Perhaps another take is that some of Apple's practices, which work fine with their current market share, would be considered draconian if they were the dominant vendor.
I imagine they'd be able to supplement it as they like. It might even be mandated some governments; I understand that's done with flour in many places, to improve public nutrition.
We have hunger, diseases, war... and all these people want to do is to get everybody to stop eating animals. Considering that it was likely the consumption of large amounts of animal protein that allowed humanity to evolve rather rapidly in the last stage of our evolution, I find PETA's goals rather ironic. Let's focus on hunger. Farming of animals can basically be considered an inefficient process for turning one type of food (grains) into other types of food (meat, dairy).
The range of successful and healthy vegetarians and vegans has shown that meat is not necessary for our survival. Do you then approve of letting other people starve because you want to get your protein in one form and not another?
They should continue, "unfortunately, there are a fair amount of countries that don't have access to the sun. " Which is a pretty ridiculous argument: by definition, all nations and all continents on this earth have access to the sun, even Antarctica. Some nations, due to their geographic position on the globe, simply have better "sunlight" than others. Event then, solar energy is available pretty much all around the world. For instance, one of the most important country in Europe for solar energy is Germany, which is not especially noted for its warm climate... Methinks that in this context "don't have access" means "don't have access to enough land to make this viable". I guess many countries are fortunate enough that most of their land is arable, and thus is probably better used to produce crops. The article suggested it for countries which have large amounts of desert, which is the perfect place for this technology.
Sure they look more attractive when sold in stores, but that's not misleading. They also look more attractive wherever you use them. I've been using one for a while, and regardless of whether the actual colours output by the lcd are better or worse, the glossy finish really adds depth to the colour. I have to agree with earlier posters that reflectiveness hasn't been an issue for me at all.
You need to sign up to download music, but the first month is free (no obligations, or at least it was when I signed up). It's made very easy to preview music. They have a pretty diverse range. The first month I was disappointed with my choices, but since then I put more effort into looking, and I've found probably one or two artists a month that I really got into. I also feel that I've barely scratched the surface. Anyhow, best of luck!
eMusic is a tiered subscription service, where you pay monthly and get a number of songs you can download each month. This amounts a cheaper price per-song, and files are mp3s with no encumbering DRM.
The downside is that although they have a lot of music, they don't have major labels or big pop artists. For me, this has actually been a real positive, since it's meant that I've found excellent artists I never would have run into or heard on the radio.
It seems that the company should provide a computer, OS, virus and spyware checkers, some automated backup facility, and all the tools that people need to do their trade. Make a DVD which slurps and reinstalls the whole box for them, then have them exercise personal responsibility. Avoid pirated software by firstly allowing people to make a business case for new software they want/need, and secondly kicking their asses if they install pirated stuff nonetheless. Oh yeah, and respect people by keeping them busy with meaningful work, and maybe they won't install crap.
Then again, a friend who worked in Spain had his company just buy each employee their own laptop, but not any software or tools they needed. It was up to them to acquire (read: pirate) whatever they needed, with the implication that they were liable if they were caught. Dodgy!
To expand on your post, logic programming languages have an inbuilt form of search. You can program rules, and these constrain the solution space of correct answers. Your code then directs the search, which is depth-first by default, and makes sure that it will explore the space efficiently and give you an answer in reasonable time.
One big difference in comparison to other language families is that the search could have multiple answers, and if the search goes down a branch which "fails", it can also backtrack and try another branch. This type of behaviour is inbuilt, and very useful for some types of problems. For other types of problems, you just end up coding in pretty much the same way that you would for imperative languages. You do get the benefit that variables are (usually) immutable, which can avoid some forms of programming errors, and that the language usually lets you program at a very high level.
Statically typed logical languages with mode systems (like Mercury) can have very good performance, and give the compiler enough information to provide warnings which catch many many more programming bugs than normal, all before your code is even run. The real downside for some applications is the small size of the ecosystem for 3rd party libraries compared to C/C++/Java/Python/Perl/etc. It may or may not be a deal-breaker for whatever you need to get done.
On the Blu-Ray and downloads topic, the argument is not about the merit of the technology, but rather on the long-term value of the technology to its investors (i.e. Sony and others). If Sony has mis-predicted the near-future marketplace, and is unable to provide good reason for us to buy content on their discs (say if everything is downloadable in good enough quality), they might have paid more than it was worth to win the format war.
