on the same page with their friends.
There's also the fact that a big company can never stand still. Just producing revenue isn't enough; they have to produce more revenue. One way to do that is to diversify, especially if you can diversify and still leverage your products in other areas. MS can do that big time.
For example, if they have a new, stronger DRM scheme (based, say, on Palladium), they may be able to get record companies to give them a price break, or even sign up those companies who don't trust Apple's FairPlay to protect their property.
MS can leverage their OS control (to give their device a performance hack that Apple can't get). Maybe they can leverage the Xbox, perhaps a plug on the side of an Xbox for their music player, or being able to build a handheld game device leveraging both the Xbox and music player platforms.
I don't know what they've gamed out, but basically, MS will try all of it. The downside, of course, is losing focus: it's usually better to make 1 good product than 10 shoddy ones. That's less about technology and more about management. MS thinks it has good management. On that, we'll have to see: the slips in the Vista schedule don't speak well to that.
Your observation answers my key question: why would they risk having their law struck down by trying it out on a site which is probably not in violation? (I haven't read the text of the law, so I can't say for certain.)
Answer: because to the Indian tribe casinos, linking to gambling site causes at least as much harm as gambling on one. Anti-gambling activists would probably care more about the actual gambling sites, but those who stand to make money from gambling demand that they limit the exposure as well.
Google Earth has a number of advantages over Google Maps:
* Because it's a native app rather than Web 2.0, it's snappier.
* There's an interface for making lists of points. For example, it's easy to create "this is my house" and "this is my office" places. That makes getting directions more convenient.
* Navigating in cities, the ability to see the buildings in 3D and get a ground-level perspective can be really useful in driving. It can be hard to figure out exactly what some directions mean just from the description.
* Printing directions is more effective, because it's not limited to the 72DPI resolution you get on a screen. (It still doesn't do as good a job as I'd like; I'm hoping version 4 does better.)
Those are things you actually want, in addition to the gee-whiz stuff. By all means continue to use Google Maps if you don't want to bother downloading something. But Google Earth isn't just fancy-but-irrelevant features.
I've worked with it on exactly one relatively simple spreadsheet, and I found numerous bugs just in that time:
* Some important formatting disappeared. Not just column widths, but numerical formatting. The difference between "currency" and "not currency" is very important for the look of the sheet.
* Re-exporting to Excel had a bug: it capitalized the sheet names, but didn't propagate that to formulas. Any formula that referenced another sheet became #VALUE
* Even for the small spreadsheet I was using (a few sheets, dozens-not-thousands of rows and columns), scrolling was very, ver slow.
This fairly simple sheet is what I think of as a canonical app for Google Spreadsheets: not mission critical, not large, not full of database lookups or macros. Maybe those are just beta complaints, but I've got to concur with your verdict: not ready for prime time.
Or at least pretending to sell prescription drugs on the Internet. I can't imagine that any of them actually send out the illegal pharmaceuticals. It's not like they're expecting to maintain a long-term relationship with you.
Maybe I'm wrong. Has anybody ever actually gotten meds from one of these guys?
I was totally with you until I got to the bit about Windows peer networking. If you know the magic button that allows me to get to other computers on the network without a 30-second hang before reporting, "No, I haven't figured out how to get to the computer that's right next to this one", I'd really like to know about it.
I think that protecting the user's own data is sufficient reason to blame this on the app writer, not the OS. Yeah, it's the OS's fault if rootkit-level harm can be done, but I think of that as a whole separate problem. Huge amounts of damage can be done even to the user's sandbox, including disclosure of private information (which isn't the OS's fault, either, if the app is giving its macros access to sockets).
There's plenty of blame to go around, but it points out a general clue: writing secure generalized systems is hard, whether it's an OS or a word processor that thinks it's an OS. Security is everybody's problem, and you have to think about it every time you get input from a user. Limiting the effect of security failure isn't the same as abnegating responsibility to prevent that failure. The more power you give that user, the more responsibility YOU have to ensure that power isn't misused.
Maybe because it tested the browsers a little too hard? I just checked it out with Firefox, and although it's more attractive, important parts of it didn't render correctly.
That could be just Coral's cache of it, however. I was having trouble getting to the live copy.
I don't actually use it for... well, what the hell was it supposed to be for? But my theater troupe uses an online ticketing service, and it's kinda nifty to be able to just scan the bar codes when they present the tickets. And all it took was a trip to Radio Shack and a downloaded driver.
