First, a clarification to the above: Twitter also works just fine with delivery over IM, so any Jabber client like Google Talk (which has a nice client on e.g. Blackberries and works on iPhones et al) will do instead, so no need to use SMS if you don't want to. Plus one can go to one's Twitter page and see a running narrative of everyone being 'followed', so it can be used in a mode with no 'instant' delivery at all.
Which leads to the main point: It's wrong to get hung up on the 'immediacy' of Twitter, because there are other technologies that already do that well (IM, email...), and this is perhaps why I think a lot of people don't get it or dismiss it as pointless. The niche that Twitter fills is asynchronous communication between 2 or more people. It's publish-subscribe.
The problem with all the other forms of instant communication we have (phone calls, SMS message, IM, email...) is that they are all interrupts: people stop whatever they are doing when you initiate communcation this way. But often one just wants to update one's own status, or make an observation about something without causing an interrupt, but still allowing the set of potentially interested people to see your updates when they are available, ready and interested.
Example: Person A is going out to drinks after work with friends. B might join if time permits, but is on a coding deadline and will potentially work rather late. A twitters progress (leaving bar X, moving to Y where there is also food), which does not interrupt B. B reaches a good stopping point, checks A's twitters, is able to meet friends at Y later in evening.
That, and variations thereof, is what Twitter is good at.
As usual, it's more fun for/.ers to have a religious debate about the relative merits of Java vs. C# rather than answer the actual question being asked. So the idea is to choose between them for the purposes of learning object-oriented programming, you say?
Either one will do almost equally well for a basic OO programming course, though you should be aware that despite all of the zealotism and hyperbole in the discussion here, they are *both* relatively poor choices for learning OOP. Both are designed to improve on the many shortcomings of C++ and C for large-scale development. As such, they are quite similar, and are roughly equivalently mediocre for erudition purposes. Furthermore, because they are meant as enterprise-class development langages, they have many features and formalisms (namespaces, assemblies, and particularly strong typing) which are designed to manage complexity and prevent errors for the 500,000 line code build. And if you're going to pursue a career in software engineering you're eventually going to want and need those things. But these same features simply get in the way and obfuscate when you're building a 200-line program to demonstrate an OO concept for a class. As the author of the Boo language (which compiles to the CLI) says in the Boo Manifesto, '"public static void main", that was a good one!'
A good choice for such a course would be Python or Ruby. Elegant, compact, forgiving, relatively pure (especially Ruby) and weakly typed, they are a pleasure to use. Ruby is so beautiful at times that you could weep. An even purer choice would be something like Smalltalk, but I think Ruby makes a better tool for teaching because it is also pragmatic enought to incorporate things like functional programming elements and idioms, which can sometimes throw light on the OO concepts and approaches by juxtaposition, and allow you to develop an understanding of when OO helps you get things done and when it gets in the way.
This doesn't solve your original problem, of course, because presumably you may choose only C# or Java. But do yourself a favor and try to bang out a few homework problems in Python or Ruby just to kick the tires. You'll be floored at how much more compact and lucid it is, and how flexible. Plus you will find in the long run that a feel for how you would implement something in a pure OO language actually greatly informs and improves your designs in languages like C# and Java, which will ultimately make you a much more skilled programmer. Which is, after all, the point.
I was at a talk by Mel Karmazin, chief executive of CBS, a few months ago. One of the questions from the audience was something like "as a network, do you feel threatened by the potential of these new Tivo devices for viewers to avoid commercials".
Turns out he has one. He said that if they found that consumers were skipping over the commercials they would find a way to put the advertising in the shows. Now, he may have just been kidding, or saying it for a bit of the shock value, but he seemed serious. That would suck.
I have a TiVo as well, and I absolutely use it to skip over intros, commercials, etc. You can take in an episode of South Park in 20 minutes. Unlike Replay, however, there is no button for 30 second skip; instead there is a 2x, 12x and 60x fast forward. I've actually come to prefer this, as I'll sometimes see something interesting or unusual zip by and go back to watch the commercial (some of them are more entertaining than the programming). So in the best of scenarios, widespread use of these devices will force commercials to be more entertaining/interesting so people will actually watch them.
