Apple takes the view that every separate boxed sale of OS/X is an upgrade because to run it you need an Apple machine, and that machine necessarily came loaded with some prior version of MacOS.
However these boxes contain a full version of Mac OS/X, not an upgrade kit. You can install it on a freshly wiped Apple box or in an emulator, it doesn't request a copy of the previous version's media.
Your first sentence in this light makes no sense. At any rate if you buy a boxed version of OS/X from the Apple store, it doesn't say "Upgrade" it says "Buy OS/X, single or family license".
Can you give an example of a good use of TPM from the consumer point of view? I thought not.
TPM is a fundamentally restrictive technology. It's there to prevent, not to help. Consumers are universally better off without it rather than with it. It is also in my opinon a fallacy that content providers require it to make money. As an example, international DVD sales really took off with region coding being defeated.
You are correct that there is no need to agree to the GPL to use software licensed under it, however in general to use software you need a license, and the GPL grants you unilaterally all rights to use the software.
In other words, you don't need to agree to the GPL because there is no counterpart agreement on your part as far as usage is concerned, but you need to be aware that the software that you are using is in fact Free to use.
Most Windows installers (even Free ones!) that F/OSS software use have a mandatory page for the licence. Putting the GPL there is not the best way, but it's OK. It should be only a page saying "this software is licensed under the GPL, therefore you can use it to your heart content. Have fun" sort of deal.
Actually (and I work in this area as a researcher), computers suck at face recognition far far less than humans.
You may think that you are good at recognising people's face but the fact is you probably can only recognise 100-1000 faces all told, of people you know rather well, and doing it in a crowd is very hard. Recognising the face of someone you met once 10 years ago is for most people almost impossible.
The US driving licence face recognition system works on litteraly millions of people, and copes very well with eyeglasses, facial hair & different hairstyle and aging. It is used to prevent people from acquiring more than one driving license in different states.
1- the license 2- the speed (OS/X's kernel is slow as molasses), not to mention how incredibly memory hungry OS/X is. I have a Linux 64-bit desktop with 512 MB of Ram that hardly ever swaps even in heavy use. MacOS thrashes about swapping with 768 MB of RAM after about 2h of work. When I have even a modest compile job I leave it thrashing around overnight rather than face the waiting 3- 64-bit. Despite what you hear, even Tiger is not 64-bit. Of course this doesn't apply to the new MacBook Pro or iMac, which sport a dual 32-bit CPU, but EMT64 core duo CPUs are in the works, I hear. I bet OS/X 10.5 still won't be 64-bit throughout. 4- the GUI. Honesly I prefer gnome or even KDE these days to slow-as-molasses, full of silly eye-candy you can't turn off, no way to get focus-follow-mouse Aqua or whatever the graphics layer on OS/X is called. 5- Poor compatibility. At my workplace and in general in my field, very few people run OS/X for the above reasons, and almost everyone runs Linux. 6- I'm sick of recompiling all the free apps under the sun to get the software I want. I'm sick of waiting for OpenOffice 2.0 to be supported on OS/X, I'm sick of having to rely on closed-source, third party shareware to do things that are taken for granted under Linux. I hate the OS/X version of most of the CLI software there is. Apple's Python is unusable on the command line, there is no command line editing or interactive history ! 7- The proprietary apps aren't worth it. MS-Office is superceded by OOo. The Gimp is enough for me over photoshop (I must have tried it once) and I don't even need Illustrator anymore thanks to Inkscape. Mac have no games, BTW (unless you are happy to pay full initial price for 3-year old games).
I have a lowly iBook 12", bought on the recollection of what NeXTStep was like about 15 years ago. I was amazed that to see that most things were there but that Apple had seen fit to make the N/S interface builder format incompatible and unreadable (and not for lack of trying). I could recompile none of my old apps. As a result I completely gave up on XCode.
I'm switching back to Linux on that notebook as soon as suspend-to-ram and wireless both work (stupid proprietary controller).
OS/X is way way way overrated. Sure it's better than windows, but that's about it, and not even on all points -- for example why can't I do reliable suspend-to-disk on my laptop ?