All with you on the crummy page-layout support in OpenOffice. But I don't find that Word is significantly better at it. Maybe a little better, but even after using it for years I find Word's notions of frames and floating images utterly baffling.
Slashdot Mirror
Indeed; I've always wondered why we don't see more attacks like this. Killing people can be extremely cost-effective; you can shut down tens of millions of dollars worth of economy for a few thousand dollars.
Arguably, the only way to prevent that is precisely these sorts of draconian measures, designed to provide every single opportunity to catch them. That would effectively destroy the US, not just in spirit but literally: we achieve what we have achieved by encouraging people to invent and invest.
So I wonder why they haven't done it yet. They've got plenty of opportunity.
I've always been leery of net neutrality legislation, not because I'm opposed to the concept but because I don't expect Congress to define it correctly. I'd actually rather see it as an RFP amending the IP standard. And there are perhaps things to be accomplished by violating neutrality that would make life better rather than worse.
But the nightmare scenario has always been there: since the number of ISPs available to most consumers are limited, that monopoly power could be used to force choices on consumers. The market could be used to reward innovative ideas that require breaking net neutrality, but monopolies break markets.
I've never really understood what the telcos expect to get from Google on this. When Google starts getting a thousand extortion bills from a thousand separate carriers, there's no way they can track which ones are valid. (Am I going to start Bob's ISP and send Google a bill for it?) I expect Google to toss them all into the trash.
And if they find that consumers are unable to reach them, I sure hope their lawyers can convince the courts that this is antitrust behavior. I trust the courts very slightly more than I trust Congress.
It phones home more than once because hard drives can be cloned. If the thing only poked its head up once, pirates would run a single legal box until WGA verified itself, then make copies of that disk and sell them. So it has to check every time if the brain wakes up in a "new body".
There's far less cause for it to phone home if it wakes up in the same old body. There's some complexity going on if you replace a hardware component; defining "a computer" is tricky.
Yeah, there are voluntary evacuations as near as 5 miles from my house, but my house is pretty far above the water table. I, and more importantly my tomatoes, are safe (though one of them fell over with the rain pounding on it.) But I think the rain has drowned my watermelon.
That's true love and home grown tomatoes.
If you've never had tomatoes grown in a garden, picked fully ripe and carried 30 feet rather than 3,000 miles, you're really missing out. Store-bought tomatoes are bred for durability, not flavor, and they're generally picked under-ripe. They redden, but not really ripen, in shipping. Garden tomatoes come in far more varieties, and taste far better.
Farm stand tomatoes are also pretty good, though they tend to be hardier, less flavorful varieties. But if you happen to find a farmer's market with, say, some Brandywines, you can get an inkling of how good a real tomato is.
So, BadAnalogyGuy (945258), if you want to try a real tomato, drop by my place. You'll have to come to Maryland, because you can't ship a real tomato. And you probably shouldn't, because once you do, you'll never be happy with store-bought red baseballs instead.
Lettuce is lettuce; get it from the store. Zucchini ships fine. But if you want a tomato, a real tomato, a garden is the only way to go. They're seasonal, and temperamental, and prone to a truly stunning variety of bugs, slugs, and blights, but taste one and you'll believe it's worth it.
Hey, big time. I'm a programmer, and like economics, programming is a very young science. There's a lot to know, but not so much that you have to go to school for years to get the important parts. (As opposed to, say, quantum physics, where there's simply no place for a "talented amateur" to make progress. But try telling that to the thousands of guys who seem to believe that Einstein was pretty smart but not as smart as they are.)
I hire programmers all the time who weren't college-educated, and sometimes people who don't even speak the language we work in. I'll train 'em in the language if they can already code.
Physicists are experts in building and applying mathematical models. They observe a complex system, try to build a model that matches it, and then report the results. EBay bidding is just another system for them to model. In some ways it's more complex than the ones they usually build, because the fundamental units are thinking human beings, but humans often behave rather more predictably than we'd like to think, at least in aggregate.
Ideally this work should be done by an economist, but increasingly economists are thinking of themselves more like physicists. The recently popular book Freakonomics actually shows things moving the other direction: an economist applying the scientific method to areas outside of pure economics, in areas like baby-naming and crime.
But since economics as a science is a relatively young science, there's a lot of work that can be done by people who aren't experts in the field but who do bring expertise from their own. Like physicists and their mathematical models.
Ya know, I'd really like to see that. Snipers who outbid me really are willing to pay more than I am, so they might as well have it. Nonetheless, they piss me off, because it means I have to wait the full length of the auction to be out-bid, preventing me from going off and bidding on some other item.
