Parent post makes a lot of sense. Think about it- can you even _buy_ a disk drive smaller than 40 gigs right now? maybe on eBay, but that's no way to run a business. If your typical dvd-quality movie is, say ~10gigs, including letterbox/fullscreen formats and the voiceover commentary by Tom Cruise and Xenu, there is no financial reason preventing Netflix from having 4 or 5 movies available for download at any given time.
The real test here is: will there be any kind of p2p functionality to offset heavy demand for new releases? jump ahead 5 years... 'Batman: the pre-school years' is released on DVD, and 10 million people all click "download" in the same 24-hour period. If nflx has peering set up, the first couple of batnerds have a regular download, and everyone else's dl gets progressively faster as more people have the whole movie available for peering... If they don't, each of those 10 million people has to wait their turn in line...
parent also makes the point that this could be faster for the customer than waiting for the USPS to ship dvd's back and forth. It's also CHEAPER in a real sense for neftlix, saving them round-trip postage and sorting/handling costs for incoming or outgoing dvd's.
Deserves? No. Expects? well, that's another matter entirely. If you've already confronted the founder of the company and pointed out that the business practices that he himself WROTE THE CODE to perform are, in fact, illegal... and he didn't seem shocked and appalled by that news... then odds are, he's already well aware of the legal status of his company's activities, he's a (wealthy? powerful?) unscrupulous bastard anyway, and won't mind squashing you like a bug if it keeps things just the way they are.
honestly, people- if your company's financial success is built on illegal behavior, and the guy who owns the majority of the company set it up to be that way, why would you think he's going to change anything just because you were bright enough to notice? The best you could hope for is that, when you try to blackmail him into splitting the profit with you, it's not just cheaper to have you killed.
Picture this: Vince Coll walks into Dutch Shultz's office.
Vince: Hey Dutch, I've been thinking. Dutch: That's a dangerous habit. Vince: You know, this Prohibition thing? That makes alcohol illegal right? Dutch: Yeah. Vince: So, all this beer we're selling, that's illegal too, right? Dutch: Yeah. That's why we make so much money. Vince: Well, I don't feel so good about breaking the law. Dutch: I don't like where you're going with this. Vince: I think maybe we should, you know, stop? Dutch: Don't make me shoot you in my office.
I can understand why Chip had the moral problems that he did, but he sure picked a naive way to try and resolve them. There IS a federal whistleblower statute, so if he went to the Feds with his first letter he would have been legally protected from retaliatory action from his employer... but keeping your job after you've turned them in doesn't do you any good if the company's only revenue stream depends on illegal activities.
To all the kids who took out $150,000 in student loans so they could afford to go to Stanford, and then graduated to hear Steve Jobs say "Dropping out of college was the best thing I ever did..." allow me to be the first to say (in my best Nelson voice)
HA HA!
welcome to the real world, suckers. The only college whose graduates I respect less than Stanford is Princeton. Maybe Steve Jobs will hire you, but I certainly won't.
I personally think we should forget about electronic voting and go back to only using paper ballots that are then hand counted.
sure, I'll agree with that. Nothing wrong with paper-only, except that we end up with goofy ballots like the butterfly design, where names don't quite line up with their respective dots.
the theory of electronic touchscreen machines is a good one- make it easy to select, in a standardized and unambiguous way, which side of an issue to vote for. I think that potential for standardization goes a long way towards making nationwide elections more uniform from precinct to precinct and state to state.
The whole point is that you don't have to have anyone counting the ballots, hopefully removing at least some human error from the process (losing count, miscounting, ballots getting stuck together, etc). Automated talliers that read scantron or punch cards do the same thing: they try to make it faster, with less chance of human error.
sorry, but I don't agree with you here. The point of using a machine might be to get a result more quickly, but the point of having an election is to count the vote the way the voter intended. Sacrificing the latter in the name of the former is throwing out the baby with the bathwater. If, somewhere in the chain from voter's hand to final tally, there is a black box - an unauditable link in the chain - it's not possible to ensure that someone isn't monkeying with the numbers. In fact, that's EXAXTLY what happened this time, in a lot of counties in Florida, and at least a few in Ohio- exit polls said one thing, official tally said a different thing, a recount was requested... and guess what? there's nothing to count! It's all just bits on a memory card, and if someone changed those bits before they were counted the first time, it's still fraud even when the bits haven't changed when you look at them the second time around.