What you propose is probably possible, but if you are getting program information from a separate source you would have to throw away the TiVo user interface completely. Most of the TiVo interface is built around the TV program's metadata (ie: actors, genre, when the program was made), which ClickTV may not be able to provide. Even if they could, it's not clear that you could easily map it (and maintain that map) to whatever TiVo's internal format is. So I would bet that you would have to abandon the TiVo's software and UI almost entirely, and substitute your own.
That would be a shame, because (1) TiVo designed an outstanding UI, and (2) most of the benefit of using TiVo (I've had one for a few months now) is in the interface and the TiVo service beyond the schedule information. Once it notices that you watch a lot of, e.g., Science Fiction shows, it starts recording other ones you might like in any free space it has. It makes suggestions based on what you've watched of other stuff you might like. It reminds you of recording conflicts when you schedule them, and does reasonably good priortization if you give it lots to record. Plus, it is a work in progress: it has an extremely active user community (check out the AVS Forum TiVo discussion group) that TiVo actually pays attention to, and it shows in the design. So if you are looking to hack it just to have a digital VCR, I'd wait, or get Replay instead (assuming *it* works in Canada). If you are interested in hacking it to avoid paying for the service, you'll be avoiding (IMHO) exactly what makes the device worth owning in the first place
Instead of worrying about the next kind of micro-storage device we can build into a digital camera, the industry should be focusing on taking it out entirely. This is what technologies like Bluetooth are supposed to make possible. You carry around storage in whatever form you like--your PDA, a 1-Gig drive in your backpack, an IBM microdrive on your belt--and the camera stores directly to the external device. This would also allow the camera to be smaller, use less battery, etc. In fact, if you used a color PDA, you potentially wouldn't need the color preview screen on the camera either.
Bluetooth products are mostly still in prototype, but AFAIK nobody is even talking about this with cameras. Seems like the way to go.
Perhaps this is the point, but creating a user-driven forum for discussing election issues won't do the majority of voters any good until we change the way the political system works.
Pre-Reagan, elections were all about party allegiance. Post-Reagan, as parties have drifted toward the center, they have been more about image (encouraged by burgeoning media coverage) and hotbutton issues like abortion. The one thing elections have never been about for most people is an objective evaluation of the candidates on all the issues, side-by-side, point-by-point. While it probably should be that way (the Democracy Project obviously wishes it so), it cannot for several reasons:
(1) People are on the whole impressionable, gullible and lazy, so they can be relied on neither to seek out objective criteria, nor to understand or take the time to evaluate them if they did.
(2) Most issues are quite complex, and require a lot of context and background knowledge. Take free trade: even an Econ 101 level discussion of comparative advantage or the lump-of-labor fallacy will cause most people's eyes to glaze over. So they leave the hard details of most issues to the politicians, and vote for them because, for instance, that person "cares about us, and won't let our jobs go overseas". Try an experiment: stop 10 people on the street, and ask them what Gore or Bush's position on free trade is.
(3) Politicians win elections by trying to be all things to all people. This is impossible if they take concrete positions, so they make vague promises that everyone would agree to (pro-education, create more jobs, etc) and avoid specifics whenever possible, relying on (1). So the current system gives them no incentive to devise a user-centered site where actual issues are discussed and positions taken.
This is all depressing, of course, but it seems to me that before you even have a chance to bootstrap voters into a world of active participation and ideal rational choice, you have to first dismantle the two-party, media-driven system that causes them to vote the way they do now.
Globalstar is another LEO satellite phone system, but it differs from Iridium in that it: (a) offers significantly cheaper and lighter phones, and (b) is not bankrupt and deorbiting its satellites. You can get a tri-mode Analog/CDMA/Satellite phone for about $1200, and there's no reason not to expect that price to fall further. They plan to roll out internet access later this year, at [a whopping] 9600 baud. I've used the phone for voice, and it's comparable to digital cellular in clarity. A really amazing device.
Admittedly $1200 isn't cheap, but I would imagine if you're the kind of person that is on boats a lot (your own or a cruise line's) it's not a big investment. Their website is http://www.globalstar.com
For those of you now wishing you had bought a new home so you could do this, there's a recent nice article in Ars Technica about how to wire your existing home properly. An interesting read.
It is actually a very thorough point-to-point remedy solution that compliments the Findings of Fact.
Agreed. In its goals, the DOJ has everyone's best interests in mind, including (as I am about to argue) Microsoft's, because it will force them to spend their energy on their products, and not the perpetuation of their monopoly.
