Just because you're paranoid doesn't mean you're wrong. While there's no proof that this is what happened, it would be a very clever tactic indeed. And while they often get accused of being greedy, rarely do they get accused of not being clever.
> I happened to like the Replay user interface better, so I got it.
I have a TiVo SA, but I'm thinking of selling it and getting two of the new Replays for upstairs and downstairs. I really like the ability to stream shows from one unit to another within the house, something TiVo is rumored to be working on, but is dragging its feet on forever. Because of this ability, there seems to also be software out there that allows you to extract shows from a unit over the LAN without having to open it up and hack it. Incidentally, if anyone knows, how many ReplayTV units can cooperate on sharing shows with each other? Just two, or can you browse any number of units?
I am quite convinced that "they" have every intention and hope for this service to fail. If (when) it fails, it gives them just another reason to say "see, we told you LEGAL online movie distribution doesn't work--people want to STEAL, not BUY!"
If they truly wanted this to succeed, they would have made it genuinely attractive and competitive with DVD. They would have admitted to themselves that the obvious downsides to this service are the excruciatingly long download times and the need to watch the movie on the PC (or to have a HTPC), so they would have tried to offset those disadvantages by making rental temptingly cheap and offering longer rental times. They would maybe rent the movies for $1.99 and let you watch them for a week or more. And why on earth not? They're not limited by a restricted number of media that they need to circulate amongst a large clientele, unlike your video store. And they have only (relatively) minor adminstration costs, and except for the download bandwidth, no recurring maintenance costs per film (rewinding/applying security tag/reshelving etc.)
But if this succeeded, it would gradually erode the very lucrative video store business, and it would open the general consumer's eyes even more to the possibilities of digital media. Worse, it would force Hollywood to face the fact that their old business model is dying.
No, I honestly think they're either using this to prove that people aren't interested in acquiring online movies legally, or possibly as a half-hearted safe bet just in case this "online thing" should turn out to be more than just a fad.
> It sounds really harsh, but it's not intended to be that way.
Well, maybe that's got something to do with the vaunted German efficiency: efficiency through not beating around the bush and calling a spade a spade. Who knows. Then again, it wasn't in Germany where my wife hung up on a rude telemarketer, only to have him call back to tell her that he thought she was a real bitch. Click. No, it was back here in the good You Ass Of A...
> Of course, you could write your own DHCP server too
That would be the most flexible solution long-term, especially if they want to do all sorts of custom lookups during a DHCP request. It's quite a trivial protocol, we implemented a custom DHCP server a few years back in a couple of days. Just look up the RFCs for BOOTP and DHCP at www.ietf.org, making sure to use the latest ones since there were multiple revisions.
I have a stash of about 1000 CDs bought in the 80's and early 90s, and I haven't bought a new CD in years. I have plenty of music to listen to for years to come, and all the CDs happily rip to MP3. Besides, good music died along with the 70s. All the new stuff coming out is just new tricks, and I'm such an old dog.
Exactly, in fact PDF and Postscript are very similar. When you render text to PS, you can end up with low-level drawing primitives such as lines and curves defining each letter, rather than high-level instructions such as "draw this string at position x,y". Once you've done that, recovering the original text amounts to highly sophisticated shape recognition and is impossible for all practical purposes. Precisely because PS and PDF support so many rendering mechanisms they are unsuitable as editable document formats.
> I want to work 14 hours a day, 7 days a week in filthy and dangerous conditions for a pitance
You read too much Dickens. Unions were great back when there were no workers' rights laws in order to force the issue, but since these sort of things have found their way into law, unions have become more and more redundant. Thank the unions for driving unskilled wages up to the point of moving manufacturing to the third world. But hey, we can always blame loss of jobs on immigrants. Germany is utterly paralyzed in its ability to introduce economic reforms, due to a considerable extent to the powerful unions. But we can always sit back and dream of the good old days of the economic miracle.
> you end up with a similar amount of "mandatory spending".
You're roughly right, in the US you get an itemized bill, while in Europe you get one lump sum payment. Except that unfortunately in Europe you're likely to make less money in the same profession. Hopefully that will slowly change with the wage transparency of the Euro.
> a) Military is only a tiny percentage of the budget
I don't consider 16% tiny by any stretch. Neither as a percentage nor as an absolute figure ($280 billion, 2001 figures), in both cases absolutely dwarfing European countries. Incidentally, the linked document contains lots of interesting figures, including comparisons with other countries.
