The best challenge for them would be to fix the many bugs of Firefox, many of which are in the very same XUL engine the DE should be based upon. Better 10 mediocre project or a great one?
They would do it simply because they can. It's the Microsoft idea of "since other do it, let's do it". I don't think it's a very good idea, actually. They should stick with what they know best.
My opinion is not shaped depending on the adoption of Macs of ANY university. As far as I am concerned moving to Macs to me is not an improvement at all. It's another proprietary platform, in fact it locks the university in terms of hardware too.
To me the best would be to adopt a university customized version of Linux. This is my opinion. A University should be totally able to control its own computing environment.
My original point was different. An hardware switch (they will be still using Windows with BootCamp), from a small university is no breakthrough. However if a majopr university (think BIG), would make the switch, then that would be something to consider and to talk about. The same would be true for a major university strongly supporting and deploying FOSS (which hasn't happened either).
This is not about opinions. It's about visibility.
Nothing to see here. Having an Apple supercomputer designed in a university, doesn't mean that university is switching. If that was the case, there would be way more US universities switching to Linux, since most of supercomputing clusters actually runs on Linux.
So I get back to sleep.
This is a very serious problem of Ubuntu that is overlooked by the developers. Problems with specific hardware like those in the TFA, can be common especially these days with so many different combinations of monitors and video cards. I'd like to see some sort of "safe mode" that kicks in when there are problems, and a GUI to allow proper reconfiguration.
If you expect a windows user to manually edit Xorg.conf, you're wrong.
You blame AOL/Yahoo/Hotmail/Gmail for your security. At the end you seem to be really worried about the VBS. As far as I know if you use Macs or Linux or any other non-Microsoft products, you can be sure to be safe from those scripts attacks. i know it's easier to blame the email providers instead of Microsoft for its poor security, which allows scripts to be executed system-wide.
So, again, looks fror the cause of your security concerns, not the consequences!
Supporting the hardware (with certified drivers) can be done without shipping the OS. If done properly you give both the choice of the distro you want AND reliable drivers.
As an alternative provide external support for some OS, either being Novell for SLED, Redhat for RHEL, or Canonical for Ubuntu. You buy the PC with the OS preinstalled and the support is provided not by Dell, but by the OS manufacturer. Something like this is already possible within Dell. They just need to expand the choice at least to business PCs.
Supporting a full-distro is demanding for a non-FOSS committed company like Dell. It's like asking MS to be committed to FOSS.
Your suggestion is the one that would make most sense in an ideal situation. In practice though, it would fail. People are attached to their distro of choice. Some companies also heavily rely on some distributions because some software they use would be certified only on those distribution.
for Dell its own distro would be a big commitment but in development and support. And then what distro should they use for theirs? A Debian based? That would alienate RedHat and Novell users. Novell? There is already too much to say about the Novell-Microsoft deal. As an example think to Oracle and their new distro based on RedHat. So far it hasn't been very successful. A distro like this would need to be certified for third party software (from IBM, Oracle, etc). I don't think Dell would go that way just to please the linux request from users.
I would quite happy if they could ship a system with No OS, at the same price or lower than with a Windows pre-loaded system, with full hardware support (read drivers). And maybe since you are at it, throw in a free CD of Ubuntu. That way I would install my distro of choice, and install the drivers (proprietary too if necessary) to make it work.
Simple because everything would be there installed and working (drivers mostly). I would have THEM to mess around with xorg.conf, have the correct ACPI support, sound and wireless working.
I can sure do it. But if they do it, well the better. In fact that's what they do with windows.
There are several forms to describe the fibonacci sequence. I find this to be the most intuitive. But then again, people add complexity to things that are generally simple.
That's not a pattern. A real one is:
LSLLSLSLLSLLSLSLLSLSL
the Fibonacci sequence. Although there is no periodicity, a local subset of tiles randomly repeat themselves (it's called self similarity). For example: LLS, LS, SLLS, etc.
The best way to know this is to take the Fourier transform and see the discrete number of component (typical of a geometry with elements that repeats themselves).
You can use brute force to form pseudo quasiperiodic tilings. However they are not really quasiperiodic, they are "approximants". So if are patient enough you can get to a point where you have a large pattern, but not necessarily being quasiperiodic. So to a degree of trial and error the Islamic artists must have developed a degree of knowledge which you seem to underestimate. point in case: the Fibonacci sequence and the golden mean. Simple mathematical rules at the base, not impossible to grasp with the arabic knowledge of math of the time. The 2D mapping may just have been their "next step". There have been long speculation of the fact that Islamic art was for long considered just a coincidence. The article here present the prove that this may indeed be not true.
Disclaimer: I (not my wife) did my PhD in quasicrystal tiling.