If you know your way around/bin/sh, and have for decades, then clearly Linux is a good choice for the desktop because it's way nicer than any of the proprietary Unices for that. On HPUX, Irix and whatnot you get X and mwm, that's it. On Solaris you might get a bit of Gnome if you are lucky. The exception is OS/X, of course, but that won't run on cheap hardware they have lying around everywhere. Notice that in this case Windows is a very very poor choice.
On the other hand if you have experience with the sheer unshakable predictability, usefullness and stability of Solaris in the server room, then compared to that good ole Linux is a bit of a toy to be honest (although slowly improving).
So JPL is in the situation where they like Linux on the desktop but not in the server room, and it makes perfect sense.
Many graduate students in applied mathematics and statistics write their new code in these languages, especially since one can find a excellent F95 compiler, free for non-commercial uses on Linux.
By non-democratic I meant that the decision to build the first, and then many nuclear plants in France was never the subject of debate in parliament, let alone amongst the people. It was made between ministers, some EdF executives and also a special-purpose public body akin to the DoE, called the CEA (atomic energy commission). The CEA has a tradition of secrecy as they were involved in (nuclear) weapons manufacturing as well. The programme was always presented as an absolute necessity and a "fait accompli".
EdF/government manufactured consent through running paid adds on prime-time TV in the 1970s, touting alternative energy sources like wind and solar as well as nuclear. However the wind and solar powerplants were tiny and for show, while the nuclear programme was huge and real.
If you talk to people who were watching French TV in the seventies they probably remember them, they were very effective adds. They appealed to the "patriotic" sense of duty and ingenuity in a subtle way. The slogan was something like "in France we don't have much oil but we have ideas". As a result there were effectively few protests.
The problem I see with this approach is that it made it very hard to have a reasonable and informed point of view. Some journalists tried quite hard to counterbalance the gov/EdF/CEA point of view on biased scientific and emotional terms. Some papers printed nice charts that showed which areas would become lethally irradiated should such and such powerplant blow up. Others wrote book-length rants about how the technology was dangerous, unpleasant and generally a disgrace.
Almost no one thought of organising any kind of reasonable and informed debate on this issue. It was just hysteria that got nowhere.
Now the situation is much better. It's actually quite easy to go and visit a nuclear plant (i've done it several times) and to get relatively accurate information about what it being done with the waste. Most of the dangerous waste is reprocessed and the rest is stored. In fact France reprocesses the waste of several countries (and Greenpeace always protests when a ship comes in full of nuclear waste).
France has a huge leading technocracy. The decision to go mostly nuclear for electricity distribution was facilitated by there existing only one state-owned electricity producer (Electricité de France a.k.a. EDF) and by a mostly non-democratic decision-making process.
On the other hand there were never any huge, organized anti-nuclear protests in France, which was hit very hard by the first oil price hike in 1973. Anti-nuclear protests in recent years have been confined to sites where nuclear wastes were due to be dumped. However protesters were able to convince government to stop building new plants for many years, probably as there were enough capacity anyway.
Specifically, nearly all of OS/X applications, the kernel and the GUI are all 32-bit on all Macs. There are only two versions of OS/X, one for servers, and one for laptops/desktops. Since there are G4 versions of either, necessarily all software has to be 32-bit only. This is why when Tiger (10.4) came out all the benchmarks in magazine showed no speed difference between "32-bit" 10.3 and "64-bit" 10.4.
However, OS/X 10.4.x is able to run 64-bit apps that can access more than 4GB of memory, but those don't have access to the GUI directly. Console apps only! If you need a GUI in front of that 64-bit app, prepare to handle the joys of ICP (inter-process communications).
Well, maybe with enough muscle mass one would be able to use time dilation effect by pushing real hard in the direction of the switch and cut back the time needed to 50 years?
I basically agree with what you wrote, indeed Apple notebooks especially PBs simply were the best laptop until about 12-18 months ago. In particular PC laptops had terrible battery life (60-90 minutes, typical...) and were prone to actually burning your lap...