It's important to remember what the DOJ is trying to remedy here. The point is not to punish Microsoft by blowing them into bits; the point is to keep them from harming other companies and consumers by abusing their monopoly power. This power stems IMHO primarily from their dominance of an office suite, a set of OSes (95/98, NT/2000) and a browser technology which have monopolistic-level market penetrations. However, the real power (and thus the potential for abuse) does not come from the quality of those products (as some of us know all too well), or even from their high level of penetration. It comes from MS's relentless tying and blurring of the lines between them, so that if you need to use one of them, you more or less roped into using them all. In other words, their incentive structure in that position is to invent new ways for you to need all things Microsoft, not to produce better products. That practice is what the DOJ wants to eliminate, and rightly so.
The originator of this thread is correct in that the post-split companies would have monopoly-level penetration in their of their product areas. However, they would have none of the ability to tie the products together, and indeed would not have any real ability or economic incentive to do so. The incentive they would be left with would be to compete by improving their products.
There are a lot of smart and talented people at Microsoft. Until now much of their collective coding hands have been tied insofar as the primary business strategy has been the market dominance of the Windows/Office empire. Split up, and with remedies properly applied, they will be forced to focus on the quality of their product to maintain market share. Everybody benefits from this, including, perhaps especially, Microsoft. It will mean that in the long run you may be seeing more, not less, of things MS. One of the reasons that surveys are finding consumers to be deeply ambivalent about the DOJ's actions is that there are huge network effect benefits to having almost everyone running the same OS and office software, and people are worried that the wrong kind of remedy will somehow mess up those benefits. If the 3-way split forces MS1, 2 and 3 to improve their products, then it seems to me those same consumers should embrace those products more, not less. So if what you secretly wish for at night is for Bill and his buddies to go down in flames, consider the possibility that the split will unlock a lot of potential in the new companies. Remember: "If you strike me down, I shall become more powerful than you can possibly imagine."
The system still suffers from a lack of drivers, but those will get there eventually.
Sadly, all new OSes are plagued by a lack of drivers. BeOS still doesn't run right on my [rather generic] laptop, because of a lack of available drivers, and the rollout of new ones is sloooow.
OSes are notoriously hard to get right, take years to develop, and when they are finally ready to show the world, only a handful of drivers are available (the few they have been developed with). This leads me to think that a nice project would be a sort of open driver API suite that the major hardware manufacturers could commit to supporting. (Sort of like DirectX, but with the standard controlled by a consortium, and for all peripherals, not just for graphics cards.) That way the OS could ship with a set of "generic" API-supported drivers that would work at least OK with most devices.
This would make the hardware manufacturers happy, because they wouldn't have to deal with every new OS on the block separately, and wouldn't have to release their proprietary hardware specs to (from their perspective) fly-by-night alternate OS developers. It would also make the alt-OS users happy, because they would be much more likely to be able to get drivers quickly. Such an API would obviously not result in optimal drivers, would be slower than custom-written drivers and would not be able to take advantage of the fancier aspects of the hardware, but at least it would be easy for the OS to provide generic-enough drivers that you could try out the OS and kick its tires without doing it in standard VGA with no printer, USB, PCMCIA and network support. Trouble is, I don't know enough about the realities of driver development to know if this is even a realistic idea. Or perhaps something like this exists already...?
Imagine a site like mysimon that compared millions of products from thousands of stores.
A laudable goal, but in order for this to work on that kind of scale you need to establish a standard, and then you need the retailers to participate. This is a classic bootstrapping problem, because retailers hate price transparency--no seller is going to sign up for a system where it has to do extra work to compete against potentially millions of other retailers based on price. However, if you can make the system large enough that consumers start to go there first, all of a sudden the retailer will need to participate, since it will start to lose customers to the service otherwise.
Then, instead of making you do all the work of signing up and purchasing from each store, the program could use a single shopping cart, and negotiate the sale from each of the stores that had the best prices, using the same user information
The single shopping cart has the same problem: in a world of increasing price transparency, one of the few "sticky" features an e-tailer has left is that you've already gone through the tedious job of filling out a user profile on their site, and not someone else's. Are you really going to go take 15 minutes to fill out a profile at Egghead.com to shave a few bucks off the price of a digital camera if you already have a profile at PC Zone? Probably not. Plus, there are perhaps some privacy issues.