On the broader gripe, people's dissatisfaction (within limits) is great. It drives consumption for new things, creates marketplaces that new innovations can slot into, and basically moves things forward. I for one am extremely glad that people today aren't happy with yesterday's horse and cart. Aren't you?
Suppose you put strong gyroscopes inside the controller, and could power them adequately. Couldn't you then provide sudden resistance at certain points in time? Imagine playing a tennis game, where it really felt like you hit the ball mid swing. Or playing a sword game, and feeling like you've struck something, as your controller suddenly stops. I suppose they can only resist acceleration, rather than provide acceleration, but it would still add a lot of realism to many games.
Can someone with a better grasp of their physics comment?
My experience is identical. I do a lot of python, C and C++ dev work for research, and I've lost days and days of time trying to get new enough libraries installing correctly, posting to forums, finally working things out. I could have installed fresh and configured ubuntu countless times over in any one of those days.
All this manual configuration can also be fragile. After running a well configured system for a long time, a change I made broke it utterly. It was so time-consuming to try to recover from that I ended up working in a linux virtual machine instead, which was so sweet I very nearly just reinstalled with linux. Only now with a fresh install of Leopard did I go through the pain again to configure OS X as I actually need. It's improving, but it's still very painful.
The current package systems for OS X just don't cut the mustard.
I'd have to disagree about fink. Despite the fact that the mac environment is so homogeneous, fink has always been really flakey for me. Perhaps it's better now, but whenever I've used fink I've had binary packages fail to install regularly, sometimes putting the package system into a bad state which is non-obvious to undo. Macports is a bit more reliable, and seems to have newer packages, but even so I still have to sometimes edit the code for packages which didn't compile correctly or to perform a compilation step manually to get something installed.
Debian and Ubuntu are so much more polished in terms of their package systems. At the moment, I can approximate the experience of installing my needed packages as I would on these linux systems, but let's not say that the experience is equivalent.
I can't comment on the Apple's own stores, but my experiences with Apple resellers in Melbourne seems to match this. I enjoy the design of their products, which are a pleasure to use (though ubuntu is less effort than OS X for my research), but in some stores the staff really seemed arrogant. I hated buying from them, but did anyway, since at the time there were few resellers to choose from. They knew you wanted the product already, and figured you'd buy it whether or not they treated you well.
Luckily there's many more resellers these days, including many stores which used to only sell PCs, so service has improved a lot.
A lot of countries have anti-discrimination laws. You would be on very sticky ground if you rejected a candidate on the grounds of information you had gleaned online (esp. if it was related to a class of discrimination). For instance if someone's online diary said "I plan to start a family in a year or two" You could not ask this type of question in an interview and if you rejected them by knowing that this was their intention you could end up being sued.
These laws need evidence to be useful though. Suppose a potential employer finds something online they don't like. They're under no obligations to hire anyone in the first place. In cases of genuine discrimination, they can only get caught by anti-discrimination laws if they're foolish enough to disclose some discriminatory grounds as the reason they didn't hire.
We all take pride in our heritage. For a long time China was in many ways in front of the rest of the world. At some point they stagnated technologically, and eventually they were essentially humiliated by foreign powers which were much more advanced and powerful, forcing them to do business (think opium, Hong Kong, Maccau...). This cultural wound is still open and fresh. They want to prove themselves to the world again, and on the whole, their government has been steadily improving. The word from students in China is that more open discussion and criticism of government is more and more readily tolerated, that the more extreme talk may still be censored, but that nobody's just disappearing. There's a feeling that things are moving forward, despite problems of corruption and inequality.
Furthermore, China has a long history of warring states, turmoil and revolution. In comparison, the stability and progress that they enjoy today is a thing to be treasured. They are grateful to the current government that there has not been another cultural revolution, and that the last 30 years have provided peace and prosperity to many, especially urban Chinese.
In light of all this, your post is basically calling for more open defiance and conflict. Honestly, the people as a whole aren't ready for that. BUT, the increasing criticism of government can only be a good thing. Once the reins have been loosened enough on this form of criticism, then expect social change to happen more rapidly, as people are more aware of problems, more eager to voice their concerns, and put more pressure on their government to fix its inadequacies.
Well, I guess the actual moments that you level are great and fun, it's just all those pesky ones that come before the next "Ding!" that suck the life out of you.
