No, humans aren't 100% and yes, you can test for that. Try a thought experiment: fill a bin with 50,000 red balls and 50,000 blue balls. Ask a human to sort them all. The result probably won't be 100%, but you can still check the result and figure out how accurate the human is without relying on a superhuman ability to tell the balls apart. Same thing for spam: if you start with a known training set, you can test humans to see how well the spam is identified by manual sorting.
Business units always pay the costs of all company services. Unless you work for the government, there's no other alternative. The only variable is at what level the costs are allocated: the crudest form is just below the CEO level, where all IT costs are rolled up en masse against the income across all business units. At the other end of the spectrum, the lowest level of business unit accounts in nickels and dimes for every byte and packet used. Rationality is somewhere in the middle.
This is commonly referred to as "Activity-Based Costing". It means that the charges are based on allocating the entire cost of the IT department back across some measure of what the IT department actually does. Since the business units can manage their needs more quickly than the IT department can alter their ability to provide, what happens is businesses pull back their demands for IT services, so the allocated costs for individual services goes up, and the businesses respond by trying to reduce demand still more. Instead, there's just a big incentive for the business units to spin up their own capabilities to do things cheaper outside the IT department. If the IT department responds as a monopoly provider with obstructionist rules, changing chargeback structures, access charges, their relationship with the business units degenerates still more.
In the long run, what this leads to is both the desire and the ability for the parent company to outsource the IT function competitively. That may well be what the upper management of your company actually wants to do. It's certainly a sign that you are headed in the opposite direction of the partnership approach. What you really need to do is sell your services at a level that makes sense to the business, so that neither side gets surprised by the costs and everybody understands what they're actually going to get in exchange for the expense. Then you've got a basis for deciding what makes sense to do internally and what makes sense to outsource.
In the past, some companies (such as Oracle) have offered licensing terms based on the number of CPUs. So, a 16-way zSeries running, say, 10,000 Linux images could really bring down the average cost for licensing, couldn't it? Of course, 10,000 Oracle instances would probably kill this box no matter how superior its I/O channels are, but it does present interesting options for an ASP offering.
Interesting stats from the article: 200 million people use Passport as a sign-on service, but according to the first paragraph on the second page of the article, only 2 million users have created "wallets". So their uptake rate is pitiful.
The next stat I'd like to see is how many of the 2 million have actually (and voluntarily...) performed a financial transaction using the thing. And how many of those were just people who created the stupid wallet so they could do business with one of the 70 sites that MS has signed up.
BTW, I just spent a few minuts nosing around the passport site, and I could not see any information (at least, without signing in) about how one might go about deleting a wallet or an entire account.
Several years ago, the Wall Street Journal published an interview with one of the Vice Presidents at Hormel about the way Spam was generally disparaged. This was before the Internet usage of the term really became mainstream, but long after the Monty Python skit. The VP mused that they didn't spend much money advertising it, but they sure sold a lot of the stuff. It looks to me like they do an interview or press release every so often, and get as much mileage out of the whole thing as they can. Seems like a good trade to me--we abuse their trademark, and they sell random pig tidbits in a can.
I wouldn't panic about the high prices just yet. It's pretty common for companies to price the "engineering" copies of their devices pretty high. They're expensive to make, and are intended for early developers who need them to build commercial products. The high price helps cover the high initial costs and, perhaps more importantly, helps ration the devices out to people who are serious about building profitable products.
The more interesting number is the $8/unit for quantity > 1 million. That's not quite at the $5 target, but it's at least in the ballpark.
Not going to help. I actually have sent log files to them. No response whatever.
BTW, @Home portscans seem to resove back to machines with names like authorized-scan1.security.home.net, and lately they've been hitting NNTP ports pretty frequently. It seems they're more interested in making sure I don't violate my agreement not to run servers and soak up their upstream bandwidth than they are in actually securing their network.
The fact of the matter is that by importing large number of H1B workers into the US, we are surpressing the natural rise in wages that would occur in the presence of a real labor shortage.
