but can it really work ? even assuming perfect voice recognition and perfect translation abilities (which we don't have), Human interpreters often have to wait till the end of a sentence to be able to translate a word, and real-time word-for-word is totally meaningless, if not impossible.
As an example, consider german, which can put verbs very very far at the end of a sentence in a stack-like manner (a bit like postfixed LISP).
To put it mildly this is a very challenging project.
According to the article you link to, a grand total of 5 Americans have sought asylum this year. This is hardly a deluge, and I don't particularly believe the Sun when it says "they freely admit it's for the free stuff".
If the parent came up with a novel way of folding the paper to produce a crane, it would most definitely be copyritable, or at least the explanations to fold the paper would be.
Do you run the 64-bit driver NVidia driver ? the 8xxx series is awful in 64-bit mode, on my machine it crashes at least once every day, doing nothing special.
I know it's a driver issue, because the 9xxx beta series is much improved, zero crash in weeks.
The parent poster is not acting irrationally, he is reasoning that the CPU market needs competition and so he supports the second fiddle. By doing so he contiously loses some FPM today, but he helps maintain the competition which ensures that his future FPM will eventually become lower at some point.
If everyone jumped ship on benchmarks alone, Intel would have died last year, we would not have Core Duo today and we would be stuck with A64 3800 forever.
For some processes you really need oil or natural gas. For example to process canola to make biodiesel fuel you need to spend some oil. It's hard to power that tractor on solar or nuclear energy. To make the fertilizer to grow that canola, the primary ingredient is natural gas.
The take-home message is while it is possible to substitute some form of energy for others, in general it is :
1- less convenient 2- more expensive
than oil.
The #1 worry is that right now a lot of the conflicts in the world revolve around oil while we are nowhere near scarcity yet. It is not that we will completely run out of oil, it is rather than the comfortable times are over.
If that land belonged to their grandfather, like you yourself are writing, then the people living in camps may indeed have a legitimate claim to that land. That they don't live on it has nothing to do with the question.
My great-great-great grandfathers lost the Confederacy, does that mean I can grab a gun and start shooting Yanks?
If you live in the South then you must have heard these people saying "the South shall rise again" perhaps ? Notice that you probably haven't been living in a camp and perhaps don't feel as disposessed, helpless and angry as many Palestinians.
I guess that Palestinians feel that they were given a raw deal by the West and the only way they'll get their claim back is through fighting. Is that so hard to understand ? (I didn't say condone).
It depends on the field. In the natural sciences, progress is often very quick and incremental, also publication delays are short, and the field is not of infinite scope like maths. Finally, researchers all use the same tools and similar approaches. This means competition is usually fierce because your neighbour's work can undermine your own.
In other fields, like computer science, maths and physics, there is basically room for everyone and new tools are developed every day. It is also quite rare that two teams work on the exact same subjet and develop the exact same method. People working in similar fields will benefit from talking to each other, not the other way around. I find there is more free sharing of resources and discoveries in these area. In these area, as soon as you show something is possible, it's almost as good as solved, so publishing incomplete solution don't deter other groups much.
Of course these are generalities, and there are exceptions either way.
Nvidia are definitely distributing their kernel extentions via their web site. You are playing with words here.
You are right that Nvidia gets away with it because essentially Linus himself has no big qualms about it. Personally I think this is a shame. Because of the laissez-faire attitude we have very poor quality free drivers, as no one has the incentive to reverse-engineer the hardware, and very poor quality proprietary hardware. Until the very latest beta drivers (9xxx series) I think Nvidia drivers really really sucked, as in lock up every 2 days on my 64-bit AMD machine (the 32-bit drivers have been sucking a little less for longer). Now they are barely adequate.
However the publishers may argue, amongst other things, that by doing their book indexing, Google are depriving the copyright owners of some control over their work without their consent. From what I've seen the scanning and OCR of non-authorized material is complete and thorough. There is also the point that you raise, that basically a digital, complete, unauthorised copy of all the disputed works is currently present on Google's servers. The fact that it is not available to the general public is mostly irrelevant. Also Google is using this data for profit (adds) without compensation to the copyright holders. This goes very much against both the spirit and the letter of copyright.
The copyright holders may also argue that web search is quite different. Web pages reachable by spiders who respect robot.txt files are inherently public and meant to be indexed. Book pages are not, to my knowledge, in any way comparable.
The fact that the Google indexing is opt out whereas all the other indexing services are opt-in (amazon, B&N, etc) makes it very different from the other offerings.
The fact that Google is not willing to switch to opt-in (in effect this would very much reduce the value of their offering vs. all of the competition) makes it clear they want to strongarm the publishers into the Google business model.
In my opinion Google is a little overconfident in this matter. At any rate it will be interesting to see what happens.
Your longest path algorithm won't work, you may have a cycle in your graph and your remaining path will never reach the endpoint. The algorithm will not terminate.
Truth is, France hasn't really been an expansionist country for most of its existence, save Louis the XIVth and Napoleonics ventures.