However things have changed very rapidly in the PC world. Now PC laptops have better features (larger screens, DVD burner standard, PCMCIA slot, wireless, BT, excellent build quality (IBM, Fujitsu, etc), faster CPU, more RAM, more disk, 64-bit, now dual-core etc etc etc) AND they have better battery life AND they are cheaper. I'm not listing OS/X because PC laptops by and large have Linux and in my mind the two are equivalent (for what I'm doing, may not be true for all).
It will be good to see what Apple is going to offer soon.
Yes, but Apple is screwed. Apple notebooks sales used to drive the company, now that everyone is waiting for Intel, everyone also acknowledges that the powerbook/iBook lines are way underpowered. Yet people with powerbooks by and large need the pro apps, such as the Adobe line.
These won't be coming in Intel form for months. I would expect a last speed bump on the PBs to bring them close to 2GHz, and an Intel iBook. They'll look great and have better raw performance than the PBs for a while, but they won't run the 3rd party pro apps satisfactorily, so they won't cannibalise PB sales to come. PB owners will wait a few more months for an even more powerful Intel chip (64-bit, dual core), and the pro apps to come with them, such as Adobe CS3.
iBooks owner will have to weather the brunt of the conversion to Intel. For a long while they will have precious few apps to run, except very slowly through Rosetta. It will be a nice laptop with few apps (unless they dual-boot to Windows!) I don't think it can be very successful just now. In 6-12 months it will be a different story.
Apple has money in the bank, they can weather the next hard few months. This upcoming iBooks is going to be more of a tech preview laptop than a real product. Expect fanzine to be full of impressive synthetic benchmarks, whipping up a huge frenzy for when the Intel PB line is finally released.
Having a real Intel product available also means Apple can put pressure on 3rd party software houses to release Intel version of their software, as there will be demand, and more readily available development platforms.
Wait, you are saying that Apple notebooks are better designed because there is more margin than for PC notebooks? Yet the iBooks are among the cheapest notebooks around. By your argument they should be using very cheap materials (to secure larger margins) and so extremely flaky and flimsy, yet they are not.
Apple has had a long field day with their notebooks because they were well designed to start with, they have not fiddled with their design as much as PC manufacturers, and they are able to charge for extras.
In contrast there is very little differential in the PC world among notebooks (they all look and feel the same, pretty much), so manufacturers have had to compete on features. This is why it is impossible to get away with shipping only a combo DVD reader/CD burner in a current PC notebook, yet this is what Apple is still doing with all 12" iBooks and some 12" PowerBooks.
Will things change when moving to Intel? I think Apple will be feeling more competitive pressure now that their main performance indicator (the CPU make and speed) will be directly comparable with PCs. It will be interesting to see.
Just in case one has been living under a rock for the past few years, CSS, the DVD encrypting scheme has been broken a long time ago.
On a general-purpose PC, including Macs, BeBox and whatnot, DVD region coding can be bypassed any old way, for example using VLC. RPC1 or 2 don't serve any purpose anymore.
Give actual examples please. I don't recall Canada, Germany or France voicing any critic when the US invaded Afghanistan. That was the right thing to do, the Taliban had to be taken down, at long last, and that was where OBL was hiding. Good job!
Now I do recall Chirac being interviewed by Time Magazine over two pages and reading a very cogent argument whereby going to Iraq was the wrong thing to do. The US would not be welcome there, OBL wasn't there and neither were any 9/11 mastermind, and it would be a dirty and costly war. Where did Chirac go wrong exactly?
Concerning your main other points, I think you are right, France should shut up on racism and multiculturalism, its model doesn't work very well, but I do remember racial riots in the US as well (LA?). On democracy however I'm not sure who should be giving lessons. Neither France nor the US has any great record of bringing real and lasting democracy anywhere. Certainly for France it didn't work very well in western Africa.
As for respect and arrogance it's mostly a matter of how things are reported in the media. The French media often report US initiatives with skepticism and incomprehension, the US media likes to report French ones in the worst possible light as often as possible lately.
The situation is not healthy. The comprehension is much better between the UK and the US because Americans can tune to the BBC and understand, while Britons can and do watch CNN easily.