Of course, it would be a truly wonderful thing to have if you could find a way to make it happen. I just wonder if it's realistic...
START is a nice attempt at natural language parsing, but (from my understanding) once it has made a sophisticated identification of what you mean, it can only direct you to the canned answers it has for various questions. What it appears to be missing (someone correct me if it has this facility--I only had time to skim their website) is any real-world 'understanding' of the relationships between the objects it is manipulating. A nice AI-like project would be to try to overlay some real-world intelligence so that it could deduce answers from relationships in its datasets.
For example, START is able to answer:
Which country has the largest population?
and
Which country has the largest area?
but not:
Which country has the largest population density?
because it does not 'know' that population density is population/area. Of course, this is a hard thing to tackle, but a proof-of-concept could be done with a limited object domain (like geograhpical data).
The same knowledge of relationships could allow search engines to be a lot more intelligent about what you are actually looking for, instead of relying purely on keywords and statistics. In other words, I should be able to go to askjeeves.com and ask "Where is the island of Java?" and not have almost all of the answers have nothing to do with Indonesia.
Of course you realize that the downward spinoff spiral is an inevitable consequence of the economics of television, since TV producers like to think they are getting something for nothing with a spinoff. The problem is that after the show is cast with a fraction of the original's budget, you end up with cheap[er] sets, worse acting and worse writing than the original. TV executives fail to realize what drove people to watch the show in the first place.
When viewers start abandoning the spinoff in droves, the reaction from the producers is not "let's hire better writers" because this costs too much. The reaction is more like: "Um...maybe if we added a Borg character with Really Big Breasts..." That may in fact get more people watching, but they won't be the same people that left the show for lack of acting and writing, and even they may eventually stop watching for the same reasons.
Re:Some are going to be pissed...
on
R.I.P. Iridium
·
· Score: 2
At least one of the Globalstar retailers is offering a $495 trade-in on a Globalstar phone. This is a pretty good deal, since the G* phones are much cheaper anyway, and by all accounts work much better. Plus, they're the only game in town at the moment, unless you enjoy GEO-style voice latency (eg:Inmarsat)
The link is worth it just for the animated GIF of the Iridium logo (the constellation Ursa Major) falling out of the sky.
Slashdot Mirror
First, a clarification to the above: Twitter also works just fine with delivery over IM, so any Jabber client like Google Talk (which has a nice client on e.g. Blackberries and works on iPhones et al) will do instead, so no need to use SMS if you don't want to. Plus one can go to one's Twitter page and see a running narrative of everyone being 'followed', so it can be used in a mode with no 'instant' delivery at all.
Which leads to the main point: It's wrong to get hung up on the 'immediacy' of Twitter, because there are other technologies that already do that well (IM, email...), and this is perhaps why I think a lot of people don't get it or dismiss it as pointless. The niche that Twitter fills is asynchronous communication between 2 or more people. It's publish-subscribe.
The problem with all the other forms of instant communication we have (phone calls, SMS message, IM, email...) is that they are all interrupts: people stop whatever they are doing when you initiate communcation this way. But often one just wants to update one's own status, or make an observation about something without causing an interrupt, but still allowing the set of potentially interested people to see your updates when they are available, ready and interested.
Example: Person A is going out to drinks after work with friends. B might join if time permits, but is on a coding deadline and will potentially work rather late. A twitters progress (leaving bar X, moving to Y where there is also food), which does not interrupt B. B reaches a good stopping point, checks A's twitters, is able to meet friends at Y later in evening.
That, and variations thereof, is what Twitter is good at.
Sadly, neither one is a good choice.
As usual, it's more fun for /.ers to have a religious debate about the relative merits of Java vs. C# rather than answer the actual question being asked. So the idea is to choose between them for the purposes of learning object-oriented programming, you say?
Either one will do almost equally well for a basic OO programming course, though you should be aware that despite all of the zealotism and hyperbole in the discussion here, they are *both* relatively poor choices for learning OOP. Both are designed to improve on the many shortcomings of C++ and C for large-scale development. As such, they are quite similar, and are roughly equivalently mediocre for erudition purposes. Furthermore, because they are meant as enterprise-class development langages, they have many features and formalisms (namespaces, assemblies, and particularly strong typing) which are designed to manage complexity and prevent errors for the 500,000 line code build. And if you're going to pursue a career in software engineering you're eventually going to want and need those things. But these same features simply get in the way and obfuscate when you're building a 200-line program to demonstrate an OO concept for a class. As the author of the Boo language (which compiles to the CLI) says in the Boo Manifesto, '"public static void main", that was a good one!'
