Yes, let's have lots of trade barriers! That _will_ help the profession.
There's a bit of a problem forming on the world trade floor... the WTO has been handing down quite a few rulings that are going against the USA, and we haven't been doing ourselves very many favors with treaties such as NAFTA.
In short, the playing field keeps getting tilted against us in the name of "fairness". It's about time we started looking out for American workers when doing these things...
Telling your users how the system works theoretically is very important to prevent GIGO situations from developing. I can't count you the number of times as a "system administrator" I've been called in to investigate a ticket that they say system seemed to muck up, only to discover that it was the result of somebody entering in information that they thought meant X but the system interpreted as Y. Users who don't know how to use the system will definitely muck it up... users who are told the right way to do things will follow it.
For those working in a one-location company, do not hide in the IT room. When a user sends an e-mail asking for help, walk out to their desk rather than e-mail back. That way, you can see exactly what they're seeing on their screen, and you can also get a feel for what's going accross their desk while they're trying to interact with the systems.
That's one thing IT workers will never be able to duplicate...
Some people clung onto their old cell phone providers even after another provider started better or cheaper service in their area simply because they wanted to keep their numbers. Number portability was the solution to that problem.
Now, it'd be relatively simple to do this, just require that ISPs offer forwarding service for up to a year after a customer cancels, and the new ISP can kick back an e-mail telling anybody who's e-mails that the user has moved to them. Of course, no ISP is going to offer this without the government ordering them to... but couldn't the FTC or FCC step in on this one?
Let's consider the users who do nothing but e-mail with their Internet connection...
- Faster speed is not much of a benefit to them. They don't download images very often, and they're fine with walking away from their computer for however long it takes while those downloads happen. - They don't particularly care about their phone callers getting busy signals, they don't get that many really important phone calls anyway. - To them, changing e-mail addresses would be a nightmare. Some are even clinging onto address that they've had since 1994. The ISP may have gone defunct, but the old domain name is still being supported by the ISP that aquired them. Look at all the legacy domains Earthlink is still supporting. - And, we're also talking about people who hate monthly bills. For retired people, they plan their budgets very carefully and even a $10/month difference bothers them.
Bottom line... not everybody wants an always-on Internet connection. Sure, everybody reading Slashdot who doesn't have one wants one... but there are a lot of people in the USA who wouldn't even know what Slashdot is.
His actual testing environment was a virtual machine setup programmed to emulate a "a plain-vanilla SoundBlaster card".
Hello... that's the worldwide standard for plain sound cards, one that for years many vendors other than Creative followed. Even if you don't have the right drivers for a card, most sound cards will gladly accept the plain SoundBlaster driver and deliver the basic features in return.
To flunk that test is a little embarassing, especially when you have to go back to Windows 3.1 to find an MS operating system that fails to figure out what to do with a funky sound card. In short... Linux distros should try to install a generic SoundBlaster sound driver if it can't autodetect the sound card.
Fred Langa's main claim to fame was as one of the key personalities in CMP's now-defunct Windows Magazine. Therefore, he's much more familiar with Windows than Linux. Let's face it, he's paid to be a pundit that writes stories that sell magazines.
Although, this doesn't exactly invalidate his point. Microsoft's got a deep driver library database included in Windows XP... containing many cards that there is no known Linux drivers for.
Didn't someone already complete the recount and showed the Gore won??
Not quite. The "media recount" showed that depending on what your standard was for reading an unclear ballot was, there was at least one possible set of rules that would have allowed Gore to win, but under the actual rule sets that were in play in 2000, and even the rule sets the Gore campaign were requesting in court, Bush was the winner.
Here is a detailed accounting of the various counting methods from PBS. In the end, Bush is the legal and legitimate president, but there's still enough wiggle room for the Gore supporters to claim they were robbed as well. The core problem is that you don't know for sure that a dimpled chad was really an intent to vote, or a voter who started to push but then pulled back without thinking they comitted, or just the result of a woman's fingernail while holding the ballot improperly.
Really, punch cards have got to go... we shouldn't have any unclarity in how we're supposed to count ballots. The box should either be marked or not marked.
I've in fact written programs that caused BSODs back in the days of Windows 3.1 and QuickBasic...
The root cause of a traditional BSOD is that a program has messed with memory directly in a way that it wasn't supposed to. It's a "General Protection Fault"... a fault that is being declared by Windows because the program has tried to mess with an area of memory it wasn't allowed to touch for the general protection of the stablity of the system. Halting the broken process there is a safety valve to prevent the whole multi-application system from going down.