Sniping works when people don't actually know what something is worth. Every time you make a bid, you literally raise the price of the item: you declare that it's worth at least this much to you. And if nobody knows what the price is, they know that it's worth a little bit more when you raise the price.
By sniping, they lower the final price by hiding the information of what they're willing to bid. The auto-bidding feature is supposed to do that for you, but information still leaks out: it tells you the value of the second-higest bid, which is probably close to that of the highest bid. It's a good proxy of the actual price, and you're raising that.
Personally, I don't get too bugged about the snipers. I know what it's worth to me and let the auto-bidders handle it. If I get sniped, it means they were willing to pay more than I was. The seller gets screwed a bit, because the sniper was hiding their willingness to pay more (and therefore probably more than they ended up paying.) But the only thing being done to me as a bidder is that I got my hopes up.
So I don't get my hopes up, and I don't bid on eBay for anything I really, really want. I use eBay not for its auction, but for its flea-market: the ability to get stuff I can't get elsewhere. I usually just pay the "buy it now" price.
In this case I'm talking about an actual monetary cost. There are some spam-fighting schemes that call for purchased "stamps" on every email you send. If your email doesn't have one, I won't accept it. Spam works only because the mail is zero-cost. Even if it cost $.00001 per email, it wouldn't be cost effective any more.
There are a number of problems with the plan. One is that you can't right now usefully charge $.00001 for anything; the overhead will kill you. There are various schemes to work around that, but none are ready for prime time.
There's also the mass-email problem: I can't maintain a joke-of-the-day email list with a million people. Switching to RSS for that would help.
There's also the fact that much spam comes from email zombies: hacked computers that send spam. If those zombies can't get access to your account, then they can send all the spam they like; the lack of a stamp means it'll be dropped. But if they do hack somebody's account sufficiently to buy stamps with it, suddenly that's real money you're losing when you get infected.
So this is all aside from bandwidth costs. There are issues associated with that, too, but as a sister post explains, those are largely under control.
The nice thing about email newsletters is that they look just like your other communcations; you use one tool to manage them.
But email is a two-way communication; RSS is really primarily one-way. That makes for a technological difference: with RSS, because it's fetch, you know you're not getting spam. Email is push, and so it's hard to distinguish newsletters from spam. And it's one more site to give your email address to, meaning one more opportunity for spammers to steal/buy it.
Getting newsletters out of the email loop will make it easier to support some anti-spam technologies. Newsletters are one of the downfalls of pay-to-send schemes, because a free newsletter emailed to a million people at $.00001 turns into real money.
I like integrating RSS into the email stream. Some email apps already support RSS, and I would like to see them show up in just a single queue of "stuff to read".
Yeah, I suffer for my art.
I have no idea what the long-term effects are. And God knows the dry ice is more fun to play with. But if you want fog right in a particular place at a particular time, nothing beats a fog machine.
I do theatrical work, not concert work. Concerts usually just crank up the fog and let it go, the more fog the better. In a play there are scenes with fog and scenes without fog, and it's really nice to be able to control that with a guy in the light booth rather than dumping pellets into a bucket. Not to mention the ability to pump fog wherever you want it.
But it sure do stink.
For stage shows, fog machines are far more controllable and produce better results than dropping dry ice in water. They use "fog juice" rather than dry ice.
Though sometimes you'll use dry ice to cool the resulting fog. The hot fog gives you a smoky, atmospheric effect. If you want ground-hugging fog, you've got to cool it down, and dry ice is a pretty good way to chill it quickly.
Possibly just because there's money to be made. That's why they have an Xbox: if they can turn a profit on it, they will.
But in the case of the iPod, they want to compete against it for the same reason Apple created it: to introduce people to the Macintosh. Apple doesn't turn much in the way of profit on iTMS; it's just there to drive people to the iPod. The iPod does turn a profit, a pretty good one, but more importantly it gets people used to the idea that Apple products just work.
The iPod has astonishing market share despite the presence of cheaper, better-equipped alternatives. If people jump from Windows to Mac on the strength of that, it jepoardizes Microsoft's market share. At this point they depend vigorously on being the default OS choice. Erode their market share a little, and you open the door to eroding it a lot, as people no longer have to buy a Wintel box just to keep on the same page with their friends.
There's also the fact that a big company can never stand still. Just producing revenue isn't enough; they have to produce more revenue. One way to do that is to diversify, especially if you can diversify and still leverage your products in other areas. MS can do that big time.
