AIUI Windows NT was POSIX compliant. But the POSIX specifications of the time left huge swathes of API defined where it was perfectly OK to return ERROR_NOT_IMPLEMENTED.
None of the mainstream Unix vendors actually did this, so most Unix code was written on the assumption that very little was not implemented. Windows, OTOH, returned ERROR_NOT_IMPLEMENTED everywhere it could. With fairly predictable results.
Apparently the DHS is looking into revising the rule.
The quickest and easiest way to guarantee that would happen would be if a senior person within the DHS was declared dead.
In fact, a similar procedure would work well for most governmental policies - have the government work by it (and any unintended consequences) for 6 months before inflicting it on the general public.
All these researchers did was discover if the people who were behind this fraud had to be a fantastically sophisticated criminal gang with insiders working at the garage, the bank clearing house and the PIN-pad manufacturers - or if anyone with a bit of motivation anf a few relatively basic tools could do it.
Would that be the FUD campaign that has so far put off remarkably few businesses, not really had much of an impact on IBM, Red Hat or Novell's bottom line and been smacked down in court so many times now it's a wonder that nobody's been brought up for wasting the courts' time?
If Microsoft is behind this VC taking an interest in SCO, I can think of only two possible reasons:
1. They don't know when to give up and look for another tack. Possible, certainly. 2. There is a lot more to this than we already know - so much so that someone's prepared to pay a lot of money to prevent that information possibly becoming public.
Point (2) could explain funding by anyone, not just Microsoft.
If the author of that linked article was such a high-up boardroom member, do you think s/he'd be making a living by writing articles and hoping they get submitted to slashdot?
Those who can, do.
Those who can't, teach.
Those who can't teach write about in magazines. Not much has changed since magazines started going digital.
He assumes that the largest market for software is in selling packaged, off the shelf products.
Well, it may be for a few things like games and OSs. But for 90% of the computing industry, the real money has always been in customised or semi-customised (based on off-the-shelf modules) solutions. Microsoft and Electronic Arts are anomalies, making good money with COTS products.
The IE error page is pretty unhelpful at the best of times. You'd have to be nuts to rely on it for error checking. You should use nslookup, dig, ping and traceroute.
Using a cookie to simulate disabling the feature means that nslookup and dig will still return erroneous results, and ping and traceroute will still work even in cases where they shouldn't.
"The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.".
It's probably not politically correct to point out that in this case, "progress" would mean "towards a Taliban-controlled state which is about half a millennium behind the rest of the world".
True, but there are three things about the Google commodity-cluster which do pose a significant issue:
1. It doesn't present the filesystem through standard POSIX calls. So you can't just set up a massive filesystem and let any application you like store files there, the application needs to support the clustering specifically. 2. It's not available outside of Google. 3. IIRC it doesn't guarantee that every block across the cluster is in perfect step. So you could wind up with User A being presented one set of records in the finance system and User B being presented a different set, and no way of knowing which is the most up to date version.
I'm sure these could be solved in time but right now they're pretty major issues.
Which in turn will deter lots of these C&D letters, especially when they're just bluffing (and they know it).
But not this one. IANAL and I haven't read the letter, but I'll put money on it it uses the DMCA's clause regarding "circumventing an effective technological measure".
In other words, in the eyes of the law the C&D is perfectly acceptable, even if you don't like it.
Google style clustering, where you know some of your hardware will fail from time to time and you're just OK with that, is the first promising alternative to mainframe uptimes since the days of VMS clusters. The best hardware from Sun or HP never came close.
Google have a very specific set of applications which, for the most part, don't really care if chunks of data from the database go missing occasionally, can be easily mirrored and it's not particularly crucial that every mirror is in perfect lockstep.
Try proposing a system like that to the IT (or, for that matter, the Finance) director of a $multi-million firm and let me know how you get on.
This sort of thing will only get more common as time goes on a people use the net for ever more and bigger media. Personally I think ISPs need to do more to bite the bullet and price their services honestly, rather than pricing them cheap and then coming up with a million and one reasons you can't have what you thought you'd paid for.
You'd hope so, wouldn't you?
It's not quite as simple as that, however. While there are a huge number of ISPs, there is 1 company providing bandwidth on a wholesale basis between customers' homes and ISP premises - the former monopoly telco, British Telecom. If the ISP doesn't want to buy bandwidth from BT - that's absolutely fine, they can organise to set up their own nationwide network infrastructure. Good luck getting permission to dig up all the roads.
Local loop unbundling should in theory eliminate that issue but I haven't seen very many ISPs take advantage of that yet.
Would this apply if their press releases had said "We will not sue!* Aren't we nice? Microsoft believes in an open development process! (* subject to the following conditions...)" and the conditions were missed out by the reporters?
