The gap between the haves and havenots will be MUCH MUCH smaller in a future where the means of material (re)production are as cheap and democratized as current information (re)production is. One of the reasons Gates is so absurdly filthy rich (as compared to just Rockefeller rich or Ford rich) is because he's got a monopoly selling something which isn't really scarce.
Hmmm...Isn't that completely contradictory?
If Bill Gates, et al., can use IP law and contract law to get a continuing monopoly on a non-scarce resource, why do you think that we will not see the same thing when technology reduces the scarcity of material items? Seems to me that technologies that reduce scarcity can be counteracted by legislation and business practices. The world won't benefit unless we have both the new technology and a new social/economic model to use it.
The technology will be the easy part. Changes to the current social/economic model will be very traumatic. Look at all the junk that's going on with the simple change from analog to digital media technology. Now expand that from a single industry to the majority of industries, and imagine the implications...
It's not the intent that bothers me so much as the method. Our military has difficulty discriminating actual threats from innocent behavior. When we place it into a situation where there are many apparent threats, it shouldn't be surprising to see fatal mistakes.
Look at some of the major incidents in Afghanistan: if our military sees/hears gunfire, it often assumes that it's under attack, and retailates - even if it's a wedding party or an ally under military maneuvers. It's the same thing that Bruce is talking about - a single failure in threat discrimination can lead to deadly consequences for civilians, and increased hostility against America. Even though there is no direct intent to injure civilians or allies, the training and structure of the system encourages poor reactions and often results in undesired and unnecessary casualties.
Yes, the terrorists are responsible for their intent to harm civilians. But we are also responsible for developing a system that tends to harm civilians in its failure modes. Time to develop a ductile military system, perchance?
I was assuming that the upgrades were free, like your original post. Yeah, even then the TCO would be horrible due to the work of testing, installing, and integrating the new versions so often.
But my concern was the security more than the TCO - a lot of security holes are not found immediately upon release, and if the warranty on the previous version is voided when a new version is available, then there would be much less incentive for the manufacturer to fix security holes in previous versions.
The reason I'm thinking about this model is that it fits very well with the "subscription with automatic upgrades" model that seems to be on its way.
Finding a new vendor is an option only if there is an alternative vendor whose product is similar enough that it doesn't add huge re-training and support costs to the TCO. Works for some products...doesn't for others...
Exactly right. The pe(rson|eople) who can fix the source should be responsible for fixing it for paying customers.
With proprietary software, that's the company's responsibility. If the source is available, but only the company is allowed to change it, then it's still the company's responsibility. If the source is freely modifiable, then it's anyone's/everyone's responsibility, no matter who was paid for the software.
In other words, if someone controls modification of the software through IP laws (copyright, patent, trademark, trade secret, whatever), then that entity has the responsibility to fix it. If they give up this control over the software, then they also give up the responsibility to warranty it, because anyone can then fix it legally.
If the warranty only applies to the latest version, then you'll see big software companies bringing out new versions every few months, so that the warranties will be voided by the new version before many of the significant bugs/security holes are found. Is this really what you want?
I don't think that'd be much of a problem. Look at it this way - under her proposal, the primary benefit is the ability to get anything that's available, and know that it's the one you want. Is that efficiency worth a quarter or nickel per track? Somehow, I think most people would think so.
Think of it as paying for the service of making easy-to-find, reliable tracks available, rather than just paying for the tracks.
FWIW, I rarely have need for UNION, but when I do it's a lot better than multiple INSERTs into a temporary table followed by a SELECT and a DROP.
It's always fun to dive into the relational algebra behind it to get a better idea what's going on. If you're interested, this page has a good summary of relational operations, although it doesn't give direct mappings between the examples and standard SQL. Interestingly enough, 6 of the 10 examples would be done with a JOIN in SQL, and only one would need a UNION...
Maybe I'm ignorant, but I can't see how to use a JOIN to do what I use UNION for. Maybe you can enlighten me. Here's an example:
SELECT DISTINCT 'foo' AS src, id,...
FROM foo
UNION
SELECT DISTINCT 'bar' AS src, id,...
FROM bar
JOIN combines data from multiple tables horizontally (into the same row/tuple) while UNION combines data from multiple tables horizontally (into separate rows/tuples). I need and use both, although JOIN far more frequently than UNION. In fact, I use UNION much more often when merging external data for import than I do within my main databases/applications.