For example, if they have a new, stronger DRM scheme (based, say, on Palladium), they may be able to get record companies to give them a price break, or even sign up those companies who don't trust Apple's FairPlay to protect their property.
MS can leverage their OS control (to give their device a performance hack that Apple can't get). Maybe they can leverage the Xbox, perhaps a plug on the side of an Xbox for their music player, or being able to build a handheld game device leveraging both the Xbox and music player platforms.
I don't know what they've gamed out, but basically, MS will try all of it. The downside, of course, is losing focus: it's usually better to make 1 good product than 10 shoddy ones. That's less about technology and more about management. MS thinks it has good management. On that, we'll have to see: the slips in the Vista schedule don't speak well to that.
Your observation answers my key question: why would they risk having their law struck down by trying it out on a site which is probably not in violation? (I haven't read the text of the law, so I can't say for certain.)
Answer: because to the Indian tribe casinos, linking to gambling site causes at least as much harm as gambling on one. Anti-gambling activists would probably care more about the actual gambling sites, but those who stand to make money from gambling demand that they limit the exposure as well.
Google Earth has a number of advantages over Google Maps:
* Because it's a native app rather than Web 2.0, it's snappier.
* There's an interface for making lists of points. For example, it's easy to create "this is my house" and "this is my office" places. That makes getting directions more convenient.
* Navigating in cities, the ability to see the buildings in 3D and get a ground-level perspective can be really useful in driving. It can be hard to figure out exactly what some directions mean just from the description.
* Printing directions is more effective, because it's not limited to the 72DPI resolution you get on a screen. (It still doesn't do as good a job as I'd like; I'm hoping version 4 does better.)
Those are things you actually want, in addition to the gee-whiz stuff. By all means continue to use Google Maps if you don't want to bother downloading something. But Google Earth isn't just fancy-but-irrelevant features.
I've worked with it on exactly one relatively simple spreadsheet, and I found numerous bugs just in that time:
* Some important formatting disappeared. Not just column widths, but numerical formatting. The difference between "currency" and "not currency" is very important for the look of the sheet.
* Re-exporting to Excel had a bug: it capitalized the sheet names, but didn't propagate that to formulas. Any formula that referenced another sheet became #VALUE
* Even for the small spreadsheet I was using (a few sheets, dozens-not-thousands of rows and columns), scrolling was very, ver slow.
This fairly simple sheet is what I think of as a canonical app for Google Spreadsheets: not mission critical, not large, not full of database lookups or macros. Maybe those are just beta complaints, but I've got to concur with your verdict: not ready for prime time.
And with a spacious 5" black-and-white screen.
God, I loved that computer.
Or at least pretending to sell prescription drugs on the Internet. I can't imagine that any of them actually send out the illegal pharmaceuticals. It's not like they're expecting to maintain a long-term relationship with you.
Maybe I'm wrong. Has anybody ever actually gotten meds from one of these guys?
Ah. Windows 2000/2003 Server. I hear life is better if you have one of those. I've never used it; it's kind of expensive.
I was totally with you until I got to the bit about Windows peer networking. If you know the magic button that allows me to get to other computers on the network without a 30-second hang before reporting, "No, I haven't figured out how to get to the computer that's right next to this one", I'd really like to know about it.
I think that protecting the user's own data is sufficient reason to blame this on the app writer, not the OS. Yeah, it's the OS's fault if rootkit-level harm can be done, but I think of that as a whole separate problem. Huge amounts of damage can be done even to the user's sandbox, including disclosure of private information (which isn't the OS's fault, either, if the app is giving its macros access to sockets).
There's plenty of blame to go around, but it points out a general clue: writing secure generalized systems is hard, whether it's an OS or a word processor that thinks it's an OS. Security is everybody's problem, and you have to think about it every time you get input from a user. Limiting the effect of security failure isn't the same as abnegating responsibility to prevent that failure. The more power you give that user, the more responsibility YOU have to ensure that power isn't misused.
Maybe because it tested the browsers a little too hard? I just checked it out with Firefox, and although it's more attractive, important parts of it didn't render correctly.
That could be just Coral's cache of it, however. I was having trouble getting to the live copy.
I don't actually use it for... well, what the hell was it supposed to be for? But my theater troupe uses an online ticketing service, and it's kinda nifty to be able to just scan the bar codes when they present the tickets. And all it took was a trip to Radio Shack and a downloaded driver.
All with you on the crummy page-layout support in OpenOffice. But I don't find that Word is significantly better at it. Maybe a little better, but even after using it for years I find Word's notions of frames and floating images utterly baffling.