Well, tell that to the countries of Maghreb, West Africa, ex-Indochina, etc etc. French colonies were quite extensive and have nothing to do with Napoleon or Louix XIV. Decolonisation was a painful process in France and is still generating discontent to this day.
Google has a number of O/Ses they could reuse, their own linux distribution for instance. The hard slog has already been done by all these volunteers and Linux works just about as well as Windows now, driverwise, with fewer and fewer exceptions as time goes by.
10 years ago, people may have needed Qt, since the state of Linux/UNIX toolkits was poor. But I think these days, nobody "needs" Qt, since there are plenty of excellent alternatives that cost you nothing under less restrictive licenses.
Please name them.
QT really is quite excellent if you want to develop for all three of Windows, Macs and Linux/Unix. I know of no toolkit that can match it.
There are portable alternatives, and I use some of them (wxWidgets, FLTK) because of the QT license, but they are not as good.
At the same time there were several socket-7. The one for Pentium 75-133 wasn't the same as the one for Pentium 166-200 due to voltage differences. I'm not familiar with socket A, but all-in-all socket 754 has more reasonably long-lived than people give it credit for.
This is an interesting commment, except that Alcatel, like any large telco would have been dead long ago if they hadn't done or sponsored a modicum of basic research, and they have, see this for example.
Meanwhile, at Bell Labs, things have been business-focused for a very long time. Remember that Thompson, Richie et al. couldn't get funding to make a new O/S, they had to pretend they were writing a text processor instead.
The first version of @acronym{UNIX} was developed on a PDP-7 which was sitting around Bell Labs. In 1971 the developers wanted to get a PDP-11 for further work on the operating system. In order to justify the cost for this system, they proposed that they would implement a document formatting system for the AT&T patents division. This first formatting program was a reimplementation of McIllroy's roff, written by J. F. Ossanna.
I'm not sure how the survey was conducted, they don't say. A proper survey is completely random, you ask random people if they have an Apple laptop, what name, year etc and if it ever needed repairs. Survey by people who volunteer their experience is near worthless, as you said. You can't derive any sort of meaningful information from them.
I tend to believe the reliability of Apple notebooks is not great though, based on my own experience. I've own a powerbook, an ibook and now a MBP. All three were in heavy use. The powerbook failed in about 30months, the ibook in 25 months. The MBP is still working so far but with lots of little issues. This fits your categories like a glove BTW.
I'm not sure about DELL, but Apple, who has earned a reputation for good engineering in general, posts about a 15% failure rate on its laptops in the first year, according to that survey. On some models, it goes as high as 4!%, and 73% over two years.
Slashdot Mirror
Because without firefox and its 20% market share or so, there wouldn't be a IE7 team at Microsoft. No need to.
Core duo is only 32-bit.
Core duo 2 is 64-bit.
Segmentation even under the simplest of conditions (piecewise constant images) is an NP-difficult optimisation problem.
People have to use hacks (more politely meta-heuristics) because the problem is unsolvable otherwise.
Also titled "On the antiquity of microbes".
On the same level :
Me/We Muhammed Ali.
but can it really work ? even assuming perfect voice recognition and perfect translation abilities (which we don't have), Human interpreters often have to wait till the end of a sentence to be able to translate a word, and real-time word-for-word is totally meaningless, if not impossible.
As an example, consider german, which can put verbs very very far at the end of a sentence in a stack-like manner (a bit like postfixed LISP).
To put it mildly this is a very challenging project.
According to the article you link to, a grand total of 5 Americans have sought asylum this year. This is hardly a deluge, and I don't particularly believe the Sun when it says "they freely admit it's for the free stuff".
I think you missed the "in a laptop" bit.
If the parent came up with a novel way of folding the paper to produce a crane, it would most definitely be copyritable, or at least the explanations to fold the paper would be.
Do you run the 64-bit driver NVidia driver ? the 8xxx series is awful in 64-bit mode, on my machine it crashes at least once every day, doing nothing special.
I know it's a driver issue, because the 9xxx beta series is much improved, zero crash in weeks.
The parent poster is not acting irrationally, he is reasoning that the CPU market needs competition and so he supports the second fiddle. By doing so he contiously loses some FPM today, but he helps maintain the competition which ensures that his future FPM will eventually become lower at some point.
If everyone jumped ship on benchmarks alone, Intel would have died last year, we would not have Core Duo today and we would be stuck with A64 3800 forever.
For some processes you really need oil or natural gas. For example to process canola to make biodiesel fuel you need to spend some oil. It's hard to power that tractor on solar or nuclear energy. To make the fertilizer to grow that canola, the primary ingredient is natural gas.
The take-home message is while it is possible to substitute some form of energy for others, in general it is :
1- less convenient
2- more expensive
than oil.
The #1 worry is that right now a lot of the conflicts in the world revolve around oil while we are nowhere near scarcity yet. It is not that we will completely run out of oil, it is rather than the comfortable times are over.
