For all the good it might possibly maybe do, I don't think breaking up Microsoft into 3 or 4 competing or even companies selling each product will help. If we make competing companies, one will still dominate. And if we make companies for each product, those products will still dominate. The solution?
Well, that's harder than you might think. One solution might be to dedicate time to monitor closely Microsoft's actions and make sure they behave themselves. The problem with that? How far do we go in monitoring Microsoft? Is bundling their software with their OS wrong? Not necessarily. For all its good, an OS is useless without software. However, things like extending Kerberos and making it proprietary at the exclusion of other OSes, well that probably goes past the line of dubious actions.
I think the best plan (without causing too much harm to the consumer, because after all, this is supposed to help the consumer) would be to sever Microsoft's deathgrip on companies such as Apple. Unlike many other industries, I believe that having global standards for the personal computers can only benefit consumers.
When Microsoft's actions are such that it makes it impossible for 3 party developers is when the problem comes into play. We should probably make Microsoft show its source (when needed), while retaining rights to the source, so that other companies may develop competing software. At this point, it'd be very inconvenient for the industry as a whole to have to adapt to any totally new OS.
Of course, anyone else's thoughts are welcome. It's hard to be perfect. But I think the solution in this case doesn't involve splitting Microsoft up.
If it could flash a dozen or two words on the screen to indicate the *themes* it's extrapolating from my current interests, I'd be fascinated. I think this would be one of the best applications of a web-related AI. They key, of course, would be constructing it so it doesn't just find words that are repeated many times on the page. (Think how many times the word 'a' is on a single page, much less months worth of browsing.) Instead, you'd have to look for relavant topics. That would be where this idea requires an AI, rather than a natural language parser. Some places to look might include what the user tends to look for in searches, the META tags of pages (although that's not as useful.) The duration of time spent at a page when the browser is active (i.e. on top). Some people leave their browsers open for long periods of time (like me) while doing other things. The ``Links'' section of pages viewed, and if the user goes there at all. Also, what section of the page is viewed for long periods of time while the browser is active. Similar tactics to figure out what the user likes. Ideally, over a period of time, the AI should be able to predict what would be appropriate or not. The last issue would be whether the user should get signifigant control over this type of AI. Of course, some basic control would be needed (so as to stop the "Hey, look, a new foot fetish site!" type suggestions from popping up while your parents are in the room.) But too much control might totally defeat the purpose. Anyway. Those are my thoughts. Enjoy. Flame. Etc.
Slashdot Mirror
Well, that's harder than you might think. One solution might be to dedicate time to monitor closely Microsoft's actions and make sure they behave themselves. The problem with that? How far do we go in monitoring Microsoft? Is bundling their software with their OS wrong? Not necessarily. For all its good, an OS is useless without software. However, things like extending Kerberos and making it proprietary at the exclusion of other OSes, well that probably goes past the line of dubious actions.
I think the best plan (without causing too much harm to the consumer, because after all, this is supposed to help the consumer) would be to sever Microsoft's deathgrip on companies such as Apple. Unlike many other industries, I believe that having global standards for the personal computers can only benefit consumers.
When Microsoft's actions are such that it makes it impossible for 3 party developers is when the problem comes into play. We should probably make Microsoft show its source (when needed), while retaining rights to the source, so that other companies may develop competing software. At this point, it'd be very inconvenient for the industry as a whole to have to adapt to any totally new OS.
Of course, anyone else's thoughts are welcome. It's hard to be perfect. But I think the solution in this case doesn't involve splitting Microsoft up.
If it could flash a dozen or two words on the screen to indicate the *themes* it's extrapolating from my current interests, I'd be fascinated. I think this would be one of the best applications of a web-related AI. They key, of course, would be constructing it so it doesn't just find words that are repeated many times on the page. (Think how many times the word 'a' is on a single page, much less months worth of browsing.) Instead, you'd have to look for relavant topics. That would be where this idea requires an AI, rather than a natural language parser. Some places to look might include what the user tends to look for in searches, the META tags of pages (although that's not as useful.) The duration of time spent at a page when the browser is active (i.e. on top). Some people leave their browsers open for long periods of time (like me) while doing other things. The ``Links'' section of pages viewed, and if the user goes there at all. Also, what section of the page is viewed for long periods of time while the browser is active. Similar tactics to figure out what the user likes. Ideally, over a period of time, the AI should be able to predict what would be appropriate or not. The last issue would be whether the user should get signifigant control over this type of AI. Of course, some basic control would be needed (so as to stop the "Hey, look, a new foot fetish site!" type suggestions from popping up while your parents are in the room.) But too much control might totally defeat the purpose. Anyway. Those are my thoughts. Enjoy. Flame. Etc.