The words we need to use are "Voter-verifiable paper BALLOT." You don't want the electronic record on the machine to be the official vote, and a little piece of paper for a souvenier; you want the goddamn piece of paper to be the official vote, so that you can COUNT THEM BY HAND if there are any questions. This is the only way to make sure that some invisible bits aren't being stolen/moved/dropped, people: THE ONLY WAY.
1. mark your choices on touchscreen machine
2. at the end of ballot, machine shows you your vote tallies for you to approve
3. you click "OK" or "CHANGE"
4. if you click "OK," the machine prints a BALLOT. This is the official ballot. It's written in plain english/thai/spanish/korean/kanji so that you can READ IT WITH YOUR EYES and you don't have to depend on a machine to tell you what it says. You take this printed BALLOT and you confirm that the votes marked on it are the votes you wanted to cast. If they don't match, you tear it up and start over, or you know there is a problem with the machine. If they do match, you put it in a BALLOT BOX, and that is your vote. No keycards, no electronic tallies, no modems to transmit secret electronic results. Your vote, on a slip of real paper, goes into a box, with thousands of other slips of paper, guaranteeing anonymity. You don't take anything away- becuase you don't need to- you've already verified that the machine printed your voter-verifiable paper BALLOT correctly, and then you voted with your BALLOT.
5. at the end of the day, people open the boxes, and count the votes.
that's it. Yes, it seems primitive. Yes, it will be slower, and it will take more work- but when you consider that the alternative is that we keep going like we are today, where Chuck Hagel can decide the outcome of half of the nation's elections, don't you think it's worth a little bit more work to make sure that election results actually reflect the intention of the voters?
yeah, WTF right? I mean, yes, the main body of my post is is slightly "offtopic" I guess, and I'd even entertain "redundant" because other posters have made the Dvorak = MoG Fan Club comparison previously. But I was responding to a post that someone else made, pointing out a flaw in his argument. I thought that was the whole point of this threaded-discussion thingy. I think it's fair to say you can't lump MoG's stalking/libel of PJ in with the NewsWeek story retraction, because the NewsWeek story was TRUE.
I'm sure that FBI_Mod_Troll is laughing into his sleeve right now.
One of the problems with your example is that Newsweek wastellingthetruth. Likewise, Bush really was AWOL, and Dan Rather was destroyed becuase he used his pulpit to point out a controversy that was politically inconvenient to the company signing his paychecks.
The cases of MoG and Jayson Blair are fundamentally different from events surrounding Rather's resignation or the retraction of the claims in the Newsweek article. Rather and Newsweek ran stories about events that were based on fact- that is they actually happened in the real world- while MoG and Blair just MADE STUFF UP. MoG carried it a step further and engaged in ad-hominem attacks on the subject of her reporting- behavior that is totally unprofessional, uncalled-for, and possibly actionable.
As the Gannon/Guckert insanity shows, it's OK to make stuff up and pass it off as fact, as long as it's an echo chamber for whatever Rove told McClellan to say. MoG's strategic failure isn't that she's reporting lies as fact, it's that she doesn't have a bunch of religious whack-jobs employed by ClearChannel to repeat her lies as if they're the truth. She does have a network of MS apologists and astroturfers who copy/paste her specious claims, and Dvorak is one of those guys.
The media has 3 audiences, and are held accountable to two masters: the audiences are (1) sheep who believe what they're told, (2) people who are willing to read past the headline and actually question the news being presented to them, and (3) the people who made the news happen and want to see it reported a certain way so (1) and (2) can know how cool/important/rich/dangerous (3) is living. The masters are (A) the corporate entities who sign the paychecks, and (B) to a much lesser extent, the news-reading public composed of (1) and (2) who vote with their eyes and dollars.