It wasn't until the NT kernel that Windows actually had decent memory management so that a process could be killed and have its memory space returned to the system without risk of the busted program having corrupted anything else. Programs still crash in Windows XP, but at least a crash in one program doesn't lead to a risk to any other program.
The OS didn't cause BSODs, the OS declared BSODs when a program acts out of bounds. Even Linux can have a "kernel panic" moment when a poorly designed driver program's flaw is exposed.
Any program crashes when it reaches an unantisipated error situation. The question comes over what part of the system is going to declare the error and shut down the errant process. If an IDE declares the error, it can highlight the bad line of code. If it's a program running natively, it'd be nice for the highest layer of the OS possible to raise the error. If it gets all the way down to the bottom of the OS and still isn't stopped... the OS is crashed.
So, it takes two to crash an OS... a program issuing an impossible instruction, and an OS that doesn't realize it needs to stop that program until it's too late.
Exactly, it stems from the Clinton era "Clipper Chip"... the theory is that the government would have the techincal ability to unlock the encryption via a backdoor, but in order to get the authority to use that backdoor they'd need the proper court approval. The main flaw that shot this plan down was that very few people believed the government's backdoor would stay secret very long, rendering whatever encryption the Clipper Chip used as worthless.
I think that estimate neglects just how few votes really decided the last election. It'd only take adding 538 additional votes for Gore in any combination of Florida districts to overturn the entire result.
If you're going to bias the election in favor of either of the two major parties, you have no need to attack the states in which your candidate is already going to win. You only need to bias enough close states to top the electoral vote balance, the popular vote doesn't matter.
As much as we say this is a nation of one-person-one-vote, that's never been the way a presdiential election is really scored.
What we need is a PKI-based smart card to verify identity. Not only would this be able to securely verify identity, but also allow digital signatures. Of course, none of this protects us from terrorists necessarily, but it would go a long way towards curbing identity theft, etc Instead of your easily-obtainable SSN securing many services, you could have a digital signature required before approval of credit, etc.
Anything that makes identity theft harder makes terrorism harder. Terrorists don't get all of their funding from overseas, some of it comes from being part of the illegal drug industry and comitting other routine profitable crimes. The better we secure ourselves against basic frauds, the harder it is for terrorists to get money...
Combo-based locks have no such "popular key"... and most are user-configurable. Sure, a brute-force attack at most would take only 1000 attempts, but that's 1000 attempts by hand which would take at least an hour or so.
I don't think it'd help the TSA much at all, because I'm sure by now they can authenticate any form of acceptable ID with a computer check to make sure that the ID's name, number, and picture all exist on one that was really issued.
What a national standard would fix would be the situations where an ID card is presented to somebody who doesn't have access to the databases it takes to verify the validity of the card, like employers or bartenders. It'd make life a little harder on somebody who intends on presenting a fake ID, that's all. Sure, fakes would be out there, but just like the way we keep adding security features to our money, those features only work when we all know where to look for them to determine when we have a fake before us.
The "escrow key" model of lock that now being distributed in the form of lugage lock leaves interesting options for a traveler...
- Leave your suitcase unlocked. The TSA can get access, and so can anybody else who wants to try to open it. - Lock your suitcase the old fashioned way. If the TSA wishes to check your bag, they'll bust your lock. Bad guys can also bust the lock. At least, if the contents are tampered with, you'll see a defeated lock when you recover your bag. - Lock your suitcase with the TSA-compliant locks. Most people can't open your bag, but TSA key holders (both good guys and bad guys) can get into your bag without having to break anything.
Nicholas Weaver has an interesting letter printed in the article where he makes the case for a need to assume that Microsoft's crown jewel, the Windows Source Code, has already fallen into the hands of black-hats, since both the Chinese and Russians have legit access, and the ease of which a determined group could steal it.
It's an intresting question. However, wouldn't we have seen more zero-day hacks in circulation from the black-hats who hold the code? Or maybe these exploits are being used, but with such infrequency that it's slipping under radars...
We already have multipurpose-use government-issued ID cards in our wallets in the form of drivers licenses or non-driver photo ID cards issued by our states.
The biggest problem with all of these is that there are 51 different issing bodies, one in every state plus one for Washington, D.C. Within each state, there are at least two formats to make non-drivers distinct from drivers, most states also have special "funny formats" for those under 21 so that they're more easily rejected when they try to purchase alcohol.