This is exactly the kind of thing I can imagine happening.
Solution 1: A solution is implemented which pays lip service to the requirement - something like ISPs poisoning the entries their DNS servers provide on demand of the BPI - or if they're really paranoid, null-routing the IP addresses. This is the kind of thing the ISPs would go for, isn't too onerous and doesn't actually do anything to solve the "problem".
Solution 2: The Great Firewall of Britain. This is what I see the Government doing if the ISPs don't. I doubt it'll be terribly effective because the government will outsource providing appropriate technology to a consultant like EDS (a company that specialises in taking money off UK government departments in exchange for half-baked systems which don't really work properly) and once the technology is ready, ISPs will be obliged to deploy it.
How is it that it caught on for the web (credit card payments over SSL), but still barely for personal communications (gpg, encrypted IM)?
That's a very good question.
One idea I've heard is that when SSL was first developed, the web was in its infancy and nobody really felt happy about the idea of sending their credit card details over it. The fact that it was relatively easy to eavesdrop on a computer network was fairly well known. This was no good to anyone who wanted to do business (OK, porn sites) over the web, and so SSL solved that problem by providing reassurance that nobody was eavesdropping.
The telephone system, on the other hand - that's been around so long that it's familiar technology and relatively few people are aware of how insecure it is. If you think GSM is bad (it's actually not that poor, and 3G introduces AES encryption), consider your land line. No encryption whatsoever and an analogue signal (so no computer equipment or specialised unusual codecs required to tap) between you and the telephone exchange.
Verified Identity Pass, a firm that offers checkpoint services at airports, has announced a $500,000 award for any solution that will make airport security checks quicker and simpler for passengers.
Translation: "We can't figure out how to do our job, so we'll pay someone else to do it instead. Yet once they give us their idea, we won't offer them a job and we won't give them royalties. Instead, they get a single one-off payment and we get to milk the idea indefinitely."
The GIMP might be very powerful and feature packed, but the learning curve to get into it is cliff shaped.
Same's true of the full version of Photoshop. There are a few extra filters which are quite useful (such as remove redeye) but an understanding of layered image editing is absolutely crucial to do anything useful with either of them.
No, but I'd imagine the Consumer Protection (Distance Selling) Regulations 2000 would apply. These give the customer a cooling off period of 7 working days.
China is one of the few contries that have a military that can take ours and who is not a trustworthy friend.
Correct me if I'm wrong, but no country in the whole of history with a military might that could significantly threaten America has ever been a trustworthy friend.
Each business I've seen has some program that's vital to the operation and only runs on windows. Wine will not fly with these people. They could give a shit about Linux, they want it to work and they don't want to learn anything else.
Even if they would be prepared to give it a go under Wine, what do you think Sage/Quicken/(whoever) will say when you call up for support and explain that you're having some trouble?
You: Hi, I'm having trouble with Quicken, I wonder if you could help? Cust. Rep: Sure, what version of Windows are you running? You: Er... I'm not, I'm running it under Wine in Ubuntu Linux Cust. Rep: <click>
I've yet to see a wireless card that didn't need a driver installing under Windows XP.
Everything needs a driver. The keyboard you're using needs a driver, the mouse needs a driver, the USB port they're plugged into needs a driver.
However keyboards, mice and USB ports are sufficiently standardised that the same driver works for all - though you may not get all the bells and whistles like the extra buttons on some keyboards. Not so wireless networking.
The only difference is that Linux tends to bundle the drivers for everything, because that's the only way it'll get supported. Windows doesn't tend to, because it's far easier to let the hardware OEM deal with that.
In a business environment deploying an new OS or OS version is expensive, and licensing is rarely the largest portion of that. I suppose if you were running your XP machines with Automatic update on pointing directly at windows update instead of at a SUS server, the activation requirement could be expensive or tedious, but that's a relatively small subsection of businesses really.
I would point out that Microsoft probably don't have a clue how widely deployed Vista is in business. Because businesses aren't going to the local PC World and buying 100 copies of Vista off the shelf - they're paying an annual fee to run Windows, and Microsoft are probably including everyone who's paid for Software Assurance in the list of "businesses which have licensed and therefore, we assume, deployed Vista".
Slashdot Mirror
AIUI Windows NT was POSIX compliant. But the POSIX specifications of the time left huge swathes of API defined where it was perfectly OK to return ERROR_NOT_IMPLEMENTED.
None of the mainstream Unix vendors actually did this, so most Unix code was written on the assumption that very little was not implemented. Windows, OTOH, returned ERROR_NOT_IMPLEMENTED everywhere it could. With fairly predictable results.
Apparently the DHS is looking into revising the rule.
The quickest and easiest way to guarantee that would happen would be if a senior person within the DHS was declared dead.