The problem with news in this country today is that group (3) and group (A) are increasingly the same people, using the news media to influence the opinions of groups (1) and (2). We have a name for this kind of media communication- it's called advertizing.
Or marketing. And when you see an ad, the important questions to ask are: "who is paying for this?" "What are they selling?" and "how much did this ad cost?" So those are the questions I (and I think many other folks who have had opportunity to appreciate what PJ is doing with Groklaw) would like to ask MoG, preferably after a subpoena and under oath.
I had Xerox pitch me on a system to do this very thing less than 2 months ago. I don't know what it's called, but I'm sure if you called and asked them they would be thrilled to sell it to you.
Thank you for blocking all ports besides 25, 56, and 80. You have given me the opportunity to learn how to set up and utilize OpenVPN software across a large private network, and the opportunity to develop a loyal client base from my fellow students who wish to continue using BitTorrent but find their normal ports blocked.
Please disregard what appears to be an extremely high volume of encrypted DNS traffic coming from my dorm room, as it is probably just an error in your reporting tools.
Clearcube does this now. They've been doing it for years. They're getting a lot of traction lately because of HIPAA. It's much easier to ensure HIPAA compliance when your computer terminals in a public space don't have any local data storage hardware or active USB ports.
the thing that prevents Clearcube from rolling it to a home use environment is that the break-even point for when it's cost effective to buy a rack and infrastrucuture to support the headless servers in the back room ends up being equal to the cost of eight or ten brand new standalone PC's... and I don't know how many family members with laptops YOU have, but I don't know anyone who has eight state-of-the-art PC's in their home.
disclaimer: I have at least six or seven PC's at home now, but one of them is a P3600, one is a Toshiba Pentium Pro laptop, one is my wife's work laptop... etc etc etc.
It's too bad this is happening, but give it some time- evolution will stop them from distributing the product without the source, becuase Iomega is a dead company that just doesn't know it's dead yet. The best thing they could do for their investors is close up shop, fire everyone, and start selling off their assets. I mean, come on, Zip disks? That was a great idea, 10 years ago when CD burners cost $500 and blank CD's were $5 each.
Now the zip disk format is the storage equivalent of the green-screen VGA monitor. The world has moved on to better things, and Iomega is stuck in 1992.
I got an email from my mother this morning. She tells me since yesterday morning, people from mortgage companies have been calling her at work, supposedly in response to her request for a quote. She asks them where they got her name, and they tell her "you signed up on a website" (her words, not mine).
She's all convinced that ChoicePoint sold her out, but she lives in Michigan, where there are no disclosure laws to protect citizens. Anybody know how she can find out if her data were sold to these fraudsters?
are there really THAT many people out there that pay the monthly fee, rather than the one time lifetime sub. fee?
Well, yeah. at $13/month, a $300 lifetime sub costs the same as 2 years of service. If you buy a box today and spend that $300, you're basically making a bet that Tivo is still going to be around in 2 years and operating its business in a substantially similar way.
I'm paying monthly, becuase I've done the math. The math says: tivo lost $32 million last year, and is sitting on assets of $63 million. So, at current subscription rates, Tivo has just enough assets to sustain themselves for two more years, if they lose as much as they did for the next two years in a row.
'but wait,' you say! 'Tivo was cash-flow positive last year, for the first time EVER! This is a mark of them moving towards a profitable business!'
unfortunately, this is only becuase they raised $75 million in Q104 by selling stock. The company is HEMMORHAGING money.
yes, I'm just trying to justify my monthly fee. but I want to be clear that there are two camps- people who don't do the math, and think that $13/month forever is actually cheaper than the $300 up front... and people like me, who think Tivo will go out of business before they are able to provide me with the monthly equivalent of $300 worth of service.