A good choice for such a course would be Python or Ruby. Elegant, compact, forgiving, relatively pure (especially Ruby) and weakly typed, they are a pleasure to use. Ruby is so beautiful at times that you could weep. An even purer choice would be something like Smalltalk, but I think Ruby makes a better tool for teaching because it is also pragmatic enought to incorporate things like functional programming elements and idioms, which can sometimes throw light on the OO concepts and approaches by juxtaposition, and allow you to develop an understanding of when OO helps you get things done and when it gets in the way.
This doesn't solve your original problem, of course, because presumably you may choose only C# or Java. But do yourself a favor and try to bang out a few homework problems in Python or Ruby just to kick the tires. You'll be floored at how much more compact and lucid it is, and how flexible. Plus you will find in the long run that a feel for how you would implement something in a pure OO language actually greatly informs and improves your designs in languages like C# and Java, which will ultimately make you a much more skilled programmer. Which is, after all, the point.
I was at a talk by Mel Karmazin, chief executive of CBS, a few months ago. One of the questions from the audience was something like "as a network, do you feel threatened by the potential of these new Tivo devices for viewers to avoid commercials".
Turns out he has one. He said that if they found that consumers were skipping over the commercials they would find a way to put the advertising in the shows. Now, he may have just been kidding, or saying it for a bit of the shock value, but he seemed serious. That would suck.
I have a TiVo as well, and I absolutely use it to skip over intros, commercials, etc. You can take in an episode of South Park in 20 minutes. Unlike Replay, however, there is no button for 30 second skip; instead there is a 2x, 12x and 60x fast forward. I've actually come to prefer this, as I'll sometimes see something interesting or unusual zip by and go back to watch the commercial (some of them are more entertaining than the programming). So in the best of scenarios, widespread use of these devices will force commercials to be more entertaining/interesting so people will actually watch them.
how far are we from not needing the service?
What you propose is probably possible, but if you are getting program information from a separate source you would have to throw away the TiVo user interface completely. Most of the TiVo interface is built around the TV program's metadata (ie: actors, genre, when the program was made), which ClickTV may not be able to provide. Even if they could, it's not clear that you could easily map it (and maintain that map) to whatever TiVo's internal format is. So I would bet that you would have to abandon the TiVo's software and UI almost entirely, and substitute your own.
That would be a shame, because (1) TiVo designed an outstanding UI, and (2) most of the benefit of using TiVo (I've had one for a few months now) is in the interface and the TiVo service beyond the schedule information. Once it notices that you watch a lot of, e.g., Science Fiction shows, it starts recording other ones you might like in any free space it has. It makes suggestions based on what you've watched of other stuff you might like. It reminds you of recording conflicts when you schedule them, and does reasonably good priortization if you give it lots to record. Plus, it is a work in progress: it has an extremely active user community (check out the AVS Forum TiVo discussion group) that TiVo actually pays attention to, and it shows in the design. So if you are looking to hack it just to have a digital VCR, I'd wait, or get Replay instead (assuming *it* works in Canada). If you are interested in hacking it to avoid paying for the service, you'll be avoiding (IMHO) exactly what makes the device worth owning in the first place
Instead of worrying about the next kind of micro-storage device we can build into a digital camera, the industry should be focusing on taking it out entirely. This is what technologies like Bluetooth are supposed to make possible. You carry around storage in whatever form you like--your PDA, a 1-Gig drive in your backpack, an IBM microdrive on your belt--and the camera stores directly to the external device. This would also allow the camera to be smaller, use less battery, etc. In fact, if you used a color PDA, you potentially wouldn't need the color preview screen on the camera either.
Bluetooth products are mostly still in prototype, but AFAIK nobody is even talking about this with cameras. Seems like the way to go.
Perhaps this is the point, but creating a user-driven forum for discussing election issues won't do the majority of voters any good until we change the way the political system works.
Pre-Reagan, elections were all about party allegiance. Post-Reagan, as parties have drifted toward the center, they have been more about image (encouraged by burgeoning media coverage) and hotbutton issues like abortion. The one thing elections have never been about for most people is an objective evaluation of the candidates on all the issues, side-by-side, point-by-point. While it probably should be that way (the Democracy Project obviously wishes it so), it cannot for several reasons:
(1) People are on the whole impressionable, gullible and lazy, so they can be relied on neither to seek out objective criteria, nor to understand or take the time to evaluate them if they did.