We are also (partially) supressing the natural exporting of jobs that occurs in the presence of cheaper sources of labor outside the United States.
Actually, Sony may know something about the Internet that most/. readers don't think about. Don't worry about what Sony does to your box. Look at the other threats:
"we will block it at your cable company". Sony controls much of the content cable companies need. Sony also makes lots of head end gear. They may already have some equipment at your cable head end that can at least block cable modem users. Take a look here for some of their DTV products. The cable companies hate what they consider piracy; they aren't going to go to bat for you.
"we will block it at your phone company" Your phone company is trying to get you to use DSL, and needs media content to make it work. Napster helps them in the short run, because users want more bandwidth. But if the phone company wants to get direct revenue from media (like VOD, for example), they're going to have to play ball with the content owners. And remember, there aren't that many phone companies to deal with, and they aren't likely to want to fight Sony on your behalf.
"we will block it at your [Internet-service provider]" How about if they are able to persuade AOL to block it? Probably not that hard.
So if AOL users go away in the short run, and broadband users go away in the short and the long run, they've at the very least bought a couple more years to crush this sort of thing.
If you think about it, the set-top boxes that receive digital video aren't all that different from game consoles (or Tivo boxes or WebTV boxes, for that matter). A future Nintendo game console could potentially receive streaming video--all it really takes is a demodulator (cable, satellite, etc.), an MPEG-2 decoder (HW or fast SW) and the right conditional access equipment. To keep up with Sony, Nintendo might even have to build or acquire a server that competes directly with nCube. Sony already makes lots of cable head-end equipment, set top boxes, and game consoles.
I can't imagine that Nintendo would limit their ability to compete in the video market over something this easy to avoid. Look for a totally different name.
VBIs? Don't be silly--digital can just send the alternate versions in separate streams.
Life is scary when your local broadcaster can sell that ad space separately. Life is really scary when your settop box causes you and your next-door neighbor to get different images.
"Mommy, how come Jennifer Aniston always drinks Bud Light on our TV and Diet Coke over at Bobby's house?"
"spending time playing with the OS because it's cool" makes Linux cost-prohibitive.... and no matter how much you play with it, Windows just isn't going to get cool.
You will be dealing with individuals. And usually, the first person dispached on a problem call is the lowest-ranked, -paid, and -trained. And a large percentage of such, if they can't resolve the problem, will resort to finger-pointing instead of calling their next level support.
That's exactly the key to dealing with this problem. I used to deal with a lot of escalations for a big company that offered services (not IBM). One of the things that happened several times was the client started screaming "we're paying $XXX/hr to you, and we expect it to be fixed now!" So the junior person panics and starts pointing fingers. The whole exercise adds tension, confusion, and bogus data to the mix. To avoid this, ask the person to do two things:
Call his or her company and ask see if anyone has encountered the problem before. (The objective is really just to give the person a graceful way to call for help.)
Tell the person you don't want to get into a circle of finger-pointing, and ask him/her to help isolate the problem so you can supply proof to your manager or the vendor who's being pointed at. (The isolation will help regardless of which vendor turns out to be at fault.)
Yes, there are different levels of cleanliness expected of different people. I'm not ready to say who gets to be an "Internet food service worker" in the analogy, but there are certainly different levels in the Internet: my aunt with her WebTV box, an ISP with 1000s of dial-up users and very few servers, a full service hosting company, a backbone provider, etc.
Right now, there isn't really a strong concensus on who should be responsible for what, and there are few sanctions against those who would shirk their responsibilities anyway. But the law in question externally imposes regulations on usage and responsibility, and that worries me. It makes everyone responsible back to a single standard. A better approach might be a delegation model of responsibility: If someone offers connectivity to another, it comes with a contract stating what responsibilities stay with the provider and which are delegated. This way, an ISP could choose to force all their users to configure their systems securely, or the ISP could elect to let users do whatever they want and accept responsibility for making sure their users behavior doesn't 'leak out' in violation of their own upstream connectivity agreements.