But, with more than a hundred formats for the best ID system we have, it's impossible for anybody to be an expert on what security measures to look for and be able to notice when they're absent.
No, this isn't an issue that'd protect us from suicide bombers or airplane hijackers... but being able to properly identify people is essential to financial transactions, and telling illegal immigrants that they don't belong here. It's not exactly a constitutional right to be able present a false ID as your own. The various issuers of drivers licenses should at least be able to agree on a common standard so those cards all look alike from jurisdiction to jurisdiction.
It's interesting to watch everyone salivate over google stock when there has been virtually no financial data published by the company (it is private after all).
Sure google is the most popular search engine and employs smart people but there's no telling what's happening on the business side of things.
They could be losing money for all we know.
But that wall is about to become transparent soon, IPO or not as Google slips past the line into having to comply with reporting rules even without an IPO. This Google IPO talk will dry up fast if things aren't as cool as people are imagining.
They have tons of cash, so why can't they just to a cash stock-buyback from all but 499 of their shareholders?
A company usually nearly-depletes its cash reserve before going IPO... So the fact that they still have cash indicates they're not so likely to IPO any time soon.
The day traders will probably have a blast playing the see-saw movements. People who buy the first day and hold for the long term are likely, *in my opinion*, lose money.
Yes. I'd agree that a buy-and-hold investor is best keeping away from Google during its early days on the market. It's going to be the most hyped IPO in a long time... and a lot of fortunes were made and blown during the dot-com era of high-flying IPOs. Day trader's paradise, average investor's nightmare...
If they did not want to do an IPO, couldn't they split up the activities? (Google ads, google servers etc.)?
That could turn ugly, as the departments would have to start charging each other for services... and there could be in-fighting that doesn't exist in the present setup.
It'd be easier to just report and not issue an IPO.
Slashdot Mirror
Yes, let's have lots of trade barriers! That _will_ help the profession.
There's a bit of a problem forming on the world trade floor... the WTO has been handing down quite a few rulings that are going against the USA, and we haven't been doing ourselves very many favors with treaties such as NAFTA.
In short, the playing field keeps getting tilted against us in the name of "fairness". It's about time we started looking out for American workers when doing these things...
Telling your users how the system works theoretically is very important to prevent GIGO situations from developing. I can't count you the number of times as a "system administrator" I've been called in to investigate a ticket that they say system seemed to muck up, only to discover that it was the result of somebody entering in information that they thought meant X but the system interpreted as Y. Users who don't know how to use the system will definitely muck it up... users who are told the right way to do things will follow it.
For those working in a one-location company, do not hide in the IT room. When a user sends an e-mail asking for help, walk out to their desk rather than e-mail back. That way, you can see exactly what they're seeing on their screen, and you can also get a feel for what's going accross their desk while they're trying to interact with the systems.
That's one thing IT workers will never be able to duplicate...
Study: 2 in 5 Web users now have broadband at home
A majority of web users don't have broadband at home...
Some people clung onto their old cell phone providers even after another provider started better or cheaper service in their area simply because they wanted to keep their numbers. Number portability was the solution to that problem.
Now, it'd be relatively simple to do this, just require that ISPs offer forwarding service for up to a year after a customer cancels, and the new ISP can kick back an e-mail telling anybody who's e-mails that the user has moved to them.
Of course, no ISP is going to offer this without the government ordering them to... but couldn't the FTC or FCC step in on this one?
Let's consider the users who do nothing but e-mail with their Internet connection...
- Faster speed is not much of a benefit to them. They don't download images very often, and they're fine with walking away from their computer for however long it takes while those downloads happen.
- They don't particularly care about their phone callers getting busy signals, they don't get that many really important phone calls anyway.
- To them, changing e-mail addresses would be a nightmare. Some are even clinging onto address that they've had since 1994. The ISP may have gone defunct, but the old domain name is still being supported by the ISP that aquired them. Look at all the legacy domains Earthlink is still supporting.
- And, we're also talking about people who hate monthly bills. For retired people, they plan their budgets very carefully and even a $10/month difference bothers them.
Bottom line... not everybody wants an always-on Internet connection. Sure, everybody reading Slashdot who doesn't have one wants one... but there are a lot of people in the USA who wouldn't even know what Slashdot is.
His actual testing environment was a virtual machine setup programmed to emulate a "a plain-vanilla SoundBlaster card".