(2) Most issues are quite complex, and require a lot of context and background knowledge. Take free trade: even an Econ 101 level discussion of comparative advantage or the lump-of-labor fallacy will cause most people's eyes to glaze over. So they leave the hard details of most issues to the politicians, and vote for them because, for instance, that person "cares about us, and won't let our jobs go overseas". Try an experiment: stop 10 people on the street, and ask them what Gore or Bush's position on free trade is.
(3) Politicians win elections by trying to be all things to all people. This is impossible if they take concrete positions, so they make vague promises that everyone would agree to (pro-education, create more jobs, etc) and avoid specifics whenever possible, relying on (1). So the current system gives them no incentive to devise a user-centered site where actual issues are discussed and positions taken.
This is all depressing, of course, but it seems to me that before you even have a chance to bootstrap voters into a world of active participation and ideal rational choice, you have to first dismantle the two-party, media-driven system that causes them to vote the way they do now.
Globalstar is another LEO satellite phone system, but it differs from Iridium in that it: (a) offers significantly cheaper and lighter phones, and (b) is not bankrupt and deorbiting its satellites. You can get a tri-mode Analog/CDMA/Satellite phone for about $1200, and there's no reason not to expect that price to fall further. They plan to roll out internet access later this year, at [a whopping] 9600 baud. I've used the phone for voice, and it's comparable to digital cellular in clarity. A really amazing device.
Admittedly $1200 isn't cheap, but I would imagine if you're the kind of person that is on boats a lot (your own or a cruise line's) it's not a big investment. Their website is http://www.globalstar.com
For those of you now wishing you had bought a new home so you could do this, there's a recent nice article in Ars Technica about how to wire your existing home properly. An interesting read.
It is actually a very thorough point-to-point remedy solution that compliments the Findings of Fact.
Agreed. In its goals, the DOJ has everyone's best interests in mind, including (as I am about to argue) Microsoft's, because it will force them to spend their energy on their products, and not the perpetuation of their monopoly.
It's important to remember what the DOJ is trying to remedy here. The point is not to punish Microsoft by blowing them into bits; the point is to keep them from harming other companies and consumers by abusing their monopoly power. This power stems IMHO primarily from their dominance of an office suite, a set of OSes (95/98, NT/2000) and a browser technology which have monopolistic-level market penetrations. However, the real power (and thus the potential for abuse) does not come from the quality of those products (as some of us know all too well), or even from their high level of penetration. It comes from MS's relentless tying and blurring of the lines between them, so that if you need to use one of them, you more or less roped into using them all. In other words, their incentive structure in that position is to invent new ways for you to need all things Microsoft, not to produce better products. That practice is what the DOJ wants to eliminate, and rightly so.
The originator of this thread is correct in that the post-split companies would have monopoly-level penetration in their of their product areas. However, they would have none of the ability to tie the products together, and indeed would not have any real ability or economic incentive to do so. The incentive they would be left with would be to compete by improving their products.
There are a lot of smart and talented people at Microsoft. Until now much of their collective coding hands have been tied insofar as the primary business strategy has been the market dominance of the Windows/Office empire. Split up, and with remedies properly applied, they will be forced to focus on the quality of their product to maintain market share. Everybody benefits from this, including, perhaps especially, Microsoft. It will mean that in the long run you may be seeing more, not less, of things MS. One of the reasons that surveys are finding consumers to be deeply ambivalent about the DOJ's actions is that there are huge network effect benefits to having almost everyone running the same OS and office software, and people are worried that the wrong kind of remedy will somehow mess up those benefits. If the 3-way split forces MS1, 2 and 3 to improve their products, then it seems to me those same consumers should embrace those products more, not less. So if what you secretly wish for at night is for Bill and his buddies to go down in flames, consider the possibility that the split will unlock a lot of potential in the new companies. Remember: "If you strike me down, I shall become more powerful than you can possibly imagine."
The system still suffers from a lack of drivers, but those will get there eventually.
Sadly, all new OSes are plagued by a lack of drivers. BeOS still doesn't run right on my [rather generic] laptop, because of a lack of available drivers, and the rollout of new ones is sloooow.