(If it makes any difference, I work primarily in Ingres and PostgreSQL)
They're all trying to do the Right Thing, but sometimes they fail really miserably
Yes, they are all trying to do what they think is the Right Thing. That's not necessarily the same as what I think is the right thing. Sometimes their perception of the Right Thing is quite different from mine.
I find that the higher up a person is in any organization, and the larger the organization, be it government or corporation, the more differences there are between their perception of the Right Thing and mine. The higher a person is, the less I am likely to implicitly trust them until I have seen convincing evidence of trustworthiness.
Yes, rockets can be aerodynamically stable. But there's a difference between stable and controlled.
I've had model rockets that were aerodynamically stable, but sure didn't end up where I wanted when a gust of wind came up during the flight. The nose turns into the wind, and the rocket usually lands in a distant tree rather than in the area I had "planned". Granted, it might take a much larger gust of wind to cause the same problem with a full-sized rocket...
Because with cable TV you are paying for access to the content, while with broadband you are paying for access to the infrastructure. They are completely different - sharing broadband does not involve redistributing content, merely access.
I must admit, this is a masterful stroke. It appears to give users additional control over their computer's security, while limiting the options in such a way that it actually concentrates that control into others' hands.
[NOTE: Since real information about Palladium is pretty fuzzy right now, I'm theorizing a bit about its capabilities for now. Only time will tell...]
It can remove my power to choose what's authorized to run on my computer. It can prevent usage of "untrusted" or "unauthorized" code. Lovely turn of phrase, that. Notice how it uses the passive to avoid any implication of *who* is trusting or authorizing the code? "Palladium is all about deciding what's trustworthy. It not only lets your computer know that you're you , but also can limit what arrives (and runs on) your computer, verifying where it comes from and who created it." The implication is that the user is in control, but who decides?
I have not yet seen anything saying how programs are authorized. It would be logical to set up a coalition to do this, and use membership agreements to control the behavior and competitiveness of its members, and exclude undesirables. We can see prior art in the way the DVD-CCA controls access to the CSS keys and uses that control to enforce region controls and lack of digital output.
It can remove my power to access information, since Palladium "can limit what arrives" on my computer. In other words, the authorization control can extend beyond code to data. If a site does not have a valid Palladium authorization (however those are issued), then Palladium may be able to prevent access to it (and tell me that it has saved me from an "unauthorized site"). Again, the key to this control rests in the authorization process.
It can remove my power to customize my computer. No, I'm not talking about case mods, I'm talking about OS and program configuration. In order to maintain a "trustworthy" system, it will have to limit access to the configuration system. Assuming they keep something like the Windows Registry, I can see two options here. They may refuse to authorize regedit, et al., and remove OS authorization from any registry touched by those programs. Or they may remove the my ability to change anything "critical" (by some definition or other) in the registry.
Ultimately, it can force a choice between "all-Palladium" and "no-Palladium". If it can refuse to run unauthorized programs or access unauthorized sites while any authorized programs are running or authorized sites are being accessed, then I cannot work in both realms at the same time. I must either choose "Palladium" ("safe") or "non-Palladium" ("dangerous"). It could also deal with these realms asymmetrically: if I try to use Palladium resources, it could automatically close all non-Palladium resources (and tell me that it has saved me from danger), but if I try to use non-Palladium resources, it might refuse to load them until I had manually closed all of my Palladium resources, and perhaps rebooted.
Faced with this choice, how many users will be willing to give up some useful non-Palladium resources rather than giving up all Palladium resources? Immanentizing the false dichotomy, anyone?
I sure hope I'm wrong about this, and that I'm just being too paranoid. Unfortunately, recent history seems to show that we need a really healthy dose of paranoia when dealing with things like this. Again, only time will tell for sure.
Yeah, it stinks that cable ISPs are [legally] gaining [monopoly] control and the ability to squelch their competition.
OTOH, this may also slow down the migration of commercial software to the network-based service model. Anything that keeps broadband prices high and single-sourced will be a disincentive for this migration.
I doubt this possible benefit outweighs the liabilities, though...and Microsoft already has an alternative - if they can't force people to use net-based application servers for their software, they can own their desktop machines with Palladium.
Slashdot Mirror
"PC Load Letter": You have not included enough politically correct phrases in your letter. Please edit it and resubmit for printing.
Hey, if God can resurrect the Apostle Peter (et al.), whose body is loooong gone, he wouldn't have any problem with ashes or diamonds.