> Actually mostly to do with > a) absorbing East Germany and the raw costs associated with that and > b) struggling to meet the monetarist demands of the ECB
Those are mitigating factors for their sorry state, but they don't negate the fact that Germany is in serious need of economic reform. There are enormous barriers to entry into business (almost impossible to get small business loans), there are ridiculous outdated retail laws (such as on discounts and closing hours), and the power of unions is ludicrous. I view the US and Europe on the two extremes of a sensible middle ground. The US have serious social equity problems, while Europe has serious economic problems (which is slowly leading to social equity problems). Maybe a consolidation of the best features of both sides could result in a better overall social and economic model.
> There is no "faceless" technician at the other end
There is no fire-able individual that gets a performance review from the company. If they're unhappy with the outsourced datacenter performance, they have only two recourses: cancel the contract or sue, and I assume that contract agreements would most likely preclude the latter. The human element is completely being overlooked in these equations. Managers like pulling their staff together into a conference room and whipping their butts in times of crisis, making them feel in control of the company. Outsourcing precludes that. Sure, it will be (and has been) tried anyway, and will (and pretty much has in the case of ASPs) fail. But be my guest.
> What consumes most of the bandwidth in an internal company network is actually "raw" data. [...] > This means that the only bandwidth being used by your company will be to display web pages.
Hmm? Database queries are actually quite network efficient and in many respects very similar to HTTP. You send a query and get back a recordset. If you used a thin client instead, most of the information inside the recordset would likely travel across the network anyway, only in the form of more bloated ASCII-inside-HTML (to be displayed say inside an HTML table). And if the web server and database server don't reside on the same machine, you'd actually DOUBLE the network traffic.
Many cases can be made for browser-based thin client computing, but reduced network traffic definitely isn't one of them. There's nothing network efficient about stateless gobs of ASCII and graphics.
Another thing is that, as you mentioned, the ASP model is mainly suitable for web applications. Unfortunatly, that is still not the majority of applications in many corporations. There are still no satisfactory web versions of office applications, and there probably never will, because they're intrinsically client-side; if you insist on serving them via a browser, they will still end up mostly executing code (ActiveX, Java, JavaScript etc.) on the client side, but inside a sandbox, adding much headache and little benefit (think saving and printing).
Sure, I never said that one couldn't come up with fancy uses of vast computing power. But do you really need to simulate each individual molecule of the game world just to shoot some monsters? In the end it comes down to cost effectiveness--games become as complex as can be reasonably cheaply done, and not (much) more so.
> renting your spare computing power from your home PC
That seems more likely, and I'm expecting any day now some enterprising company to introduce something like that. Of course, it probably won't work out for data security reasons, but they will try anyway.
> I can't see supercomputing cycles as being something that is commodity, > or for that matter, something I (or any company) needs to buy on a metered basis.
Exactly. They're making a very odd proposal at at time at which more than ever computing power is becoming a cheap commodity, and when each new generation of processors enables new types of applications never before possible. Except that this set of applications is becoming smaller and smaller, leaving mostly finite element simulation and such. While there certainly is a market for that, it's hardly so huge that it should become a new utility.
Besides, given the potential emergence of quantum computing over the next one or two decades, which will make current super computing obsolete in most respects, and given that the legislation of new utilities can take a decade anyway, why bother?
> They can simply pay for CPU cycles just as they pay for electricity
Oh, so they don't need CPU cycles to connect to those rented cycles? They submit data and receive results by US Postal Service? Given that the equipment on customer premises is usually powerful enough for most typical processing required by your generic ACME anyway, what extra cycles exactly do they need? Today's company needs lots of STORAGE and NETWORK BANDWIDTH to store and transmit endless Outlook memos with large attachments, their processing demands are usually more than adequately met. And your typical ACME will be far too distrustful to let some faceless third party store those Outlook messages for them. Never mind that a big fat external WAN pipe to that third party is a lot more expensive than a big fat internal LAN pipe.
> If you guys actually read the (many) articles on Sam's speach, > you'd see its nothing like timesharing either.