I'm not at all sure this is true- I'm running Trillian on the MS messenger network from a corp environment that blocks everything except ports 25, 56, and 80. I'm pretty sure messenger doesn't run over DNS.
Ok, let's say you manage an IT department responsible for 100 desktop/laptop computers, and you have to buy all new machines. Where do you begin? Where can you gain efficiencies from standardization?
The best-case scenario is to make your org as much a monoculture as possible. If every machine is the same, you don't waste time trying to figure out whether it's the Jan2005 version of that Fujitsu hard disk that's giving you trouble, or if it's likeley to be something else becuase there have never been problems with the Nov 2004 Fujitsus. If every machine is the same, you don't have to troubleshoot how your new netword drivers cooperate with the code coming from your legacy AIX boxes running your custom-coded business database application.
Next best is if you have slightly different hardware, but just one software image, and regular automated backups. If something goes wrong on one box, it will be wrong on all boxes. If something is funny with just one box, you can fix it by wiping it clean and re-installing the original image, and then dragging custom files over from the backup.
of course, in normal businesses, this isn't possible. So you have maybe 3 kinds of boxes, with one image each, that you give to 90% of the folks you support. This leaves you with ten people who have an oddball machine, for an effective total of 13 different computers for you to manage (and you just hope that those ten 'special' people won't take up 50% of your support time).
Ok, now think big for a second. Think about managing an organization with 100,000 PC's. Now think about buying 100,000 BRAND-NEW PC's to replace the boxes you have today. You have the same basic desires- minimize hardware differences, consolidate images, standardize across the userbase. But if you call Dell today, I bet you a million bucks that they won't be able to ship you 100,000 identical PC's all at once- not tomorrow, not next month.
So what happens when you tell Dell you want to order 30,000 of PC model A, 40,000 of model B, and 25,000 of model C? They'll ask you for a roll-out schedule, and a list of delivery addresses, and give you a product roadmap for when they plan to run out of P4 3.2 GHz chips and move up to P4 3.8GHz chips, so you can sync your roll-out plans to their product roadmap. But that's not the point of the higher price for business. The higher price is necessary because you want a single hardware/software image, so when you order 95,000 PC's they have to order 95,000 of the same ethernet card to make sure that they can provide you with a stable hardware image. They pay up front for the components that go in your boxes- but you don't pay for your boxes until they ship (or even better, you don't pay until they have landed on your dock and you've had a chance to look them over). So Dell has to pay for inventory costs, and for the float on the money they used to buy your Ethernet cards ahead of when they can bill you for the computers those cards have gone into.
Contrast this with an order for 100 PCs- Dell just builds them all and ships them to you, because they have 100 of each of the relevant parts in stock.
now for the really complicated situation, which explains the different prices: Consider buying 1 PC per day, each day, for 100 days, from the consumer/small business line. If you open all of those computers and look at the parts, odds are good that several hardware components will have changed between the first order you place in January and the last one you place in April- for these lines, Dell just builds with whatever parts they happen to have on hand that day. The business-oriented products line is more expensive becuase Dell buys a large inventory of identical parts, and is prepared to commit to its large business customers that the hardware won't change more than (e.g.) once every six months. The argument is that the higher up-front cost of the equipment is more than offset by the lower overall cost to the organization that comes from managing just a few hardware configurations instead of 100,000 different PC's.
and this is a great argument, except it falls through when you realize that if you are in fact planning to buy 100,000 PC's, from ANY tier-1 vendor, you won't be paying anything near list price for them. Oh well.
Nothing is stopping customers from doing exactly what you propose, with their contracts.
and all it will take is a critical mass -say, 10% of the total healthcare market in the US- to take this stance, and you'll see it start to happen. look at this for a hint of what's going on...
And as I'm sure you know, there are different flavors of DICOM produced by different vendors. Last time I checked, Siemens DICOM doesn't play nice with GE DICOM. Yes, there are standards, but they're GOVERNMENT standards, not customer standards. They all have loopholes big enough to drive a truck through, and the vendors exploit these loopholes to lock customers into a one-vendor package.