I'd try ClockingIT, which is a rails-based interface. I had been using roundup, which is nice because it's very easily customizable, but moved to ClockingIT because of the free hosting they provide, and the simple workflow. The nicest thing is that you interactively use the page itself to log your hours, so you don't have to manage the time side of things. It supports multiple customers, projects, milestones, etc, and has a built-in wiki, forum and reporting interface. It's really quite nice.
The real problem with VB6 from my experience has nothing to do with the language itself, but rather with two related issues. On the management side, a good friend worked for a software dev shop whose idea was: hire talented architects, then write it in VB so that even cheap, unexperienced coders can do the job. Surprise, surprise, the cheap, unexperienced coders made such a hash of it that they needed to spend the money for experienced and talented coders anyway. By then the cost to the company meant its creditors started to scale everything down. Coder side: people may be under the illusion that because they can code something (good), their code is just as good as someone with better depth of knowledge and experience (bad).
Hence the hostility from experienced coders, who are undervalued by management due to that fallacy. I really don't think anyone's unhappy that more people can program and script things themselves. As long as they're maintaining it, it's all good. People just want to be valued by their actual contribution.
It also assumes that no hierarchy is being used. Clearly there is a management hierarchy for the Linux kernel, with Linus at the top. He's not communicating with every developer directly, nor should he. What The Mythical Man Month is saying is that a project with this many developers would be infeasible without a large hierarchy supporting the bottom level developers. So this story is just: more developers, more management. What a surprise, hey?
Personally, I don't see any conflict between the world being created by some God, even in 7 days, and its being formed over billions of years by natural processes. One is a faith based way of experiencing the world, the other is a sensory based, practical, and logical way. They are both useful.
I disagree. These two worldviews are constantly battling with each other, and have been for a very long time (note the historical persecution of breakthrough scientific minds for religious reasons). They suggest very different ways of living your life, one way based on evidence and reasoning, the other based on dogmatism and saying, "This much knowledge is enough for me." Which one of these two views is going to provide our next medical breakthrough? Which one has historically driven humanity forwards? I'm not so sure they're both useful.
There's a data sparseness issue, from a statistical perspective. There's only 122 data points. Can you really make predictions from such a limited set? Even if you do find many, the "interesting" wars (WW1, WW2, Vietnam, etc.) may all be statistical outliers if the bulk of the data is made up of many smaller conflicts.
There's also a methodological issue. Lets say you use some data for testing, and hold out some for training. You can even shift these and use 10-fold cross-validation or whatever. You test your model, and it's bad, so you go and do some remodeling, and come up with one that scores better on this data. Despite your efforts, you've just successfully tuned yourself to your data, by preferring the model that worked better on it. The only way to know how you really would have gone is to hold out another separate random subset of the data initially. You then train and cross-validate on the rest, and finally, on your best model, you test it on the extra holdout and work out how your model went.
Did they do that here? I don't know, but the need to hold out enough wars as data points to get confidence in the model makes data sparseness even worse. Put simply, there's not enough data to make any useful conclusions from these models.
Slashdot Mirror
Perhaps another take is that some of Apple's practices, which work fine with their current market share, would be considered draconian if they were the dominant vendor.
I imagine they'd be able to supplement it as they like. It might even be mandated some governments; I understand that's done with flour in many places, to improve public nutrition.
Sure they look more attractive when sold in stores, but that's not misleading. They also look more attractive wherever you use them. I've been using one for a while, and regardless of whether the actual colours output by the lcd are better or worse, the glossy finish really adds depth to the colour. I have to agree with earlier posters that reflectiveness hasn't been an issue for me at all.
I have to agree. Most educational games that I tried were neither good games nor good educational tools.
You need to sign up to download music, but the first month is free (no obligations, or at least it was when I signed up). It's made very easy to preview music. They have a pretty diverse range. The first month I was disappointed with my choices, but since then I put more effort into looking, and I've found probably one or two artists a month that I really got into. I also feel that I've barely scratched the surface. Anyhow, best of luck!
eMusic is a tiered subscription service, where you pay monthly and get a number of songs you can download each month. This amounts a cheaper price per-song, and files are mp3s with no encumbering DRM.
The downside is that although they have a lot of music, they don't have major labels or big pop artists. For me, this has actually been a real positive, since it's meant that I've found excellent artists I never would have run into or heard on the radio.
I highly recommend it.