I'm a French-speaking Aussie BTW, and I've lived and worked in the USA.
In 1989 ObjC was a much better OO language than C++ but the world has moved on since C++ was simply an Object Oriented Language.
Personally I think C++ in 1989 was far less defensible than it is now. Compare the NIH Class library of 1990 or so with the 1998 ISO C++ Library, in particular the STL, or far more interestingly the BOOST libraries.
Thanks for the answers, I have other questions/remarks :
- How does the service/song expire ? Presumably you can upload the songs to a WMV player. Do those keep playing forever as long as you don't sync ? In other words can you keep a selection somewhere that you want to play for a long time without necessarily purchasing the songs and endlessly paying for the subscription?
- If you subscribe to your favourite cable channel, you can still tape the shows and keep them forever if you so wish.
- I like to purchase whole albums. Itunes is great for that. As yahoo requires a CC# for me to sample the system, I can't find out if albums are available, is that the case ?
If this were 1990, the title would read "neural network predicts movie success" and the discussion would be about the impending success of strong AI.
Reading TFA, it's impossible to know whether this study has any value without seing a proper article, as submited to a reputable stats journal.
First of all this sounds like simple statistical classification with pretty obvious variables. However making classification work is not always trivial.
Methodology is the key here. The sample of 800 movies is rather small, and the details on the chosen explanatory variables is sketchy. With enough variables, even meaningless ones, one can explain anything on a training sample. However with proper classification techniques, using for example jacknife/resubstitution/cross-validation one can find out if the classification model has any actual predictive values.
As someone said "anybody can predict the past", and someone else "prediction is rather difficult, especially about the future".
Slashdot Mirror
Apple takes the view that every separate boxed sale of OS/X is an upgrade because to run it you need an Apple machine, and that machine necessarily came loaded with some prior version of MacOS.
However these boxes contain a full version of Mac OS/X, not an upgrade kit. You can install it on a freshly wiped Apple box or in an emulator, it doesn't request a copy of the previous version's media.
Your first sentence in this light makes no sense. At any rate if you buy a boxed version of OS/X from the Apple store, it doesn't say "Upgrade" it says "Buy OS/X, single or family license".
Can you give an example of a good use of TPM from the consumer point of view? I thought not.
TPM is a fundamentally restrictive technology. It's there to prevent, not to help. Consumers are universally better off without it rather than with it. It is also in my opinon a fallacy that content providers require it to make money. As an example, international DVD sales really took off with region coding being defeated.
Can Apple buy Dell now, sell off its assets and redistribute the wealth to the shareholders ?
You are correct that there is no need to agree to the GPL to use software licensed under it, however in general to use software you need a license, and the GPL grants you unilaterally all rights to use the software.
In other words, you don't need to agree to the GPL because there is no counterpart agreement on your part as far as usage is concerned, but you need to be aware that the software that you are using is in fact Free to use.
Most Windows installers (even Free ones!) that F/OSS software use have a mandatory page for the licence. Putting the GPL there is not the best way, but it's OK. It should be only a page saying "this software is licensed under the GPL, therefore you can use it to your heart content. Have fun" sort of deal.
Actually (and I work in this area as a researcher), computers suck at face recognition far far less than humans.
You may think that you are good at recognising people's face but the fact is you probably can only recognise 100-1000 faces all told, of people you know rather well, and doing it in a crowd is very hard. Recognising the face of someone you met once 10 years ago is for most people almost impossible.
The US driving licence face recognition system works on litteraly millions of people, and copes very well with eyeglasses, facial hair & different hairstyle and aging. It is used to prevent people from acquiring more than one driving license in different states.
1- the license
2- the speed (OS/X's kernel is slow as molasses), not to mention how incredibly memory hungry OS/X is. I have a Linux 64-bit desktop with 512 MB of Ram that hardly ever swaps even in heavy use. MacOS thrashes about swapping with 768 MB of RAM after about 2h of work. When I have even a modest compile job I leave it thrashing around overnight rather than face the waiting
3- 64-bit. Despite what you hear, even Tiger is not 64-bit. Of course this doesn't apply to the new MacBook Pro or iMac, which sport a dual 32-bit CPU, but EMT64 core duo CPUs are in the works, I hear. I bet OS/X 10.5 still won't be 64-bit throughout.