Let's take a look at the analogy to real-world viruses. Everybody could be required to take a class in basic precautions. Then, we would be required to take the usual precautionary steps at all times: cleaning, face mask, gloves, condom etc. If you pass a virus to another person, you are liable for fines up to $15,000 -- more if it's an incurable disease. (We could call it "the burden of breathing.")
What's really wrong with this scheme is exactly the same thing that's wrong with the "Burden of Connecting" suggestion. Sure, we'd love to prosecute the guy who comes to the office sick and infects everyone, but really we don't expect everyone to be disinfected to the standards of a health care worker at all times. We should have similar expectations for computer viruses: sometimes one of your friends will catch one, and if you aren't careful you may get it too. But we expect the professionals to contain major outbreaks and to have much higher standards for cleanliness.
Serving WML really involves three major components: the phone, a WAP gateway, and an HTTP server.
The phone has a conversation with the gateway using the WAP protocols over some bearer network. That conversation may be either session or non-session oriented, and may be either secure or insecure. Your carrier's GSM network is one option for the bearer; a dial-up PPP connection is another. (If your phone is using a dial-up connection, it will need to connect with a WAP gateway using UDP ports 9200-9203.) In any case, the phone does not actually open an HTTP connection directly to your HTTP server; the connection the phone sees is to the gateway.
The WAP gateway makes the actual connection with the HTTP server. It's basically a proxy server. The WAP gateway may open a separate HTTP session on behalf of each phone, or it may open a new session for every request.
I'm not aware of any WAP enabled phones that store their own cookies. However, the gateway may support cookies on behalf of the phone, though some will downgrade persistent cookies to session cookies. Even if the gateway stores persistent cookies, there's still the problem that the phone may not always use the same gateway, so the persistent cookie store may not be available.
The result of all this is that your HTTP server may or may not see session cookies, and persistent cookies may or may not really persist. About the only thing that is pretty reliable is that you'll get lots of requests from different users using the same gateway IP.
One possibility to deal with missing session cookies is to use URL mangling. This is discussed at some length in another thread, so I won't go into the technique here. However, there are a couple caveats:
The maximum length of a URL on a phone may be significantly less than for a web browser. Even if you don't hit the max URL length, long URLs still increase the size of your HTTP response. It's pretty easy to send back more than a phone can handle (some choke on 1500 bytes or so.)
The gateway transcodes the WML and WMLScript text into a binary format for transmission to the phone, and most commercial WAP gateways do some caching. URL mangling makes caching more difficult, since the mangling introduces small differences between otherwise similar responses. This problem is compounded by the fact that some of the gateways still use HTTP 1.0.
The controversy over the.tv domain goes back some time. Here's an article published in the September 98 issue of Wired. According to the article, Tuvalu has been a bit slow selling the TLD partly because they've been burned in the past selling phone numbers and passports.
Also, someone at WebTV registered themselves as the administrator for.tv some time before this article was published and rather irritated the officials in the country of Tuvalu. Interestingly, Microsoft has since then been quite prominent in the efforts to propagate the tv: protocol designation.
I know many of us would prefer that Amazon not hold the patents they have, but the fact is they do and intend to use them to defend against their direct competitors. The question then becomes: "How can Amazon protect its fiscal interests without getting the rest of us caught in the crossfire?"
Bezos can show that he is truly serious by announcing that Amazon will not now or ever enforce that one click patent. He can keep and enforce any other patents he has.
Or he could enforce it, but offer to license it for $1. Amazon controls the licenses and can withhold it from direct competitors (like B&N) while still allowing everyone else to use it.
Alternately, Bezos can announce that he will only enforce Amazon patents for three years. After which anyone can use them.
He could do this by offering $0 or $1 licenses to all comers in three years.