If you live in the South then you must have heard these people saying "the South shall rise again" perhaps ? Notice that you probably haven't been living in a camp and perhaps don't feel as disposessed, helpless and angry as many Palestinians.
I guess that Palestinians feel that they were given a raw deal by the West and the only way they'll get their claim back is through fighting. Is that so hard to understand ? (I didn't say condone).
It depends on the field. In the natural sciences, progress is often very quick and incremental, also publication delays are short, and the field is not of infinite scope like maths. Finally, researchers all use the same tools and similar approaches. This means competition is usually fierce because your neighbour's work can undermine your own.
In other fields, like computer science, maths and physics, there is basically room for everyone and new tools are developed every day. It is also quite rare that two teams work on the exact same subjet and develop the exact same method. People working in similar fields will benefit from talking to each other, not the other way around. I find there is more free sharing of resources and discoveries in these area. In these area, as soon as you show something is possible, it's almost as good as solved, so publishing incomplete solution don't deter other groups much.
Of course these are generalities, and there are exceptions either way.
Nvidia are definitely distributing their kernel extentions via their web site. You are playing with words here.
You are right that Nvidia gets away with it because essentially Linus himself has no big qualms about it. Personally I think this is a shame. Because of the laissez-faire attitude we have very poor quality free drivers, as no one has the incentive to reverse-engineer the hardware, and very poor quality proprietary hardware. Until the very latest beta drivers (9xxx series) I think Nvidia drivers really really sucked, as in lock up every 2 days on my 64-bit AMD machine (the 32-bit drivers have been sucking a little less for longer). Now they are barely adequate.
However the publishers may argue, amongst other things, that by doing their book indexing, Google are depriving the copyright owners of some control over their work without their consent. From what I've seen the scanning and OCR of non-authorized material is complete and thorough. There is also the point that you raise, that basically a digital, complete, unauthorised copy of all the disputed works is currently present on Google's servers. The fact that it is not available to the general public is mostly irrelevant. Also Google is using this data for profit (adds) without compensation to the copyright holders. This goes very much against both the spirit and the letter of copyright.
The copyright holders may also argue that web search is quite different. Web pages reachable by spiders who respect robot.txt files are inherently public and meant to be indexed. Book pages are not, to my knowledge, in any way comparable.
The fact that the Google indexing is opt out whereas all the other indexing services are opt-in (amazon, B&N, etc) makes it very different from the other offerings.
The fact that Google is not willing to switch to opt-in (in effect this would very much reduce the value of their offering vs. all of the competition) makes it clear they want to strongarm the publishers into the Google business model.
In my opinion Google is a little overconfident in this matter. At any rate it will be interesting to see what happens.
Your longest path algorithm won't work, you may have a cycle in your graph and your remaining path will never reach the endpoint. The algorithm will not terminate.
Well, tell that to the countries of Maghreb, West Africa, ex-Indochina, etc etc. French colonies were quite extensive and have nothing to do with Napoleon or Louix XIV. Decolonisation was a painful process in France and is still generating discontent to this day.
Yes, it has changed, the SNCF site is now quite browser-agnostic.
Google has a number of O/Ses they could reuse, their own linux distribution for instance. The hard slog has already been done by all these volunteers and Linux works just about as well as Windows now, driverwise, with fewer and fewer exceptions as time goes by.
Please name them.
QT really is quite excellent if you want to develop for all three of Windows, Macs and Linux/Unix. I know of no toolkit that can match it.
There are portable alternatives, and I use some of them (wxWidgets, FLTK) because of the QT license, but they are not as good.
At the same time there were several socket-7. The one for Pentium 75-133 wasn't the same as the one for Pentium 166-200 due to voltage differences. I'm not familiar with socket A, but all-in-all socket 754 has more reasonably long-lived than people give it credit for.
Meanwhile, at Bell Labs, things have been business-focused for a very long time. Remember that Thompson, Richie et al. couldn't get funding to make a new O/S, they had to pretend they were writing a text processor instead.
Not really, all the AMD-based laptops use socket 754, and some 754 desktop motherboards accept notebook CPU.
Hi,
I'm not sure how the survey was conducted, they don't say. A proper survey is completely random, you ask random people if they have an Apple laptop, what name, year etc and if it ever needed repairs. Survey by people who volunteer their experience is near worthless, as you said. You can't derive any sort of meaningful information from them.
I tend to believe the reliability of Apple notebooks is not great though, based on my own experience. I've own a powerbook, an ibook and now a MBP. All three were in heavy use. The powerbook failed in about 30months, the ibook in 25 months. The MBP is still working so far but with lots of little issues. This fits your categories like a glove BTW.
I'm not sure about DELL, but Apple, who has earned a reputation for good engineering in general, posts about a 15% failure rate on its laptops in the first year, according to that survey. On some models, it goes as high as 4!%, and 73% over two years.
So 15% is within normal by that standard.