Then why do they keep bringing up super computing? Do they want to be an ASP or a cluster provider, or I guess both? The ASP concept will simply never take off; even though it can look good on paper, it's doomed due to the distrustful human nature. Most companies are simply too paranoid to trust their data to third parties. I mean, if IS is too paranoid to let their workforce change their own desktop wallpaper or resolution, do you honestly think they're going to entrust their accounting and mission critical data to some guy at the other end of a WAN? Keep dreaming.
As far as clustering is concerned, for the handful of companies that have super computing needs, that might come in handy (even though Watson's infamous worldwide computer demand assessment might not be far off the mark in that case). But the vast majority of companies have a much higher demand for high throughput servers and networks and large storage capacities than for super computing. The bottleneck in most departmental servers nowadays are not so much the CPUs but storage and network.
Basically what it amounts to is that consumer CPUs are getting so fast and storage so ample that the domain of problems they can't handle on the desktop is getting smaller and smaller, leaving mainly finite element simulation screaming for more power. I mean, if a pocket music player is approaching the computing power of a 1980 Cray, the case for commoditized clustering (to the extent that they're advocating) is pretty weak. Methinks it's just IBM having a hard time breaking with their own past.
> but it will be far cheaper for you to simply rent the time.
How??? What will you be likely be doing that requires such enormous computing cycles at home? Your most likely cycle sink will be games, and even heavy 3D games are approaching the day where they don't saturate a consumer CPU anymore (esp. given a decent video card). So basically you're going to have a "terminal" at home that connects to a remote computer cluster to rent the processing time for--wordprocessing, which the terminal itself has more than ample processing power for. For all but a very select few super computer-type applications this entire model is flawed and merely a symptom of the fact that IBM simply can't break with its past.
> So why not make it GSM based, use anywhere and pay a single fee to one vendor?
Because I would be using this device 90% of the time at home, and why should I pay (VERY HEFTY) usage fees for SLOW GPRS when I have an idle broadband access point sitting there that wouldn't cost me any extra? The answer is because me using my broadband with the Pogo device doesn't generate any revenue for either Pogo or their wireless service partners. It's just like TiVo--we have the technology and infrastructure to not charge you monthly fees for the guide service, but we won't do that because it would earn us no money.
Or what phone companies want? Because this particular user would much rather have a Pogo with 802.11 connectivity that can roam around the house and backyard. The form factor is perfect, the functionality is perfect, the connectivity is useless. Of course, I do realize that they're doing the old razor-and-blades thing here, expecting to not make much money on the device and instead tying it to a lucrative perpetual service. But that's their problem, not mine, and I certainly won't be buying this one. As soon as it has WiFi, I will.
> Wow, TWO whole US universities? That's definitely > enough of a sample to make blanket statements
Relax, sonny, you're not the only one who took statistics and is aware of the concept of significant sample size. Notice that the US sample size is twice the Australian one <g>, so if you want to bitch, bitch on their behalf. Maybe they gave me an unfairly positive impression. Besides, I talk to plenty of people that have been through CS programs all over the place to have a feeling that the top schools' programs are the exception rather than the rule. And yes, Purdue would definitely be more in the exception group.
After having taken computer science courses at one Australian (UQ) and two US universities, I have to say that on average Aussie courses are way ahead. But then again Aussie unis push harder in most disciplines anyway. Heck, even highschool does--the only serious math I did at US unis beyond Australian senior high school level were differential equations 1 & 2.
> this device uses hardware dedicated to the sole purpose of decoing video/audio
Even MP3 decoders are often not true custom ASICs nowadays. I checked out a couple recently, and while they certainly look and behave like a single chip custom IC, internally they are based on a RISC core that executes regular old machine code. In fact, they offer the ability to upload new code to the chip. Given the increasing complexity of functionality expected of single-chip solutions nowadays, you're going to see this trend more and more. Rather than design and manufacture custom silicon--custom down to the individual gate level--it's cheaper and quicker to take an existing micro design and integrate it into a larger IC that implements the desired functionality with relatively little new(ly designed) silicon and lots of software.
> legendary Indian astrophysicist > brilliant Indian efforts
Sounds just like ole time communist propaganda. Our brilliant scientists figured out how to convert 4-ply toilet paper into 2-ply paper that will last twice as long. Soon India will take over the world.
Slashdot Mirror
Just because you're paranoid doesn't mean you're wrong. While there's no proof that this is what happened, it would be a very clever tactic indeed. And while they often get accused of being greedy, rarely do they get accused of not being clever.