There's a really simple way for Bezos to demonstrate that Amazon isn't going after the little guys: offer to license the patents to them for extremely cheap (like $1 or so). Licensing patents is perfectly legal and, if anything, would strengthen the case against those who haven't paid licensing fees. The law doesn't require Amazon to license to all comers, or to offer the same terms to everyone. Amazon can establish or decline commercial relationships with whoever they want. Bezos can put his money where his mouth is. Who knows, they could even make a profit at something...
Slashdot Mirror
No, humans aren't 100% and yes, you can test for that. Try a thought experiment: fill a bin with 50,000 red balls and 50,000 blue balls. Ask a human to sort them all. The result probably won't be 100%, but you can still check the result and figure out how accurate the human is without relying on a superhuman ability to tell the balls apart. Same thing for spam: if you start with a known training set, you can test humans to see how well the spam is identified by manual sorting.
How do you know there have been no false positives? Are you reading your spam?
Business units always pay the costs of all company services. Unless you work for the government, there's no other alternative. The only variable is at what level the costs are allocated: the crudest form is just below the CEO level, where all IT costs are rolled up en masse against the income across all business units. At the other end of the spectrum, the lowest level of business unit accounts in nickels and dimes for every byte and packet used. Rationality is somewhere in the middle.
This is commonly referred to as "Activity-Based Costing". It means that the charges are based on allocating the entire cost of the IT department back across some measure of what the IT department actually does. Since the business units can manage their needs more quickly than the IT department can alter their ability to provide, what happens is businesses pull back their demands for IT services, so the allocated costs for individual services goes up, and the businesses respond by trying to reduce demand still more. Instead, there's just a big incentive for the business units to spin up their own capabilities to do things cheaper outside the IT department. If the IT department responds as a monopoly provider with obstructionist rules, changing chargeback structures, access charges, their relationship with the business units degenerates still more.
In the long run, what this leads to is both the desire and the ability for the parent company to outsource the IT function competitively. That may well be what the upper management of your company actually wants to do. It's certainly a sign that you are headed in the opposite direction of the partnership approach. What you really need to do is sell your services at a level that makes sense to the business, so that neither side gets surprised by the costs and everybody understands what they're actually going to get in exchange for the expense. Then you've got a basis for deciding what makes sense to do internally and what makes sense to outsource.
In the past, some companies (such as Oracle) have offered licensing terms based on the number of CPUs. So, a 16-way zSeries running, say, 10,000 Linux images could really bring down the average cost for licensing, couldn't it? Of course, 10,000 Oracle instances would probably kill this box no matter how superior its I/O channels are, but it does present interesting options for an ASP offering.
Interesting stats from the article: 200 million people use Passport as a sign-on service, but according to the first paragraph on the second page of the article, only 2 million users have created "wallets". So their uptake rate is pitiful.
The next stat I'd like to see is how many of the 2 million have actually (and voluntarily...) performed a financial transaction using the thing. And how many of those were just people who created the stupid wallet so they could do business with one of the 70 sites that MS has signed up.
BTW, I just spent a few minuts nosing around the passport site, and I could not see any information (at least, without signing in) about how one might go about deleting a wallet or an entire account.
Several years ago, the Wall Street Journal published an interview with one of the Vice Presidents at Hormel about the way Spam was generally disparaged. This was before the Internet usage of the term really became mainstream, but long after the Monty Python skit. The VP mused that they didn't spend much money advertising it, but they sure sold a lot of the stuff. It looks to me like they do an interview or press release every so often, and get as much mileage out of the whole thing as they can. Seems like a good trade to me--we abuse their trademark, and they sell random pig tidbits in a can.
Two daughters. 6 & 3.5
I wouldn't panic about the high prices just yet. It's pretty common for companies to price the "engineering" copies of their devices pretty high. They're expensive to make, and are intended for early developers who need them to build commercial products. The high price helps cover the high initial costs and, perhaps more importantly, helps ration the devices out to people who are serious about building profitable products.
The more interesting number is the $8/unit for quantity > 1 million. That's not quite at the $5 target, but it's at least in the ballpark.