4- the GUI. Honesly I prefer gnome or even KDE these days to slow-as-molasses, full of silly eye-candy you can't turn off, no way to get focus-follow-mouse Aqua or whatever the graphics layer on OS/X is called.
5- Poor compatibility. At my workplace and in general in my field, very few people run OS/X for the above reasons, and almost everyone runs Linux.
6- I'm sick of recompiling all the free apps under the sun to get the software I want. I'm sick of waiting for OpenOffice 2.0 to be supported on OS/X, I'm sick of having to rely on closed-source, third party shareware to do things that are taken for granted under Linux. I hate the OS/X version of most of the CLI software there is. Apple's Python is unusable on the command line, there is no command line editing or interactive history !
7- The proprietary apps aren't worth it. MS-Office is superceded by OOo. The Gimp is enough for me over photoshop (I must have tried it once) and I don't even need Illustrator anymore thanks to Inkscape. Mac have no games, BTW (unless you are happy to pay full initial price for 3-year old games).
I have a lowly iBook 12", bought on the recollection of what NeXTStep was like about 15 years ago. I was amazed that to see that most things were there but that Apple had seen fit to make the N/S interface builder format incompatible and unreadable (and not for lack of trying). I could recompile none of my old apps. As a result I completely gave up on XCode.
I'm switching back to Linux on that notebook as soon as suspend-to-ram and wireless both work (stupid proprietary controller).
OS/X is way way way overrated. Sure it's better than windows, but that's about it, and not even on all points -- for example why can't I do reliable suspend-to-disk on my laptop ?
This is really simple.
/bin/sh, and have for decades, then clearly Linux is a good choice for the desktop because it's way nicer than any of the proprietary Unices for that. On HPUX, Irix and whatnot you get X and mwm, that's it. On Solaris you might get a bit of Gnome if you are lucky. The exception is OS/X, of course, but that won't run on cheap hardware they have lying around everywhere. Notice that in this case Windows is a very very poor choice.
If you know your way around
On the other hand if you have experience with the sheer unshakable predictability, usefullness and stability of Solaris in the server room, then compared to that good ole Linux is a bit of a toy to be honest (although slowly improving).
So JPL is in the situation where they like Linux on the desktop but not in the server room, and it makes perfect sense.
FORTRAN is still an evolving language, look for FORTRAN 90, 95, and 2003.
Many graduate students in applied mathematics and statistics write their new code in these languages, especially since one can find a excellent F95 compiler, free for non-commercial uses on Linux.
By non-democratic I meant that the decision to build the first, and then many nuclear plants in France was never the subject of debate in parliament, let alone amongst the people. It was made between ministers, some EdF executives and also a special-purpose public body akin to the DoE, called the CEA (atomic energy commission). The CEA has a tradition of secrecy as they were involved in (nuclear) weapons manufacturing as well. The programme was always presented as an absolute necessity and a "fait accompli".
EdF/government manufactured consent through running paid adds on prime-time TV in the 1970s, touting alternative energy sources like wind and solar as well as nuclear. However the wind and solar powerplants were tiny and for show, while the nuclear programme was huge and real.
If you talk to people who were watching French TV in the seventies they probably remember them, they were very effective adds. They appealed to the "patriotic" sense of duty and ingenuity in a subtle way. The slogan was something like "in France we don't have much oil but we have ideas". As a result there were effectively few protests.
The problem I see with this approach is that it made it very hard to have a reasonable and informed point of view. Some journalists tried quite hard to counterbalance the gov/EdF/CEA point of view on biased scientific and emotional terms. Some papers printed nice charts that showed which areas would become lethally irradiated should such and such powerplant blow up. Others wrote book-length rants about how the technology was dangerous, unpleasant and generally a disgrace.
Almost no one thought of organising any kind of reasonable and informed debate on this issue. It was just hysteria that got nowhere.