> I happened to like the Replay user interface better, so I got it.
I have a TiVo SA, but I'm thinking of selling it and getting two of the new Replays for upstairs and downstairs. I really like the ability to stream shows from one unit to another within the house, something TiVo is rumored to be working on, but is dragging its feet on forever. Because of this ability, there seems to also be software out there that allows you to extract shows from a unit over the LAN without having to open it up and hack it. Incidentally, if anyone knows, how many ReplayTV units can cooperate on sharing shows with each other? Just two, or can you browse any number of units?
I am quite convinced that "they" have every intention and hope for this service to fail. If (when) it fails, it gives them just another reason to say "see, we told you LEGAL online movie distribution doesn't work--people want to STEAL, not BUY!"
If they truly wanted this to succeed, they would have made it genuinely attractive and competitive with DVD. They would have admitted to themselves that the obvious downsides to this service are the excruciatingly long download times and the need to watch the movie on the PC (or to have a HTPC), so they would have tried to offset those disadvantages by making rental temptingly cheap and offering longer rental times. They would maybe rent the movies for $1.99 and let you watch them for a week or more. And why on earth not? They're not limited by a restricted number of media that they need to circulate amongst a large clientele, unlike your video store. And they have only (relatively) minor adminstration costs, and except for the download bandwidth, no recurring maintenance costs per film (rewinding/applying security tag/reshelving etc.)
But if this succeeded, it would gradually erode the very lucrative video store business, and it would open the general consumer's eyes even more to the possibilities of digital media. Worse, it would force Hollywood to face the fact that their old business model is dying.
No, I honestly think they're either using this to prove that people aren't interested in acquiring online movies legally, or possibly as a half-hearted safe bet just in case this "online thing" should turn out to be more than just a fad.
> It sounds really harsh, but it's not intended to be that way.
Well, maybe that's got something to do with the vaunted German efficiency: efficiency through not beating around the bush and calling a spade a spade. Who knows. Then again, it wasn't in Germany where my wife hung up on a rude telemarketer, only to have him call back to tell her that he thought she was a real bitch. Click. No, it was back here in the good You Ass Of A...
> Of course, you could write your own DHCP server too
That would be the most flexible solution long-term, especially if they want to do all sorts of custom lookups during a DHCP request. It's quite a trivial protocol, we implemented a custom DHCP server a few years back in a couple of days. Just look up the RFCs for BOOTP and DHCP at www.ietf.org, making sure to use the latest ones since there were multiple revisions.
I have a stash of about 1000 CDs bought in the 80's and early 90s, and I haven't bought a new CD in years. I have plenty of music to listen to for years to come, and all the CDs happily rip to MP3. Besides, good music died along with the 70s. All the new stuff coming out is just new tricks, and I'm such an old dog.
> it's been seen as a companion to postscript
Exactly, in fact PDF and Postscript are very similar. When you render text to PS, you can end up with low-level drawing primitives such as lines and curves defining each letter, rather than high-level instructions such as "draw this string at position x,y". Once you've done that, recovering the original text amounts to highly sophisticated shape recognition and is impossible for all practical purposes. Precisely because PS and PDF support so many rendering mechanisms they are unsuitable as editable document formats.
> I want to work 14 hours a day, 7 days a week in filthy and dangerous conditions for a pitance
You read too much Dickens. Unions were great back when there were no workers' rights laws in order to force the issue, but since these sort of things have found their way into law, unions have become more and more redundant. Thank the unions for driving unskilled wages up to the point of moving manufacturing to the third world. But hey, we can always blame loss of jobs on immigrants. Germany is utterly paralyzed in its ability to introduce economic reforms, due to a considerable extent to the powerful unions. But we can always sit back and dream of the good old days of the economic miracle.
> you end up with a similar amount of "mandatory spending".
You're roughly right, in the US you get an itemized bill, while in Europe you get one lump sum payment. Except that unfortunately in Europe you're likely to make less money in the same profession. Hopefully that will slowly change with the wage transparency of the Euro.
> a) Military is only a tiny percentage of the budget
I don't consider 16% tiny by any stretch. Neither as a percentage nor as an absolute figure ($280 billion, 2001 figures), in both cases absolutely dwarfing European countries. Incidentally, the linked document contains lots of interesting figures, including comparisons with other countries.