It seems that the company should provide a computer, OS, virus and spyware checkers, some automated backup facility, and all the tools that people need to do their trade. Make a DVD which slurps and reinstalls the whole box for them, then have them exercise personal responsibility. Avoid pirated software by firstly allowing people to make a business case for new software they want/need, and secondly kicking their asses if they install pirated stuff nonetheless. Oh yeah, and respect people by keeping them busy with meaningful work, and maybe they won't install crap.
Then again, a friend who worked in Spain had his company just buy each employee their own laptop, but not any software or tools they needed. It was up to them to acquire (read: pirate) whatever they needed, with the implication that they were liable if they were caught. Dodgy!
To expand on your post, logic programming languages have an inbuilt form of search. You can program rules, and these constrain the solution space of correct answers. Your code then directs the search, which is depth-first by default, and makes sure that it will explore the space efficiently and give you an answer in reasonable time.
One big difference in comparison to other language families is that the search could have multiple answers, and if the search goes down a branch which "fails", it can also backtrack and try another branch. This type of behaviour is inbuilt, and very useful for some types of problems. For other types of problems, you just end up coding in pretty much the same way that you would for imperative languages. You do get the benefit that variables are (usually) immutable, which can avoid some forms of programming errors, and that the language usually lets you program at a very high level.
Statically typed logical languages with mode systems (like Mercury) can have very good performance, and give the compiler enough information to provide warnings which catch many many more programming bugs than normal, all before your code is even run. The real downside for some applications is the small size of the ecosystem for 3rd party libraries compared to C/C++/Java/Python/Perl/etc. It may or may not be a deal-breaker for whatever you need to get done.
http://www.youtube.com/watch?v=IEZcE1jixh4
Half whale... half cow... but tastier than both!
On the Blu-Ray and downloads topic, the argument is not about the merit of the technology, but rather on the long-term value of the technology to its investors (i.e. Sony and others). If Sony has mis-predicted the near-future marketplace, and is unable to provide good reason for us to buy content on their discs (say if everything is downloadable in good enough quality), they might have paid more than it was worth to win the format war.
On the broader gripe, people's dissatisfaction (within limits) is great. It drives consumption for new things, creates marketplaces that new innovations can slot into, and basically moves things forward. I for one am extremely glad that people today aren't happy with yesterday's horse and cart. Aren't you?
Suppose you put strong gyroscopes inside the controller, and could power them adequately. Couldn't you then provide sudden resistance at certain points in time? Imagine playing a tennis game, where it really felt like you hit the ball mid swing. Or playing a sword game, and feeling like you've struck something, as your controller suddenly stops. I suppose they can only resist acceleration, rather than provide acceleration, but it would still add a lot of realism to many games.
Can someone with a better grasp of their physics comment?
My experience is identical. I do a lot of python, C and C++ dev work for research, and I've lost days and days of time trying to get new enough libraries installing correctly, posting to forums, finally working things out. I could have installed fresh and configured ubuntu countless times over in any one of those days.
All this manual configuration can also be fragile. After running a well configured system for a long time, a change I made broke it utterly. It was so time-consuming to try to recover from that I ended up working in a linux virtual machine instead, which was so sweet I very nearly just reinstalled with linux. Only now with a fresh install of Leopard did I go through the pain again to configure OS X as I actually need. It's improving, but it's still very painful.
The current package systems for OS X just don't cut the mustard.
I'd have to disagree about fink. Despite the fact that the mac environment is so homogeneous, fink has always been really flakey for me. Perhaps it's better now, but whenever I've used fink I've had binary packages fail to install regularly, sometimes putting the package system into a bad state which is non-obvious to undo. Macports is a bit more reliable, and seems to have newer packages, but even so I still have to sometimes edit the code for packages which didn't compile correctly or to perform a compilation step manually to get something installed.
Debian and Ubuntu are so much more polished in terms of their package systems. At the moment, I can approximate the experience of installing my needed packages as I would on these linux systems, but let's not say that the experience is equivalent.
I can't comment on the Apple's own stores, but my experiences with Apple resellers in Melbourne seems to match this. I enjoy the design of their products, which are a pleasure to use (though ubuntu is less effort than OS X for my research), but in some stores the staff really seemed arrogant. I hated buying from them, but did anyway, since at the time there were few resellers to choose from. They knew you wanted the product already, and figured you'd buy it whether or not they treated you well.
Luckily there's many more resellers these days, including many stores which used to only sell PCs, so service has improved a lot.