If you are a Siemens sales guy, which one is better for you- a Siemens patient monitor that listens to a GE pulse ox, or a Siemens monitor that only works correctly with other Siemens equipment?
All the vendors make stuff that works. It just doesn't always play nice with the other kids. Standards compliance *on paper* is worthless if the box doesn't work with your other stuff when you plug it in. Publishing another set of standards won't fix this situation unless customers have a uniform, objective test for interoperability, and obtain the contractual right to RETURN a system that fails this test to the vendor for full refund (and some $$$ penalty for the inconvenience of being a guinea pig)...
An organized national health care system would produce "reference systems" for components of the OR suite, and provide them to the vendors with the understanding that if the vendor wanted to sell anything, their products would need to successfully interoperate with the reference system. Fortunatley for the continued financial well-being of GE and Siemens, the health care system in the US is about as far from organized as you can get.
To reply to parent's parent's parent's post- the issue is not standards. The issue is ENFORCEMENT of standards BY CUSTOMERS rather than by the government. HL7 was written by the vendors so that customers can't use "standards compliance" to change the market dynamic. DICOM was written to fix/extend HL7, but didn't change the approach. You can write RFC's all day, and turn them into a standard if you want, but the real problem is that to drive change in the market for healthcare devices, you need to take power from the vendors and put it into the hands of the customers, and the only way to do that is with contracts that carry financial penalties for the vendor if they fail an objective interoperability test.
Slashdot Mirror
you say IM is for conversation, email is for documentation.and several similar comments.
I log all my IM conversations: user, time and date. don't you? If you're using Trillian, you're doing it automatically.
The real test here is: will there be any kind of p2p functionality to offset heavy demand for new releases? jump ahead 5 years... 'Batman: the pre-school years' is released on DVD, and 10 million people all click "download" in the same 24-hour period. If nflx has peering set up, the first couple of batnerds have a regular download, and everyone else's dl gets progressively faster as more people have the whole movie available for peering... If they don't, each of those 10 million people has to wait their turn in line...
parent also makes the point that this could be faster for the customer than waiting for the USPS to ship dvd's back and forth. It's also CHEAPER in a real sense for neftlix, saving them round-trip postage and sorting/handling costs for incoming or outgoing dvd's.
honestly, people- if your company's financial success is built on illegal behavior, and the guy who owns the majority of the company set it up to be that way, why would you think he's going to change anything just because you were bright enough to notice? The best you could hope for is that, when you try to blackmail him into splitting the profit with you, it's not just cheaper to have you killed.
Picture this: Vince Coll walks into Dutch Shultz's office.
I can understand why Chip had the moral problems that he did, but he sure picked a naive way to try and resolve them. There IS a federal whistleblower statute, so if he went to the Feds with his first letter he would have been legally protected from retaliatory action from his employer... but keeping your job after you've turned them in doesn't do you any good if the company's only revenue stream depends on illegal activities.
You just gave me goosebumps on a Saturday morning. Thanks!
To all the kids who took out $150,000 in student loans so they could afford to go to Stanford, and then graduated to hear Steve Jobs say "Dropping out of college was the best thing I ever did..." allow me to be the first to say (in my best Nelson voice)
HA HA!
welcome to the real world, suckers. The only college whose graduates I respect less than Stanford is Princeton. Maybe Steve Jobs will hire you, but I certainly won't.
sure, I'll agree with that. Nothing wrong with paper-only, except that we end up with goofy ballots like the butterfly design, where names don't quite line up with their respective dots.
the theory of electronic touchscreen machines is a good one- make it easy to select, in a standardized and unambiguous way, which side of an issue to vote for. I think that potential for standardization goes a long way towards making nationwide elections more uniform from precinct to precinct and state to state.