Not going to help. I actually have sent log files to them. No response whatever.
BTW, @Home portscans seem to resove back to machines with names like authorized-scan1.security.home.net, and lately they've been hitting NNTP ports pretty frequently. It seems they're more interested in making sure I don't violate my agreement not to run servers and soak up their upstream bandwidth than they are in actually securing their network.
OK, I'll throw in the 26 cents.
And hey, the first five miles of towing are free.
The fact of the matter is that by importing large number of H1B workers into the US, we are surpressing the natural rise in wages that would occur in the presence of a real labor shortage.
We are also (partially) supressing the natural exporting of jobs that occurs in the presence of cheaper sources of labor outside the United States.
Actually, Sony may know something about the Internet that most /. readers don't think about. Don't worry about what Sony does to your box. Look at the other threats:
"we will block it at your cable company". Sony controls much of the content cable companies need. Sony also makes lots of head end gear. They may already have some equipment at your cable head end that can at least block cable modem users. Take a look here for some of their DTV products. The cable companies hate what they consider piracy; they aren't going to go to bat for you.
"we will block it at your phone company" Your phone company is trying to get you to use DSL, and needs media content to make it work. Napster helps them in the short run, because users want more bandwidth. But if the phone company wants to get direct revenue from media (like VOD, for example), they're going to have to play ball with the content owners. And remember, there aren't that many phone companies to deal with, and they aren't likely to want to fight Sony on your behalf.
"we will block it at your [Internet-service provider]" How about if they are able to persuade AOL to block it? Probably not that hard.
So if AOL users go away in the short run, and broadband users go away in the short and the long run, they've at the very least bought a couple more years to crush this sort of thing.
I'd say nCube is in a pretty closely related industry. They build servers to stream high quality digital video. (They also build equipment to insert ads into TV shows, but that's a little different.)
If you think about it, the set-top boxes that receive digital video aren't all that different from game consoles (or Tivo boxes or WebTV boxes, for that matter). A future Nintendo game console could potentially receive streaming video--all it really takes is a demodulator (cable, satellite, etc.), an MPEG-2 decoder (HW or fast SW) and the right conditional access equipment. To keep up with Sony, Nintendo might even have to build or acquire a server that competes directly with nCube. Sony already makes lots of cable head-end equipment, set top boxes, and game consoles.
I can't imagine that Nintendo would limit their ability to compete in the video market over something this easy to avoid. Look for a totally different name.
VBIs? Don't be silly--digital can just send the alternate versions in separate streams.
Life is scary when your local broadcaster can sell that ad space separately. Life is really scary when your settop box causes you and your next-door neighbor to get different images.
"Mommy, how come Jennifer Aniston always drinks Bud Light on our TV and Diet Coke over at Bobby's house?"
"spending time playing with the OS because it's cool" makes Linux cost-prohibitive. ... and no matter how much you play with it, Windows just isn't going to get cool.
You will be dealing with individuals. And usually, the first person dispached on a problem call is the lowest-ranked, -paid, and -trained. And a large percentage of such, if they can't resolve the problem, will resort to finger-pointing instead of calling their next level support.
That's exactly the key to dealing with this problem. I used to deal with a lot of escalations for a big company that offered services (not IBM). One of the things that happened several times was the client started screaming "we're paying $XXX/hr to you, and we expect it to be fixed now!" So the junior person panics and starts pointing fingers. The whole exercise adds tension, confusion, and bogus data to the mix. To avoid this, ask the person to do two things:
Yes, there are different levels of cleanliness expected of different people. I'm not ready to say who gets to be an "Internet food service worker" in the analogy, but there are certainly different levels in the Internet: my aunt with her WebTV box, an ISP with 1000s of dial-up users and very few servers, a full service hosting company, a backbone provider, etc.