Now the situation is much better. It's actually quite easy to go and visit a nuclear plant (i've done it several times) and to get relatively accurate information about what it being done with the waste. Most of the dangerous waste is reprocessed and the rest is stored. In fact France reprocesses the waste of several countries (and Greenpeace always protests when a ship comes in full of nuclear waste).
Anyway, this is a different debate.
France has a huge leading technocracy. The decision to go mostly nuclear for electricity distribution was facilitated by there existing only one state-owned electricity producer (Electricité de France a.k.a. EDF) and by a mostly non-democratic decision-making process.
On the other hand there were never any huge, organized anti-nuclear protests in France, which was hit very hard by the first oil price hike in 1973. Anti-nuclear protests in recent years have been confined to sites where nuclear wastes were due to be dumped. However protesters were able to convince government to stop building new plants for many years, probably as there were enough capacity anyway.
Specifically, nearly all of OS/X applications, the kernel and the GUI are all 32-bit on all Macs. There are only two versions of OS/X, one for servers, and one for laptops/desktops. Since there are G4 versions of either, necessarily all software has to be 32-bit only. This is why when Tiger (10.4) came out all the benchmarks in magazine showed no speed difference between "32-bit" 10.3 and "64-bit" 10.4.
However, OS/X 10.4.x is able to run 64-bit apps that can access more than 4GB of memory, but those don't have access to the GUI directly. Console apps only! If you need a GUI in front of that 64-bit app, prepare to handle the joys of ICP (inter-process communications).
Well, maybe with enough muscle mass one would be able to use time dilation effect by pushing real hard in the direction of the switch and cut back the time needed to 50 years?
I basically agree with what you wrote, indeed Apple notebooks especially PBs simply were the best laptop until about 12-18 months ago. In particular PC laptops had terrible battery life (60-90 minutes, typical...) and were prone to actually burning your lap...
However things have changed very rapidly in the PC world. Now PC laptops have better features (larger screens, DVD burner standard, PCMCIA slot, wireless, BT, excellent build quality (IBM, Fujitsu, etc), faster CPU, more RAM, more disk, 64-bit, now dual-core etc etc etc) AND they have better battery life AND they are cheaper. I'm not listing OS/X because PC laptops by and large have Linux and in my mind the two are equivalent (for what I'm doing, may not be true for all).
It will be good to see what Apple is going to offer soon.
Yes, but Apple is screwed. Apple notebooks sales used to drive the company, now that everyone is waiting for Intel, everyone also acknowledges that the powerbook/iBook lines are way underpowered. Yet people with powerbooks by and large need the pro apps, such as the Adobe line.
These won't be coming in Intel form for months. I would expect a last speed bump on the PBs to bring them close to 2GHz, and an Intel iBook. They'll look great and have better raw performance than the PBs for a while, but they won't run the 3rd party pro apps satisfactorily, so they won't cannibalise PB sales to come. PB owners will wait a few more months for an even more powerful Intel chip (64-bit, dual core), and the pro apps to come with them, such as Adobe CS3.
iBooks owner will have to weather the brunt of the conversion to Intel. For a long while they will have precious few apps to run, except very slowly through Rosetta. It will be a nice laptop with few apps (unless they dual-boot to Windows!) I don't think it can be very successful just now. In 6-12 months it will be a different story.
Apple has money in the bank, they can weather the next hard few months. This upcoming iBooks is going to be more of a tech preview laptop than a real product. Expect fanzine to be full of impressive synthetic benchmarks, whipping up a huge frenzy for when the Intel PB line is finally released.
Having a real Intel product available also means Apple can put pressure on 3rd party software houses to release Intel version of their software, as there will be demand, and more readily available development platforms.
Wait, you are saying that Apple notebooks are better designed because there is more margin than for PC notebooks? Yet the iBooks are among the cheapest notebooks around. By your argument they should be using very cheap materials (to secure larger margins) and so extremely flaky and flimsy, yet they are not.
Apple has had a long field day with their notebooks because they were well designed to start with, they have not fiddled with their design as much as PC manufacturers, and they are able to charge for extras.