The whole point is that you don't have to have anyone counting the ballots, hopefully removing at least some human error from the process (losing count, miscounting, ballots getting stuck together, etc). Automated talliers that read scantron or punch cards do the same thing: they try to make it faster, with less chance of human error.
sorry, but I don't agree with you here. The point of using a machine might be to get a result more quickly, but the point of having an election is to count the vote the way the voter intended. Sacrificing the latter in the name of the former is throwing out the baby with the bathwater. If, somewhere in the chain from voter's hand to final tally, there is a black box - an unauditable link in the chain - it's not possible to ensure that someone isn't monkeying with the numbers. In fact, that's EXAXTLY what happened this time, in a lot of counties in Florida, and at least a few in Ohio- exit polls said one thing, official tally said a different thing, a recount was requested... and guess what? there's nothing to count! It's all just bits on a memory card, and if someone changed those bits before they were counted the first time, it's still fraud even when the bits haven't changed when you look at them the second time around.
that's it. Yes, it seems primitive. Yes, it will be slower, and it will take more work- but when you consider that the alternative is that we keep going like we are today, where Chuck Hagel can decide the outcome of half of the nation's elections, don't you think it's worth a little bit more work to make sure that election results actually reflect the intention of the voters?
Fox can't dictate how you watch television. But try recording their lineup, stripping the commercials, and putting them on the Internet
Don't mind if I do!
Girl... friend? Gym? What are these things you speak of?
yeah, WTF right? I mean, yes, the main body of my post is is slightly "offtopic" I guess, and I'd even entertain "redundant" because other posters have made the Dvorak = MoG Fan Club comparison previously. But I was responding to a post that someone else made, pointing out a flaw in his argument. I thought that was the whole point of this threaded-discussion thingy. I think it's fair to say you can't lump MoG's stalking/libel of PJ in with the NewsWeek story retraction, because the NewsWeek story was TRUE.
I'm sure that FBI_Mod_Troll is laughing into his sleeve right now.
The cases of MoG and Jayson Blair are fundamentally different from events surrounding Rather's resignation or the retraction of the claims in the Newsweek article. Rather and Newsweek ran stories about events that were based on fact- that is they actually happened in the real world- while MoG and Blair just MADE STUFF UP. MoG carried it a step further and engaged in ad-hominem attacks on the subject of her reporting- behavior that is totally unprofessional, uncalled-for, and possibly actionable.
As the Gannon/Guckert insanity shows, it's OK to make stuff up and pass it off as fact, as long as it's an echo chamber for whatever Rove told McClellan to say. MoG's strategic failure isn't that she's reporting lies as fact, it's that she doesn't have a bunch of religious whack-jobs employed by ClearChannel to repeat her lies as if they're the truth. She does have a network of MS apologists and astroturfers who copy/paste her specious claims, and Dvorak is one of those guys.
The media has 3 audiences, and are held accountable to two masters: the audiences are (1) sheep who believe what they're told, (2) people who are willing to read past the headline and actually question the news being presented to them, and (3) the people who made the news happen and want to see it reported a certain way so (1) and (2) can know how cool/important/rich/dangerous (3) is living. The masters are (A) the corporate entities who sign the paychecks, and (B) to a much lesser extent, the news-reading public composed of (1) and (2) who vote with their eyes and dollars.
The problem with news in this country today is that group (3) and group (A) are increasingly the same people, using the news media to influence the opinions of groups (1) and (2). We have a name for this kind of media communication- it's called advertizing.
Or marketing. And when you see an ad, the important questions to ask are: "who is paying for this?" "What are they selling?" and "how much did this ad cost?" So those are the questions I (and I think many other folks who have had opportunity to appreciate what PJ is doing with Groklaw) would like to ask MoG, preferably after a subpoena and under oath.
Yeah, but do you have pointy hair? Well, do you?
/just kidding...
Where in San Francisco is this Tenderloin district you speak of?
google maps is your friend. but blowjobs in the tenderloin are probably NOT.
Would that research be the result of 11 years of 1-handed typing?
I had Xerox pitch me on a system to do this very thing less than 2 months ago. I don't know what it's called, but I'm sure if you called and asked them they would be thrilled to sell it to you.