Hello... that's the worldwide standard for plain sound cards, one that for years many vendors other than Creative followed. Even if you don't have the right drivers for a card, most sound cards will gladly accept the plain SoundBlaster driver and deliver the basic features in return.
To flunk that test is a little embarassing, especially when you have to go back to Windows 3.1 to find an MS operating system that fails to figure out what to do with a funky sound card. In short... Linux distros should try to install a generic SoundBlaster sound driver if it can't autodetect the sound card.
The point is, XP's got drivers for some of those historic cards. If it got a driver into Windows 95, it still works in Windows XP.
Linux's driver history doesn't go back that far... so some hardware that works with Windows just will never work with Linux.
Fred Langa's main claim to fame was as one of the key personalities in CMP's now-defunct Windows Magazine. Therefore, he's much more familiar with Windows than Linux. Let's face it, he's paid to be a pundit that writes stories that sell magazines.
Although, this doesn't exactly invalidate his point. Microsoft's got a deep driver library database included in Windows XP... containing many cards that there is no known Linux drivers for.
Didn't someone already complete the recount and showed the Gore won??
Not quite. The "media recount" showed that depending on what your standard was for reading an unclear ballot was, there was at least one possible set of rules that would have allowed Gore to win, but under the actual rule sets that were in play in 2000, and even the rule sets the Gore campaign were requesting in court, Bush was the winner.
Here is a detailed accounting of the various counting methods from PBS. In the end, Bush is the legal and legitimate president, but there's still enough wiggle room for the Gore supporters to claim they were robbed as well. The core problem is that you don't know for sure that a dimpled chad was really an intent to vote, or a voter who started to push but then pulled back without thinking they comitted, or just the result of a woman's fingernail while holding the ballot improperly.
Really, punch cards have got to go... we shouldn't have any unclarity in how we're supposed to count ballots. The box should either be marked or not marked.
I've in fact written programs that caused BSODs back in the days of Windows 3.1 and QuickBasic...
The root cause of a traditional BSOD is that a program has messed with memory directly in a way that it wasn't supposed to. It's a "General Protection Fault"... a fault that is being declared by Windows because the program has tried to mess with an area of memory it wasn't allowed to touch for the general protection of the stablity of the system. Halting the broken process there is a safety valve to prevent the whole multi-application system from going down.
It wasn't until the NT kernel that Windows actually had decent memory management so that a process could be killed and have its memory space returned to the system without risk of the busted program having corrupted anything else. Programs still crash in Windows XP, but at least a crash in one program doesn't lead to a risk to any other program.
The OS didn't cause BSODs, the OS declared BSODs when a program acts out of bounds. Even Linux can have a "kernel panic" moment when a poorly designed driver program's flaw is exposed.
Any program crashes when it reaches an unantisipated error situation. The question comes over what part of the system is going to declare the error and shut down the errant process. If an IDE declares the error, it can highlight the bad line of code. If it's a program running natively, it'd be nice for the highest layer of the OS possible to raise the error. If it gets all the way down to the bottom of the OS and still isn't stopped... the OS is crashed.
So, it takes two to crash an OS... a program issuing an impossible instruction, and an OS that doesn't realize it needs to stop that program until it's too late.
Exactly, it stems from the Clinton era "Clipper Chip"... the theory is that the government would have the techincal ability to unlock the encryption via a backdoor, but in order to get the authority to use that backdoor they'd need the proper court approval. The main flaw that shot this plan down was that very few people believed the government's backdoor would stay secret very long, rendering whatever encryption the Clipper Chip used as worthless.
I think that estimate neglects just how few votes really decided the last election. It'd only take adding 538 additional votes for Gore in any combination of Florida districts to overturn the entire result.
If you're going to bias the election in favor of either of the two major parties, you have no need to attack the states in which your candidate is already going to win. You only need to bias enough close states to top the electoral vote balance, the popular vote doesn't matter.
As much as we say this is a nation of one-person-one-vote, that's never been the way a presdiential election is really scored.
What we need is a PKI-based smart card to verify identity. Not only would this be able to securely verify identity, but also allow digital signatures. Of course, none of this protects us from terrorists necessarily, but it would go a long way towards curbing identity theft, etc Instead of your easily-obtainable SSN securing many services, you could have a digital signature required before approval of credit, etc.