Right now, there isn't really a strong concensus on who should be responsible for what, and there are few sanctions against those who would shirk their responsibilities anyway. But the law in question externally imposes regulations on usage and responsibility, and that worries me. It makes everyone responsible back to a single standard. A better approach might be a delegation model of responsibility: If someone offers connectivity to another, it comes with a contract stating what responsibilities stay with the provider and which are delegated. This way, an ISP could choose to force all their users to configure their systems securely, or the ISP could elect to let users do whatever they want and accept responsibility for making sure their users behavior doesn't 'leak out' in violation of their own upstream connectivity agreements.
Let's take a look at the analogy to real-world viruses. Everybody could be required to take a class in basic precautions. Then, we would be required to take the usual precautionary steps at all times: cleaning, face mask, gloves, condom etc. If you pass a virus to another person, you are liable for fines up to $15,000 -- more if it's an incurable disease. (We could call it "the burden of breathing.")
What's really wrong with this scheme is exactly the same thing that's wrong with the "Burden of Connecting" suggestion. Sure, we'd love to prosecute the guy who comes to the office sick and infects everyone, but really we don't expect everyone to be disinfected to the standards of a health care worker at all times. We should have similar expectations for computer viruses: sometimes one of your friends will catch one, and if you aren't careful you may get it too. But we expect the professionals to contain major outbreaks and to have much higher standards for cleanliness.
Serving WML really involves three major components: the phone, a WAP gateway, and an HTTP server.
The phone has a conversation with the gateway using the WAP protocols over some bearer network. That conversation may be either session or non-session oriented, and may be either secure or insecure. Your carrier's GSM network is one option for the bearer; a dial-up PPP connection is another. (If your phone is using a dial-up connection, it will need to connect with a WAP gateway using UDP ports 9200-9203.) In any case, the phone does not actually open an HTTP connection directly to your HTTP server; the connection the phone sees is to the gateway.
The WAP gateway makes the actual connection with the HTTP server. It's basically a proxy server. The WAP gateway may open a separate HTTP session on behalf of each phone, or it may open a new session for every request.
I'm not aware of any WAP enabled phones that store their own cookies. However, the gateway may support cookies on behalf of the phone, though some will downgrade persistent cookies to session cookies. Even if the gateway stores persistent cookies, there's still the problem that the phone may not always use the same gateway, so the persistent cookie store may not be available.
The result of all this is that your HTTP server may or may not see session cookies, and persistent cookies may or may not really persist. About the only thing that is pretty reliable is that you'll get lots of requests from different users using the same gateway IP.
One possibility to deal with missing session cookies is to use URL mangling. This is discussed at some length in another thread, so I won't go into the technique here. However, there are a couple caveats:
WAP devices do support client side scripting, with a language called WMLScript. It's essentially a stripped down version of ECMAScript.
The controversy over the .tv domain goes back some time. Here's an article published in the September 98 issue of Wired. According to the article, Tuvalu has been a bit slow selling the TLD partly because they've been burned in the past selling phone numbers and passports.
Also, someone at WebTV registered themselves as the administrator for .tv some time before this article was published and rather irritated the officials in the country of Tuvalu. Interestingly, Microsoft has since then been quite prominent in the efforts to propagate the tv: protocol designation.
I know many of us would prefer that Amazon not hold the patents they have, but the fact is they do and intend to use them to defend against their direct competitors. The question then becomes: "How can Amazon protect its fiscal interests without getting the rest of us caught in the crossfire?"
Or he could enforce it, but offer to license it for $1. Amazon controls the licenses and can withhold it from direct competitors (like B&N) while still allowing everyone else to use it.
He could do this by offering $0 or $1 licenses to all comers in three years.
There's a really simple way for Bezos to demonstrate that Amazon isn't going after the little guys: offer to license the patents to them for extremely cheap (like $1 or so). Licensing patents is perfectly legal and, if anything, would strengthen the case against those who haven't paid licensing fees. The law doesn't require Amazon to license to all comers, or to offer the same terms to everyone. Amazon can establish or decline commercial relationships with whoever they want. Bezos can put his money where his mouth is. Who knows, they could even make a profit at something...