Dear IT Department,
Thank you for blocking all ports besides 25, 56, and 80. You have given me the opportunity to learn how to set up and utilize OpenVPN software across a large private network, and the opportunity to develop a loyal client base from my fellow students who wish to continue using BitTorrent but find their normal ports blocked.
Please disregard what appears to be an extremely high volume of encrypted DNS traffic coming from my dorm room, as it is probably just an error in your reporting tools.
Best regards,
the thing that prevents Clearcube from rolling it to a home use environment is that the break-even point for when it's cost effective to buy a rack and infrastrucuture to support the headless servers in the back room ends up being equal to the cost of eight or ten brand new standalone PC's... and I don't know how many family members with laptops YOU have, but I don't know anyone who has eight state-of-the-art PC's in their home.
disclaimer: I have at least six or seven PC's at home now, but one of them is a P3600, one is a Toshiba Pentium Pro laptop, one is my wife's work laptop... etc etc etc.
Now the zip disk format is the storage equivalent of the green-screen VGA monitor. The world has moved on to better things, and Iomega is stuck in 1992.
I got an email from my mother this morning. She tells me since yesterday morning, people from mortgage companies have been calling her at work, supposedly in response to her request for a quote. She asks them where they got her name, and they tell her "you signed up on a website" (her words, not mine).
She's all convinced that ChoicePoint sold her out, but she lives in Michigan, where there are no disclosure laws to protect citizens. Anybody know how she can find out if her data were sold to these fraudsters?
where was that EULA link again?
Well, yeah. at $13/month, a $300 lifetime sub costs the same as 2 years of service. If you buy a box today and spend that $300, you're basically making a bet that Tivo is still going to be around in 2 years and operating its business in a substantially similar way.
I'm paying monthly, becuase I've done the math. The math says: tivo lost $32 million last year, and is sitting on assets of $63 million. So, at current subscription rates, Tivo has just enough assets to sustain themselves for two more years, if they lose as much as they did for the next two years in a row.
'but wait,' you say! 'Tivo was cash-flow positive last year, for the first time EVER! This is a mark of them moving towards a profitable business!'
unfortunately, this is only becuase they raised $75 million in Q104 by selling stock. The company is HEMMORHAGING money.
yes, I'm just trying to justify my monthly fee. but I want to be clear that there are two camps- people who don't do the math, and think that $13/month forever is actually cheaper than the $300 up front... and people like me, who think Tivo will go out of business before they are able to provide me with the monthly equivalent of $300 worth of service.
I'm not at all sure this is true- I'm running Trillian on the MS messenger network from a corp environment that blocks everything except ports 25, 56, and 80. I'm pretty sure messenger doesn't run over DNS.
The best-case scenario is to make your org as much a monoculture as possible. If every machine is the same, you don't waste time trying to figure out whether it's the Jan2005 version of that Fujitsu hard disk that's giving you trouble, or if it's likeley to be something else becuase there have never been problems with the Nov 2004 Fujitsus. If every machine is the same, you don't have to troubleshoot how your new netword drivers cooperate with the code coming from your legacy AIX boxes running your custom-coded business database application.
Next best is if you have slightly different hardware, but just one software image, and regular automated backups. If something goes wrong on one box, it will be wrong on all boxes. If something is funny with just one box, you can fix it by wiping it clean and re-installing the original image, and then dragging custom files over from the backup.
of course, in normal businesses, this isn't possible. So you have maybe 3 kinds of boxes, with one image each, that you give to 90% of the folks you support. This leaves you with ten people who have an oddball machine, for an effective total of 13 different computers for you to manage (and you just hope that those ten 'special' people won't take up 50% of your support time).
Ok, now think big for a second. Think about managing an organization with 100,000 PC's. Now think about buying 100,000 BRAND-NEW PC's to replace the boxes you have today. You have the same basic desires- minimize hardware differences, consolidate images, standardize across the userbase. But if you call Dell today, I bet you a million bucks that they won't be able to ship you 100,000 identical PC's all at once- not tomorrow, not next month.