Anything that makes identity theft harder makes terrorism harder. Terrorists don't get all of their funding from overseas, some of it comes from being part of the illegal drug industry and comitting other routine profitable crimes. The better we secure ourselves against basic frauds, the harder it is for terrorists to get money...
Combo-based locks have no such "popular key"... and most are user-configurable. Sure, a brute-force attack at most would take only 1000 attempts, but that's 1000 attempts by hand which would take at least an hour or so.
I don't think it'd help the TSA much at all, because I'm sure by now they can authenticate any form of acceptable ID with a computer check to make sure that the ID's name, number, and picture all exist on one that was really issued.
What a national standard would fix would be the situations where an ID card is presented to somebody who doesn't have access to the databases it takes to verify the validity of the card, like employers or bartenders. It'd make life a little harder on somebody who intends on presenting a fake ID, that's all. Sure, fakes would be out there, but just like the way we keep adding security features to our money, those features only work when we all know where to look for them to determine when we have a fake before us.
How long is it until somebody buys up some of these TSA-unlockable locks and reverse engineers their way into a duplicate of the TSA key?
The "escrow key" model of lock that now being distributed in the form of lugage lock leaves interesting options for a traveler...
- Leave your suitcase unlocked. The TSA can get access, and so can anybody else who wants to try to open it.
- Lock your suitcase the old fashioned way. If the TSA wishes to check your bag, they'll bust your lock. Bad guys can also bust the lock. At least, if the contents are tampered with, you'll see a defeated lock when you recover your bag.
- Lock your suitcase with the TSA-compliant locks. Most people can't open your bag, but TSA key holders (both good guys and bad guys) can get into your bag without having to break anything.
Hmm.. which option to chose?
Nicholas Weaver has an interesting letter printed in the article where he makes the case for a need to assume that Microsoft's crown jewel, the Windows Source Code, has already fallen into the hands of black-hats, since both the Chinese and Russians have legit access, and the ease of which a determined group could steal it.
It's an intresting question. However, wouldn't we have seen more zero-day hacks in circulation from the black-hats who hold the code? Or maybe these exploits are being used, but with such infrequency that it's slipping under radars...
We already have multipurpose-use government-issued ID cards in our wallets in the form of drivers licenses or non-driver photo ID cards issued by our states.
The biggest problem with all of these is that there are 51 different issing bodies, one in every state plus one for Washington, D.C. Within each state, there are at least two formats to make non-drivers distinct from drivers, most states also have special "funny formats" for those under 21 so that they're more easily rejected when they try to purchase alcohol.
But, with more than a hundred formats for the best ID system we have, it's impossible for anybody to be an expert on what security measures to look for and be able to notice when they're absent.
No, this isn't an issue that'd protect us from suicide bombers or airplane hijackers... but being able to properly identify people is essential to financial transactions, and telling illegal immigrants that they don't belong here. It's not exactly a constitutional right to be able present a false ID as your own. The various issuers of drivers licenses should at least be able to agree on a common standard so those cards all look alike from jurisdiction to jurisdiction.
A company usually nearly-depletes its cash reserve before going IPO
Is there a particular strategic reason for this?
Yes. There's no need to go do a funding-for-ownership trade in an IPO if there's plenty of spare funds lying around.
It's interesting to watch everyone salivate over google stock when there has been virtually no financial data published by the company (it is private after all). Sure google is the most popular search engine and employs smart people but there's no telling what's happening on the business side of things. They could be losing money for all we know. But that wall is about to become transparent soon, IPO or not as Google slips past the line into having to comply with reporting rules even without an IPO. This Google IPO talk will dry up fast if things aren't as cool as people are imagining.
They have tons of cash, so why can't they just to a cash stock-buyback from all but 499 of their shareholders?
A company usually nearly-depletes its cash reserve before going IPO... So the fact that they still have cash indicates they're not so likely to IPO any time soon.
The day traders will probably have a blast playing the see-saw movements. People who buy the first day and hold for the long term are likely, *in my opinion*, lose money.
Yes. I'd agree that a buy-and-hold investor is best keeping away from Google during its early days on the market. It's going to be the most hyped IPO in a long time... and a lot of fortunes were made and blown during the dot-com era of high-flying IPOs. Day trader's paradise, average investor's nightmare...
If they did not want to do an IPO, couldn't they split up the activities? (Google ads, google servers etc.)?
That could turn ugly, as the departments would have to start charging each other for services... and there could be in-fighting that doesn't exist in the present setup.
It'd be easier to just report and not issue an IPO.