In contrast there is very little differential in the PC world among notebooks (they all look and feel the same, pretty much), so manufacturers have had to compete on features. This is why it is impossible to get away with shipping only a combo DVD reader/CD burner in a current PC notebook, yet this is what Apple is still doing with all 12" iBooks and some 12" PowerBooks.
Will things change when moving to Intel? I think Apple will be feeling more competitive pressure now that their main performance indicator (the CPU make and speed) will be directly comparable with PCs. It will be interesting to see.
Hilbert was working on GR as well, but he didn't have the physical insights Einstein had.
Why, using Bicycle Repairman, of course !
I thought, but I must be wrong, that cash was legal tender everywhere and no one could refuse cash for payment for any reason.
Just in case one has been living under a rock for the past few years, CSS, the DVD encrypting scheme has been broken a long time ago.
On a general-purpose PC, including Macs, BeBox and whatnot, DVD region coding can be bypassed any old way, for example using VLC. RPC1 or 2 don't serve any purpose anymore.
> whether we're right or wrong
Give actual examples please. I don't recall Canada, Germany or France voicing any critic when the US invaded Afghanistan. That was the right thing to do, the Taliban had to be taken down, at long last, and that was where OBL was hiding. Good job!
Now I do recall Chirac being interviewed by Time Magazine over two pages and reading a very cogent argument whereby going to Iraq was the wrong thing to do. The US would not be welcome there, OBL wasn't there and neither were any 9/11 mastermind, and it would be a dirty and costly war. Where did Chirac go wrong exactly?
Concerning your main other points, I think you are right, France should shut up on racism and multiculturalism, its model doesn't work very well, but I do remember racial riots in the US as well (LA?). On democracy however I'm not sure who should be giving lessons. Neither France nor the US has any great record of bringing real and lasting democracy anywhere. Certainly for France it didn't work very well in western Africa.
As for respect and arrogance it's mostly a matter of how things are reported in the media. The French media often report US initiatives with skepticism and incomprehension, the US media likes to report French ones in the worst possible light as often as possible lately.
The situation is not healthy. The comprehension is much better between the UK and the US because Americans can tune to the BBC and understand, while Britons can and do watch CNN easily.
I'm a French-speaking Aussie BTW, and I've lived and worked in the USA.
Some of us like having zero-overhead generic container classes. It comes in rather handy in practice.
There is no such thing as a perfect language.
In 1989 ObjC was a much better OO language than C++ but the world has moved on since C++ was simply an Object Oriented Language.
Personally I think C++ in 1989 was far less defensible than it is now. Compare the NIH Class library of 1990 or so with the 1998 ISO C++ Library, in particular the STL, or far more interestingly the BOOST libraries.
Thanks for the answers, I have other questions/remarks :
- How does the service/song expire ? Presumably you can upload the songs to a WMV player. Do those keep playing forever as long as you don't sync ? In other words can you keep a selection somewhere that you want to play for a long time without necessarily purchasing the songs and endlessly paying for the subscription?
- If you subscribe to your favourite cable channel, you can still tape the shows and keep them forever if you so wish.
- I like to purchase whole albums. Itunes is great for that. As yahoo requires a CC# for me to sample the system, I can't find out if albums are available, is that the case ?
- thanks for the tip.
Best.
If this were 1990, the title would read "neural network predicts movie success" and the discussion would be about the impending success of strong AI.
Reading TFA, it's impossible to know whether this study has any value without seing a proper article, as submited to a reputable stats journal.
First of all this sounds like simple statistical classification with pretty obvious variables. However making classification work is not always trivial.
Methodology is the key here. The sample of 800 movies is rather small, and the details on the chosen explanatory variables is sketchy. With enough variables, even meaningless ones, one can explain anything on a training sample. However with proper classification techniques, using for example jacknife/resubstitution/cross-validation one can find out if the classification model has any actual predictive values.
As someone said "anybody can predict the past", and someone else "prediction is rather difficult, especially about the future".
In this case, however, some of the priests were both gay and paedophile (i.e. they raped male children).