I do agree with most of your other points.
> Actually mostly to do with
> a) absorbing East Germany and the raw costs associated with that and
> b) struggling to meet the monetarist demands of the ECB
Those are mitigating factors for their sorry state, but they don't negate the fact that Germany is in serious need of economic reform. There are enormous barriers to entry into business (almost impossible to get small business loans), there are ridiculous outdated retail laws (such as on discounts and closing hours), and the power of unions is ludicrous. I view the US and Europe on the two extremes of a sensible middle ground. The US have serious social equity problems, while Europe has serious economic problems (which is slowly leading to social equity problems). Maybe a consolidation of the best features of both sides could result in a better overall social and economic model.
> There is no "faceless" technician at the other end
There is no fire-able individual that gets a performance review from the company. If they're unhappy with the outsourced datacenter performance, they have only two recourses: cancel the contract or sue, and I assume that contract agreements would most likely preclude the latter. The human element is completely being overlooked in these equations. Managers like pulling their staff together into a conference room and whipping their butts in times of crisis, making them feel in control of the company. Outsourcing precludes that. Sure, it will be (and has been) tried anyway, and will (and pretty much has in the case of ASPs) fail. But be my guest.
> What consumes most of the bandwidth in an internal company network is actually "raw" data.
[...]
> This means that the only bandwidth being used by your company will be to display web pages.
Hmm? Database queries are actually quite network efficient and in many respects very similar to HTTP. You send a query and get back a recordset. If you used a thin client instead, most of the information inside the recordset would likely travel across the network anyway, only in the form of more bloated ASCII-inside-HTML (to be displayed say inside an HTML table). And if the web server and database server don't reside on the same machine, you'd actually DOUBLE the network traffic.
Many cases can be made for browser-based thin client computing, but reduced network traffic definitely isn't one of them. There's nothing network efficient about stateless gobs of ASCII and graphics.
Another thing is that, as you mentioned, the ASP model is mainly suitable for web applications. Unfortunatly, that is still not the majority of applications in many corporations. There are still no satisfactory web versions of office applications, and there probably never will, because they're intrinsically client-side; if you insist on serving them via a browser, they will still end up mostly executing code (ActiveX, Java, JavaScript etc.) on the client side, but inside a sandbox, adding much headache and little benefit (think saving and printing).
> Can you say "physics engine"?
Sure, I never said that one couldn't come up with fancy uses of vast computing power. But do you really need to simulate each individual molecule of the game world just to shoot some monsters? In the end it comes down to cost effectiveness--games become as complex as can be reasonably cheaply done, and not (much) more so.
> renting your spare computing power from your home PC
That seems more likely, and I'm expecting any day now some enterprising company to introduce something like that. Of course, it probably won't work out for data security reasons, but they will try anyway.
> I can't see supercomputing cycles as being something that is commodity,
> or for that matter, something I (or any company) needs to buy on a metered basis.
Exactly. They're making a very odd proposal at at time at which more than ever computing power is becoming a cheap commodity, and when each new generation of processors enables new types of applications never before possible. Except that this set of applications is becoming smaller and smaller, leaving mostly finite element simulation and such. While there certainly is a market for that, it's hardly so huge that it should become a new utility.
Besides, given the potential emergence of quantum computing over the next one or two decades, which will make current super computing obsolete in most respects, and given that the legislation of new utilities can take a decade anyway, why bother?
> They can simply pay for CPU cycles just as they pay for electricity
Oh, so they don't need CPU cycles to connect to those rented cycles? They submit data and receive results by US Postal Service? Given that the equipment on customer premises is usually powerful enough for most typical processing required by your generic ACME anyway, what extra cycles exactly do they need? Today's company needs lots of STORAGE and NETWORK BANDWIDTH to store and transmit endless Outlook memos with large attachments, their processing demands are usually more than adequately met. And your typical ACME will be far too distrustful to let some faceless third party store those Outlook messages for them. Never mind that a big fat external WAN pipe to that third party is a lot more expensive than a big fat internal LAN pipe.
> If you guys actually read the (many) articles on Sam's speach,
> you'd see its nothing like timesharing either.