So what happens when you tell Dell you want to order 30,000 of PC model A, 40,000 of model B, and 25,000 of model C? They'll ask you for a roll-out schedule, and a list of delivery addresses, and give you a product roadmap for when they plan to run out of P4 3.2 GHz chips and move up to P4 3.8GHz chips, so you can sync your roll-out plans to their product roadmap. But that's not the point of the higher price for business. The higher price is necessary because you want a single hardware/software image, so when you order 95,000 PC's they have to order 95,000 of the same ethernet card to make sure that they can provide you with a stable hardware image. They pay up front for the components that go in your boxes- but you don't pay for your boxes until they ship (or even better, you don't pay until they have landed on your dock and you've had a chance to look them over). So Dell has to pay for inventory costs, and for the float on the money they used to buy your Ethernet cards ahead of when they can bill you for the computers those cards have gone into.
Contrast this with an order for 100 PCs- Dell just builds them all and ships them to you, because they have 100 of each of the relevant parts in stock.
now for the really complicated situation, which explains the different prices: Consider buying 1 PC per day, each day, for 100 days, from the consumer/small business line. If you open all of those computers and look at the parts, odds are good that several hardware components will have changed between the first order you place in January and the last one you place in April- for these lines, Dell just builds with whatever parts they happen to have on hand that day. The business-oriented products line is more expensive becuase Dell buys a large inventory of identical parts, and is prepared to commit to its large business customers that the hardware won't change more than (e.g.) once every six months. The argument is that the higher up-front cost of the equipment is more than offset by the lower overall cost to the organization that comes from managing just a few hardware configurations instead of 100,000 different PC's.
and this is a great argument, except it falls through when you realize that if you are in fact planning to buy 100,000 PC's, from ANY tier-1 vendor, you won't be paying anything near list price for them. Oh well.
Nothing is stopping customers from doing exactly what you propose, with their contracts.
and all it will take is a critical mass -say, 10% of the total healthcare market in the US- to take this stance, and you'll see it start to happen. look at this for a hint of what's going on...
I thought [DICOM and HL7] were distinct...?
And as I'm sure you know, there are different flavors of DICOM produced by different vendors. Last time I checked, Siemens DICOM doesn't play nice with GE DICOM. Yes, there are standards, but they're GOVERNMENT standards, not customer standards. They all have loopholes big enough to drive a truck through, and the vendors exploit these loopholes to lock customers into a one-vendor package.
If you are a Siemens sales guy, which one is better for you- a Siemens patient monitor that listens to a GE pulse ox, or a Siemens monitor that only works correctly with other Siemens equipment?
All the vendors make stuff that works. It just doesn't always play nice with the other kids. Standards compliance *on paper* is worthless if the box doesn't work with your other stuff when you plug it in. Publishing another set of standards won't fix this situation unless customers have a uniform, objective test for interoperability, and obtain the contractual right to RETURN a system that fails this test to the vendor for full refund (and some $$$ penalty for the inconvenience of being a guinea pig)...
An organized national health care system would produce "reference systems" for components of the OR suite, and provide them to the vendors with the understanding that if the vendor wanted to sell anything, their products would need to successfully interoperate with the reference system. Fortunatley for the continued financial well-being of GE and Siemens, the health care system in the US is about as far from organized as you can get.
To reply to parent's parent's parent's post- the issue is not standards. The issue is ENFORCEMENT of standards BY CUSTOMERS rather than by the government. HL7 was written by the vendors so that customers can't use "standards compliance" to change the market dynamic. DICOM was written to fix/extend HL7, but didn't change the approach. You can write RFC's all day, and turn them into a standard if you want, but the real problem is that to drive change in the market for healthcare devices, you need to take power from the vendors and put it into the hands of the customers, and the only way to do that is with contracts that carry financial penalties for the vendor if they fail an objective interoperability test.