OSes are notoriously hard to get right, take years to develop, and when they are finally ready to show the world, only a handful of drivers are available (the few they have been developed with). This leads me to think that a nice project would be a sort of open driver API suite that the major hardware manufacturers could commit to supporting. (Sort of like DirectX, but with the standard controlled by a consortium, and for all peripherals, not just for graphics cards.) That way the OS could ship with a set of "generic" API-supported drivers that would work at least OK with most devices.
This would make the hardware manufacturers happy, because they wouldn't have to deal with every new OS on the block separately, and wouldn't have to release their proprietary hardware specs to (from their perspective) fly-by-night alternate OS developers. It would also make the alt-OS users happy, because they would be much more likely to be able to get drivers quickly. Such an API would obviously not result in optimal drivers, would be slower than custom-written drivers and would not be able to take advantage of the fancier aspects of the hardware, but at least it would be easy for the OS to provide generic-enough drivers that you could try out the OS and kick its tires without doing it in standard VGA with no printer, USB, PCMCIA and network support. Trouble is, I don't know enough about the realities of driver development to know if this is even a realistic idea. Or perhaps something like this exists already...?
Imagine a site like mysimon that compared millions of products from thousands of stores.
A laudable goal, but in order for this to work on that kind of scale you need to establish a standard, and then you need the retailers to participate. This is a classic bootstrapping problem, because retailers hate price transparency--no seller is going to sign up for a system where it has to do extra work to compete against potentially millions of other retailers based on price. However, if you can make the system large enough that consumers start to go there first, all of a sudden the retailer will need to participate, since it will start to lose customers to the service otherwise.
Then, instead of making you do all the work of signing up and purchasing from each store, the program could use a single shopping cart, and negotiate the sale from each of the stores that had the best prices, using the same user information
The single shopping cart has the same problem: in a world of increasing price transparency, one of the few "sticky" features an e-tailer has left is that you've already gone through the tedious job of filling out a user profile on their site, and not someone else's. Are you really going to go take 15 minutes to fill out a profile at Egghead.com to shave a few bucks off the price of a digital camera if you already have a profile at PC Zone? Probably not. Plus, there are perhaps some privacy issues.
Of course, it would be a truly wonderful thing to have if you could find a way to make it happen. I just wonder if it's realistic...
START is a nice attempt at natural language parsing, but (from my understanding) once it has made a sophisticated identification of what you mean, it can only direct you to the canned answers it has for various questions. What it appears to be missing (someone correct me if it has this facility--I only had time to skim their website) is any real-world 'understanding' of the relationships between the objects it is manipulating. A nice AI-like project would be to try to overlay some real-world intelligence so that it could deduce answers from relationships in its datasets.
For example, START is able to answer:
Which country has the largest population?
and
Which country has the largest area?
but not:
Which country has the largest population density?
because it does not 'know' that population density is population/area. Of course, this is a hard thing to tackle, but a proof-of-concept could be done with a limited object domain (like geograhpical data).
The same knowledge of relationships could allow search engines to be a lot more intelligent about what you are actually looking for, instead of relying purely on keywords and statistics. In other words, I should be able to go to askjeeves.com and ask "Where is the island of Java?" and not have almost all of the answers have nothing to do with Indonesia.
Of course you realize that the downward spinoff spiral is an inevitable consequence of the economics of television, since TV producers like to think they are getting something for nothing with a spinoff. The problem is that after the show is cast with a fraction of the original's budget, you end up with cheap[er] sets, worse acting and worse writing than the original. TV executives fail to realize what drove people to watch the show in the first place.
When viewers start abandoning the spinoff in droves, the reaction from the producers is not "let's hire better writers" because this costs too much. The reaction is more like: "Um...maybe if we added a Borg character with Really Big Breasts..." That may in fact get more people watching, but they won't be the same people that left the show for lack of acting and writing, and even they may eventually stop watching for the same reasons.
At least one of the Globalstar retailers is offering a $495 trade-in on a Globalstar phone. This is a pretty good deal, since the G* phones are much cheaper anyway, and by all accounts work much better. Plus, they're the only game in town at the moment, unless you enjoy GEO-style voice latency (eg:Inmarsat)
The link is worth it just for the animated GIF of the Iridium logo (the constellation Ursa Major) falling out of the sky.