In fact, a similar procedure would work well for most governmental policies - have the government work by it (and any unintended consequences) for 6 months before inflicting it on the general public.
That's Jeremy Paxman. He really doesn't hold back - and it can be a thing of beauty to watch.
Frankly, we could do with a few more interviewers like him - unfortunately, there's only a couple of shows where that kind of thing goes on.
The thing is, there are already examples of tampered PIN pads being used to capture people's card details - for example, see:
http://news.bbc.co.uk/1/hi/england/4980190.stm
All these researchers did was discover if the people who were behind this fraud had to be a fantastically sophisticated criminal gang with insiders working at the garage, the bank clearing house and the PIN-pad manufacturers - or if anyone with a bit of motivation anf a few relatively basic tools could do it.
It turns out that pretty much anyone could do it.
Would that be the FUD campaign that has so far put off remarkably few businesses, not really had much of an impact on IBM, Red Hat or Novell's bottom line and been smacked down in court so many times now it's a wonder that nobody's been brought up for wasting the courts' time?
If Microsoft is behind this VC taking an interest in SCO, I can think of only two possible reasons:
1. They don't know when to give up and look for another tack. Possible, certainly.
2. There is a lot more to this than we already know - so much so that someone's prepared to pay a lot of money to prevent that information possibly becoming public.
Point (2) could explain funding by anyone, not just Microsoft.
If the author of that linked article was such a high-up boardroom member, do you think s/he'd be making a living by writing articles and hoping they get submitted to slashdot?
Those who can, do.
Those who can't, teach.
Those who can't teach write about in magazines. Not much has changed since magazines started going digital.
He assumes that the largest market for software is in selling packaged, off the shelf products.
Well, it may be for a few things like games and OSs. But for 90% of the computing industry, the real money has always been in customised or semi-customised (based on off-the-shelf modules) solutions. Microsoft and Electronic Arts are anomalies, making good money with COTS products.
The IE error page is pretty unhelpful at the best of times. You'd have to be nuts to rely on it for error checking. You should use nslookup, dig, ping and traceroute.
Using a cookie to simulate disabling the feature means that nslookup and dig will still return erroneous results, and ping and traceroute will still work even in cases where they shouldn't.
I think George Bernard Shaw said it best:
"The reasonable man adapts himself to the world: the unreasonable one persists in trying to adapt the world to himself. Therefore all progress depends on the unreasonable man.".
It's probably not politically correct to point out that in this case, "progress" would mean "towards a Taliban-controlled state which is about half a millennium behind the rest of the world".
True, but there are three things about the Google commodity-cluster which do pose a significant issue:
1. It doesn't present the filesystem through standard POSIX calls. So you can't just set up a massive filesystem and let any application you like store files there, the application needs to support the clustering specifically.
2. It's not available outside of Google.
3. IIRC it doesn't guarantee that every block across the cluster is in perfect step. So you could wind up with User A being presented one set of records in the finance system and User B being presented a different set, and no way of knowing which is the most up to date version.
I'm sure these could be solved in time but right now they're pretty major issues.
Which in turn will deter lots of these C&D letters, especially when they're just bluffing (and they know it).
But not this one. IANAL and I haven't read the letter, but I'll put money on it it uses the DMCA's clause regarding "circumventing an effective technological measure".
In other words, in the eyes of the law the C&D is perfectly acceptable, even if you don't like it.
Google style clustering, where you know some of your hardware will fail from time to time and you're just OK with that, is the first promising alternative to mainframe uptimes since the days of VMS clusters. The best hardware from Sun or HP never came close.
Google have a very specific set of applications which, for the most part, don't really care if chunks of data from the database go missing occasionally, can be easily mirrored and it's not particularly crucial that every mirror is in perfect lockstep.
Try proposing a system like that to the IT (or, for that matter, the Finance) director of a $multi-million firm and let me know how you get on.
This sort of thing will only get more common as time goes on a people use the net for ever more and bigger media. Personally I think ISPs need to do more to bite the bullet and price their services honestly, rather than pricing them cheap and then coming up with a million and one reasons you can't have what you thought you'd paid for.
You'd hope so, wouldn't you?
It's not quite as simple as that, however. While there are a huge number of ISPs, there is 1 company providing bandwidth on a wholesale basis between customers' homes and ISP premises - the former monopoly telco, British Telecom. If the ISP doesn't want to buy bandwidth from BT - that's absolutely fine, they can organise to set up their own nationwide network infrastructure. Good luck getting permission to dig up all the roads.
Local loop unbundling should in theory eliminate that issue but I haven't seen very many ISPs take advantage of that yet.
unlimited bits at the contract rate or I get to sue!
Sure, no problem. Your contract says up to 10 Mbps, with no lower bound.