People say that it's a coincidence. What Lu and Steinhardt are
demonstrating is deep knowledge of advanced math. The best example of
quasiperiodic tiling (as they are called), is the Fibonacci sequence.
To build it you need to differnet tiles (say a long segment "L" and a
short segment "S") and two combining rules:
1. at every S you change it with a L
2. at every L you change it with LS
so you build the different generations of the sequence as follow:
S
LS
LSL
LSLLS
LSLLSLSL
LSLLSLSLLSLLS
LSLLSLSLLSLLSLSLLSLSL
etc...
You can go at infinity with this. You won't find periodicity or a
pattern that repeat itself. Now to the point: does this means that you
take the two segments and you put them together randomly you get the F.
sequence? No, by any chance. The rules are simple (and the Fibonacci
sequence is old (~1200), so I would not be surprised if the Islamic
mathematicians were aware of it, so they "ported" it in 2D (the Penrose
tiling is the 2D version of the F. sequence).
By the way the story goes even back in time further: the ratio between
the number of L and S for a significantly large sequence, is tau, the
golden mean (again the same is true for the Penrose tiling). The golden
mean was a key number (sqrt5+1)/2~1.6... in the greek world, where it
was used as a proportion standard to build building and temples. It's
also a key element in fractal growth, in key dimensions of our body,
etc.
So the Islamic artists (scientists?) of the time were a bit like
today's scientists. they gathered previous studies and assembled
together using some new insights.
Why don't give them credits for it, instead of stupidly saying: "well
they just got lucky?".
You take the tiles and you randomly try to tile your space, and I garantee you (actually you can demonstrate it) that you can't cover all the space, you will have blanks. There are specific "combining rules" to be followed to be able to form those nice and ordered patterns.
Having a pretty geometry has nothing to do with it.
See the answer below. Our addiction to oil is so huge that even using all the fertile land to produce ethanol will make up to only about 40% of our needs, without having any land left for food crops. THe problem is not the land itself. It's the water supply. After oil we are addicted to water, which has been used most of the time not in th optimal way. Chemical contamination, overuse for electrical production, reduced the amount usable, for agriculture.
In addition, let's not forget the desertification of the North American continent, and of Europe. Once the land becomes desert, there is usually no way back.
For this reason I strongly opposed ethanol and the huge investments regarding its use instead of crop for food. Let's spend research money to improve solar and nuclear. With a solid alternative source of energy, we could extract potable water from the sea. It's a dream, I know, but poper use of nuclear energy would allow that.
So I am in favor of real alternative sources of energy, which do not rely on current primary needs, and ethanol is not one of those.
Very true. Ethanol is viable only if produced from non-food-related products. Simple math, we don't have enough usable and fertile land to grow food crops AND fuel crops.
Much of Physics doesn't run on Macs, but Windows. A vast majority of Electron microscopes, probe microscopes runs on Windows too (Omicron, Philips, Veeco). It's funny that some of the best electron microscopy software actually runs on Macs, but only for PPC, so big centers (like the NCEM at the Lawrence Berkeley Lab) runs on old G4. Raman probes runs on Windows for the most part. A lot of custom built equipment uses National Instruments LabView, which runs bot Windows and Mac, but it's commolny installed in PCs, being usually cheaper. So unless you are doing advanced imaging world, most of the equipment runs on PCs.
So you are probably right about biology, linguistics and possibly others. But the physics and material science world is mainly a Windows world.
P.S. It's funny because sometimes you see in labs old Macs (from the late 80s and 90s) running very old equipment. Macs were the standard in Physics labs. At some point that changed....
Slashdot Mirror
The best challenge for them would be to fix the many bugs of Firefox, many of which are in the very same XUL engine the DE should be based upon. Better 10 mediocre project or a great one?
They would do it simply because they can. It's the Microsoft idea of "since other do it, let's do it". I don't think it's a very good idea, actually. They should stick with what they know best.
My opinion is not shaped depending on the adoption of Macs of ANY university. As far as I am concerned moving to Macs to me is not an improvement at all. It's another proprietary platform, in fact it locks the university in terms of hardware too. To me the best would be to adopt a university customized version of Linux. This is my opinion. A University should be totally able to control its own computing environment. My original point was different. An hardware switch (they will be still using Windows with BootCamp), from a small university is no breakthrough. However if a majopr university (think BIG), would make the switch, then that would be something to consider and to talk about. The same would be true for a major university strongly supporting and deploying FOSS (which hasn't happened either). This is not about opinions. It's about visibility.
Thanks for your note. I am really glad they are working on it.
You also need to add the $$$ for licenses of MS products...