Then why do they keep bringing up super computing? Do they want to be an ASP or a cluster provider, or I guess both? The ASP concept will simply never take off; even though it can look good on paper, it's doomed due to the distrustful human nature. Most companies are simply too paranoid to trust their data to third parties. I mean, if IS is too paranoid to let their workforce change their own desktop wallpaper or resolution, do you honestly think they're going to entrust their accounting and mission critical data to some guy at the other end of a WAN? Keep dreaming.
As far as clustering is concerned, for the handful of companies that have super computing needs, that might come in handy (even though Watson's infamous worldwide computer demand assessment might not be far off the mark in that case). But the vast majority of companies have a much higher demand for high throughput servers and networks and large storage capacities than for super computing. The bottleneck in most departmental servers nowadays are not so much the CPUs but storage and network.
Basically what it amounts to is that consumer CPUs are getting so fast and storage so ample that the domain of problems they can't handle on the desktop is getting smaller and smaller, leaving mainly finite element simulation screaming for more power. I mean, if a pocket music player is approaching the computing power of a 1980 Cray, the case for commoditized clustering (to the extent that they're advocating) is pretty weak. Methinks it's just IBM having a hard time breaking with their own past.
> but it will be far cheaper for you to simply rent the time.
How??? What will you be likely be doing that requires such enormous computing cycles at home? Your most likely cycle sink will be games, and even heavy 3D games are approaching the day where they don't saturate a consumer CPU anymore (esp. given a decent video card). So basically you're going to have a "terminal" at home that connects to a remote computer cluster to rent the processing time for--wordprocessing, which the terminal itself has more than ample processing power for. For all but a very select few super computer-type applications this entire model is flawed and merely a symptom of the fact that IBM simply can't break with its past.
> So why not make it GSM based, use anywhere and pay a single fee to one vendor?
Because I would be using this device 90% of the time at home, and why should I pay (VERY HEFTY) usage fees for SLOW GPRS when I have an idle broadband access point sitting there that wouldn't cost me any extra? The answer is because me using my broadband with the Pogo device doesn't generate any revenue for either Pogo or their wireless service partners. It's just like TiVo--we have the technology and infrastructure to not charge you monthly fees for the guide service, but we won't do that because it would earn us no money.
Or what phone companies want? Because this particular user would much rather have a Pogo with 802.11 connectivity that can roam around the house and backyard. The form factor is perfect, the functionality is perfect, the connectivity is useless. Of course, I do realize that they're doing the old razor-and-blades thing here, expecting to not make much money on the device and instead tying it to a lucrative perpetual service. But that's their problem, not mine, and I certainly won't be buying this one. As soon as it has WiFi, I will.
> Wow, TWO whole US universities? That's definitely
> enough of a sample to make blanket statements
Relax, sonny, you're not the only one who took statistics and is aware of the concept of significant sample size. Notice that the US sample size is twice the Australian one <g>, so if you want to bitch, bitch on their behalf. Maybe they gave me an unfairly positive impression. Besides, I talk to plenty of people that have been through CS programs all over the place to have a feeling that the top schools' programs are the exception rather than the rule. And yes, Purdue would definitely be more in the exception group.
After having taken computer science courses at one Australian (UQ) and two US universities, I have to say that on average Aussie courses are way ahead. But then again Aussie unis push harder in most disciplines anyway. Heck, even highschool does--the only serious math I did at US unis beyond Australian senior high school level were differential equations 1 & 2.
...this player won't be breaking any price records (except upwardly maybe). Call me when Apex sells one.
> this device uses hardware dedicated to the sole purpose of decoing video/audio
Even MP3 decoders are often not true custom ASICs nowadays. I checked out a couple recently, and while they certainly look and behave like a single chip custom IC, internally they are based on a RISC core that executes regular old machine code. In fact, they offer the ability to upload new code to the chip. Given the increasing complexity of functionality expected of single-chip solutions nowadays, you're going to see this trend more and more. Rather than design and manufacture custom silicon--custom down to the individual gate level--it's cheaper and quicker to take an existing micro design and integrate it into a larger IC that implements the desired functionality with relatively little new(ly designed) silicon and lots of software.
> legendary Indian astrophysicist
> brilliant Indian efforts
Sounds just like ole time communist propaganda. Our brilliant scientists figured out how to convert 4-ply toilet paper into 2-ply paper that will last twice as long. Soon India will take over the world.