You may download as much as you like on the throttled 100Kbps service. Have fun!
Would this apply if their press releases had said "We will not sue!* Aren't we nice? Microsoft believes in an open development process! (* subject to the following conditions...)" and the conditions were missed out by the reporters?
This is exactly the kind of thing I can imagine happening.
Solution 1: A solution is implemented which pays lip service to the requirement - something like ISPs poisoning the entries their DNS servers provide on demand of the BPI - or if they're really paranoid, null-routing the IP addresses. This is the kind of thing the ISPs would go for, isn't too onerous and doesn't actually do anything to solve the "problem".
Solution 2: The Great Firewall of Britain. This is what I see the Government doing if the ISPs don't. I doubt it'll be terribly effective because the government will outsource providing appropriate technology to a consultant like EDS (a company that specialises in taking money off UK government departments in exchange for half-baked systems which don't really work properly) and once the technology is ready, ISPs will be obliged to deploy it.
How is it that it caught on for the web (credit card payments over SSL), but still barely for personal communications (gpg, encrypted IM)?
That's a very good question.
One idea I've heard is that when SSL was first developed, the web was in its infancy and nobody really felt happy about the idea of sending their credit card details over it. The fact that it was relatively easy to eavesdrop on a computer network was fairly well known. This was no good to anyone who wanted to do business (OK, porn sites) over the web, and so SSL solved that problem by providing reassurance that nobody was eavesdropping.
The telephone system, on the other hand - that's been around so long that it's familiar technology and relatively few people are aware of how insecure it is. If you think GSM is bad (it's actually not that poor, and 3G introduces AES encryption), consider your land line. No encryption whatsoever and an analogue signal (so no computer equipment or specialised unusual codecs required to tap) between you and the telephone exchange.
Verified Identity Pass, a firm that offers checkpoint services at airports, has announced a $500,000 award for any solution that will make airport security checks quicker and simpler for passengers.
Translation: "We can't figure out how to do our job, so we'll pay someone else to do it instead. Yet once they give us their idea, we won't offer them a job and we won't give them royalties. Instead, they get a single one-off payment and we get to milk the idea indefinitely."
The GIMP might be very powerful and feature packed, but the learning curve to get into it is cliff shaped.
Same's true of the full version of Photoshop. There are a few extra filters which are quite useful (such as remove redeye) but an understanding of layered image editing is absolutely crucial to do anything useful with either of them.
No, but I'd imagine the Consumer Protection (Distance Selling) Regulations 2000 would apply. These give the customer a cooling off period of 7 working days.
Source:
UK Office of Fair Trading
China is one of the few contries that have a military that can take ours and who is not a trustworthy friend.
Correct me if I'm wrong, but no country in the whole of history with a military might that could significantly threaten America has ever been a trustworthy friend.
Each business I've seen has some program that's vital to the operation and only runs on windows. Wine will not fly with these people. They could give a shit about Linux, they want it to work and they don't want to learn anything else.
Even if they would be prepared to give it a go under Wine, what do you think Sage/Quicken/(whoever) will say when you call up for support and explain that you're having some trouble?
You: Hi, I'm having trouble with Quicken, I wonder if you could help?
Cust. Rep: Sure, what version of Windows are you running?
You: Er... I'm not, I'm running it under Wine in Ubuntu Linux
Cust. Rep: <click>
I've yet to see a wireless card that didn't need a driver installing under Windows XP.
Everything needs a driver. The keyboard you're using needs a driver, the mouse needs a driver, the USB port they're plugged into needs a driver.
However keyboards, mice and USB ports are sufficiently standardised that the same driver works for all - though you may not get all the bells and whistles like the extra buttons on some keyboards. Not so wireless networking.
The only difference is that Linux tends to bundle the drivers for everything, because that's the only way it'll get supported. Windows doesn't tend to, because it's far easier to let the hardware OEM deal with that.
You should actually try using Linux.
You'll be amazed how trouble free updating ALL of your installed software is.
Today, you are correct and the GP is wrong.
Should the Linux desktop ever become sufficiently commonplace, I suspect the reverse will be true.
In a business environment deploying an new OS or OS version is expensive, and licensing is rarely the largest portion of that. I suppose if you were running your XP machines with Automatic update on pointing directly at windows update instead of at a SUS server, the activation requirement could be expensive or tedious, but that's a relatively small subsection of businesses really.
I would point out that Microsoft probably don't have a clue how widely deployed Vista is in business. Because businesses aren't going to the local PC World and buying 100 copies of Vista off the shelf - they're paying an annual fee to run Windows, and Microsoft are probably including everyone who's paid for Software Assurance in the list of "businesses which have licensed and therefore, we assume, deployed Vista".