These laws need evidence to be useful though. Suppose a potential employer finds something online they don't like. They're under no obligations to hire anyone in the first place. In cases of genuine discrimination, they can only get caught by anti-discrimination laws if they're foolish enough to disclose some discriminatory grounds as the reason they didn't hire.
We all take pride in our heritage. For a long time China was in many ways in front of the rest of the world. At some point they stagnated technologically, and eventually they were essentially humiliated by foreign powers which were much more advanced and powerful, forcing them to do business (think opium, Hong Kong, Maccau...). This cultural wound is still open and fresh. They want to prove themselves to the world again, and on the whole, their government has been steadily improving. The word from students in China is that more open discussion and criticism of government is more and more readily tolerated, that the more extreme talk may still be censored, but that nobody's just disappearing. There's a feeling that things are moving forward, despite problems of corruption and inequality.
Furthermore, China has a long history of warring states, turmoil and revolution. In comparison, the stability and progress that they enjoy today is a thing to be treasured. They are grateful to the current government that there has not been another cultural revolution, and that the last 30 years have provided peace and prosperity to many, especially urban Chinese.
In light of all this, your post is basically calling for more open defiance and conflict. Honestly, the people as a whole aren't ready for that. BUT, the increasing criticism of government can only be a good thing. Once the reins have been loosened enough on this form of criticism, then expect social change to happen more rapidly, as people are more aware of problems, more eager to voice their concerns, and put more pressure on their government to fix its inadequacies.
Well, I guess the actual moments that you level are great and fun, it's just all those pesky ones that come before the next "Ding!" that suck the life out of you.
I'd try ClockingIT, which is a rails-based interface. I had been using roundup, which is nice because it's very easily customizable, but moved to ClockingIT because of the free hosting they provide, and the simple workflow. The nicest thing is that you interactively use the page itself to log your hours, so you don't have to manage the time side of things. It supports multiple customers, projects, milestones, etc, and has a built-in wiki, forum and reporting interface. It's really quite nice.
The real problem with VB6 from my experience has nothing to do with the language itself, but rather with two related issues. On the management side, a good friend worked for a software dev shop whose idea was: hire talented architects, then write it in VB so that even cheap, unexperienced coders can do the job. Surprise, surprise, the cheap, unexperienced coders made such a hash of it that they needed to spend the money for experienced and talented coders anyway. By then the cost to the company meant its creditors started to scale everything down. Coder side: people may be under the illusion that because they can code something (good), their code is just as good as someone with better depth of knowledge and experience (bad).
Hence the hostility from experienced coders, who are undervalued by management due to that fallacy. I really don't think anyone's unhappy that more people can program and script things themselves. As long as they're maintaining it, it's all good. People just want to be valued by their actual contribution.
It also assumes that no hierarchy is being used. Clearly there is a management hierarchy for the Linux kernel, with Linus at the top. He's not communicating with every developer directly, nor should he. What The Mythical Man Month is saying is that a project with this many developers would be infeasible without a large hierarchy supporting the bottom level developers. So this story is just: more developers, more management. What a surprise, hey?
I disagree. These two worldviews are constantly battling with each other, and have been for a very long time (note the historical persecution of breakthrough scientific minds for religious reasons). They suggest very different ways of living your life, one way based on evidence and reasoning, the other based on dogmatism and saying, "This much knowledge is enough for me." Which one of these two views is going to provide our next medical breakthrough? Which one has historically driven humanity forwards? I'm not so sure they're both useful.
There's a data sparseness issue, from a statistical perspective. There's only 122 data points. Can you really make predictions from such a limited set? Even if you do find many, the "interesting" wars (WW1, WW2, Vietnam, etc.) may all be statistical outliers if the bulk of the data is made up of many smaller conflicts.
There's also a methodological issue. Lets say you use some data for testing, and hold out some for training. You can even shift these and use 10-fold cross-validation or whatever. You test your model, and it's bad, so you go and do some remodeling, and come up with one that scores better on this data. Despite your efforts, you've just successfully tuned yourself to your data, by preferring the model that worked better on it. The only way to know how you really would have gone is to hold out another separate random subset of the data initially. You then train and cross-validate on the rest, and finally, on your best model, you test it on the extra holdout and work out how your model went.
Did they do that here? I don't know, but the need to hold out enough wars as data points to get confidence in the model makes data sparseness even worse. Put simply, there's not enough data to make any useful conclusions from these models.