There's nothing that can't be reversed with time travel or alternate universes...the ultimate plot extrication tools! :(
Hmmm...Isn't that completely contradictory?
If Bill Gates, et al., can use IP law and contract law to get a continuing monopoly on a non-scarce resource, why do you think that we will not see the same thing when technology reduces the scarcity of material items? Seems to me that technologies that reduce scarcity can be counteracted by legislation and business practices. The world won't benefit unless we have both the new technology and a new social/economic model to use it.
The technology will be the easy part. Changes to the current social/economic model will be very traumatic. Look at all the junk that's going on with the simple change from analog to digital media technology. Now expand that from a single industry to the majority of industries, and imagine the implications...
It's not the intent that bothers me so much as the method. Our military has difficulty discriminating actual threats from innocent behavior. When we place it into a situation where there are many apparent threats, it shouldn't be surprising to see fatal mistakes.
Look at some of the major incidents in Afghanistan: if our military sees/hears gunfire, it often assumes that it's under attack, and retailates - even if it's a wedding party or an ally under military maneuvers. It's the same thing that Bruce is talking about - a single failure in threat discrimination can lead to deadly consequences for civilians, and increased hostility against America. Even though there is no direct intent to injure civilians or allies, the training and structure of the system encourages poor reactions and often results in undesired and unnecessary casualties.
Yes, the terrorists are responsible for their intent to harm civilians. But we are also responsible for developing a system that tends to harm civilians in its failure modes. Time to develop a ductile military system, perchance?
I was assuming that the upgrades were free, like your original post. Yeah, even then the TCO would be horrible due to the work of testing, installing, and integrating the new versions so often.
But my concern was the security more than the TCO - a lot of security holes are not found immediately upon release, and if the warranty on the previous version is voided when a new version is available, then there would be much less incentive for the manufacturer to fix security holes in previous versions.
The reason I'm thinking about this model is that it fits very well with the "subscription with automatic upgrades" model that seems to be on its way.
Finding a new vendor is an option only if there is an alternative vendor whose product is similar enough that it doesn't add huge re-training and support costs to the TCO. Works for some products...doesn't for others...
Exactly right. The pe(rson|eople) who can fix the source should be responsible for fixing it for paying customers.
With proprietary software, that's the company's responsibility. If the source is available, but only the company is allowed to change it, then it's still the company's responsibility. If the source is freely modifiable, then it's anyone's/everyone's responsibility, no matter who was paid for the software.
In other words, if someone controls modification of the software through IP laws (copyright, patent, trademark, trade secret, whatever), then that entity has the responsibility to fix it. If they give up this control over the software, then they also give up the responsibility to warranty it, because anyone can then fix it legally.
If the warranty only applies to the latest version, then you'll see big software companies bringing out new versions every few months, so that the warranties will be voided by the new version before many of the significant bugs/security holes are found. Is this really what you want?
I don't think that'd be much of a problem. Look at it this way - under her proposal, the primary benefit is the ability to get anything that's available, and know that it's the one you want. Is that efficiency worth a quarter or nickel per track? Somehow, I think most people would think so.
Think of it as paying for the service of making easy-to-find, reliable tracks available, rather than just paying for the tracks.
Any time...
FWIW, I rarely have need for UNION, but when I do it's a lot better than multiple INSERTs into a temporary table followed by a SELECT and a DROP.
It's always fun to dive into the relational algebra behind it to get a better idea what's going on. If you're interested, this page has a good summary of relational operations, although it doesn't give direct mappings between the examples and standard SQL. Interestingly enough, 6 of the 10 examples would be done with a JOIN in SQL, and only one would need a UNION...
And just which FM would that be, now?
Maybe I'm ignorant, but I can't see how to use a JOIN to do what I use UNION for. Maybe you can enlighten me. Here's an example:
SELECT DISTINCT 'foo' AS src, id,JOIN combines data from multiple tables horizontally (into the same row/tuple) while UNION combines data from multiple tables horizontally (into separate rows/tuples). I need and use both, although JOIN far more frequently than UNION. In fact, I use UNION much more often when merging external data for import than I do within my main databases/applications.
(If it makes any difference, I work primarily in Ingres and PostgreSQL)
Yeah, and Linus once said that Linux was so dependent on i386 that it couldn't be ported to anything else.
Things can change.
Yes, they are all trying to do what they think is the Right Thing. That's not necessarily the same as what I think is the right thing. Sometimes their perception of the Right Thing is quite different from mine.
I find that the higher up a person is in any organization, and the larger the organization, be it government or corporation, the more differences there are between their perception of the Right Thing and mine. The higher a person is, the less I am likely to implicitly trust them until I have seen convincing evidence of trustworthiness.
Joke. Manacle. Bonds. Slavery. Duh.
Yes, rockets can be aerodynamically stable. But there's a difference between stable and controlled.
I've had model rockets that were aerodynamically stable, but sure didn't end up where I wanted when a gust of wind came up during the flight. The nose turns into the wind, and the rocket usually lands in a distant tree rather than in the area I had "planned". Granted, it might take a much larger gust of wind to cause the same problem with a full-sized rocket...
Because with cable TV you are paying for access to the content, while with broadband you are paying for access to the infrastructure. They are completely different - sharing broadband does not involve redistributing content, merely access.
I must admit, this is a masterful stroke. It appears to give users additional control over their computer's security, while limiting the options in such a way that it actually concentrates that control into others' hands.
[NOTE: Since real information about Palladium is pretty fuzzy right now, I'm theorizing a bit about its capabilities for now. Only time will tell...]
It can remove my power to choose what's authorized to run on my computer. It can prevent usage of "untrusted" or "unauthorized" code. Lovely turn of phrase, that. Notice how it uses the passive to avoid any implication of *who* is trusting or authorizing the code? "Palladium is all about deciding what's trustworthy. It not only lets your computer know that you're you , but also can limit what arrives (and runs on) your computer, verifying where it comes from and who created it." The implication is that the user is in control, but who decides?
I have not yet seen anything saying how programs are authorized. It would be logical to set up a coalition to do this, and use membership agreements to control the behavior and competitiveness of its members, and exclude undesirables. We can see prior art in the way the DVD-CCA controls access to the CSS keys and uses that control to enforce region controls and lack of digital output.
It can remove my power to access information, since Palladium "can limit what arrives" on my computer. In other words, the authorization control can extend beyond code to data. If a site does not have a valid Palladium authorization (however those are issued), then Palladium may be able to prevent access to it (and tell me that it has saved me from an "unauthorized site"). Again, the key to this control rests in the authorization process.
It can remove my power to customize my computer. No, I'm not talking about case mods, I'm talking about OS and program configuration. In order to maintain a "trustworthy" system, it will have to limit access to the configuration system. Assuming they keep something like the Windows Registry, I can see two options here. They may refuse to authorize regedit, et al., and remove OS authorization from any registry touched by those programs. Or they may remove the my ability to change anything "critical" (by some definition or other) in the registry.
Ultimately, it can force a choice between "all-Palladium" and "no-Palladium". If it can refuse to run unauthorized programs or access unauthorized sites while any authorized programs are running or authorized sites are being accessed, then I cannot work in both realms at the same time. I must either choose "Palladium" ("safe") or "non-Palladium" ("dangerous"). It could also deal with these realms asymmetrically: if I try to use Palladium resources, it could automatically close all non-Palladium resources (and tell me that it has saved me from danger), but if I try to use non-Palladium resources, it might refuse to load them until I had manually closed all of my Palladium resources, and perhaps rebooted.
Faced with this choice, how many users will be willing to give up some useful non-Palladium resources rather than giving up all Palladium resources? Immanentizing the false dichotomy, anyone?
I sure hope I'm wrong about this, and that I'm just being too paranoid. Unfortunately, recent history seems to show that we need a really healthy dose of paranoia when dealing with things like this. Again, only time will tell for sure.
Yeah, it stinks that cable ISPs are [legally] gaining [monopoly] control and the ability to squelch their competition.
OTOH, this may also slow down the migration of commercial software to the network-based service model. Anything that keeps broadband prices high and single-sourced will be a disincentive for this migration.
I doubt this possible benefit outweighs the liabilities, though...and Microsoft already has an alternative - if they can't force people to use net-based application servers for their software, they can own their desktop machines with Palladium.
Appropriate, too. Seems like they are often more like emcees than like representatives...
It was Benjamin Franklin:
When I first heard it, I thought he should have used "float" instead of "swim"...
Then you need to check out the Nakamichi mobile equipment. No flash, just good sound...
Of course. Bogosort isn't quite so bogo if it's massively parallelized! ;)
Aha! A DNA bogo-sort!