Nothing to see here. Having an Apple supercomputer designed in a university, doesn't mean that university is switching. If that was the case, there would be way more US universities switching to Linux, since most of supercomputing clusters actually runs on Linux. So I get back to sleep.
This is a very serious problem of Ubuntu that is overlooked by the developers. Problems with specific hardware like those in the TFA, can be common especially these days with so many different combinations of monitors and video cards. I'd like to see some sort of "safe mode" that kicks in when there are problems, and a GUI to allow proper reconfiguration. If you expect a windows user to manually edit Xorg.conf, you're wrong.
Wake me up when a major US university does the switch...
You blame AOL/Yahoo/Hotmail/Gmail for your security. At the end you seem to be really worried about the VBS. As far as I know if you use Macs or Linux or any other non-Microsoft products, you can be sure to be safe from those scripts attacks. i know it's easier to blame the email providers instead of Microsoft for its poor security, which allows scripts to be executed system-wide. So, again, looks fror the cause of your security concerns, not the consequences!
Reason #1: Windows + Office market share: 95%; Linux + OpenOffice: ~1-2% Reason #2: Ubuntu + Openoffice is free. Bundling MS Office costs you big $.
Supporting the hardware (with certified drivers) can be done without shipping the OS. If done properly you give both the choice of the distro you want AND reliable drivers. As an alternative provide external support for some OS, either being Novell for SLED, Redhat for RHEL, or Canonical for Ubuntu. You buy the PC with the OS preinstalled and the support is provided not by Dell, but by the OS manufacturer. Something like this is already possible within Dell. They just need to expand the choice at least to business PCs. Supporting a full-distro is demanding for a non-FOSS committed company like Dell. It's like asking MS to be committed to FOSS.
Your suggestion is the one that would make most sense in an ideal situation. In practice though, it would fail. People are attached to their distro of choice. Some companies also heavily rely on some distributions because some software they use would be certified only on those distribution. for Dell its own distro would be a big commitment but in development and support. And then what distro should they use for theirs? A Debian based? That would alienate RedHat and Novell users. Novell? There is already too much to say about the Novell-Microsoft deal. As an example think to Oracle and their new distro based on RedHat. So far it hasn't been very successful. A distro like this would need to be certified for third party software (from IBM, Oracle, etc). I don't think Dell would go that way just to please the linux request from users. I would quite happy if they could ship a system with No OS, at the same price or lower than with a Windows pre-loaded system, with full hardware support (read drivers). And maybe since you are at it, throw in a free CD of Ubuntu. That way I would install my distro of choice, and install the drivers (proprietary too if necessary) to make it work.
Simple because everything would be there installed and working (drivers mostly). I would have THEM to mess around with xorg.conf, have the correct ACPI support, sound and wireless working. I can sure do it. But if they do it, well the better. In fact that's what they do with windows.
There are several forms to describe the fibonacci sequence. I find this to be the most intuitive. But then again, people add complexity to things that are generally simple.
That's not a pattern. A real one is: LSLLSLSLLSLLSLSLLSLSL the Fibonacci sequence. Although there is no periodicity, a local subset of tiles randomly repeat themselves (it's called self similarity). For example: LLS, LS, SLLS, etc. The best way to know this is to take the Fourier transform and see the discrete number of component (typical of a geometry with elements that repeats themselves).
You can use brute force to form pseudo quasiperiodic tilings. However they are not really quasiperiodic, they are "approximants". So if are patient enough you can get to a point where you have a large pattern, but not necessarily being quasiperiodic. So to a degree of trial and error the Islamic artists must have developed a degree of knowledge which you seem to underestimate. point in case: the Fibonacci sequence and the golden mean. Simple mathematical rules at the base, not impossible to grasp with the arabic knowledge of math of the time. The 2D mapping may just have been their "next step". There have been long speculation of the fact that Islamic art was for long considered just a coincidence. The article here present the prove that this may indeed be not true. Disclaimer: I (not my wife) did my PhD in quasicrystal tiling.
Luckily copyright laws do not apply to science.
People say that it's a coincidence. What Lu and Steinhardt are demonstrating is deep knowledge of advanced math. The best example of quasiperiodic tiling (as they are called), is the Fibonacci sequence. To build it you need to differnet tiles (say a long segment "L" and a short segment "S") and two combining rules:
1. at every S you change it with a L
2. at every L you change it with LS
so you build the different generations of the sequence as follow:
S
LS
LSL
LSLLS
LSLLSLSL
LSLLSLSLLSLLS
LSLLSLSLLSLLSLSLLSLSL
etc...
You can go at infinity with this. You won't find periodicity or a pattern that repeat itself. Now to the point: does this means that you take the two segments and you put them together randomly you get the F. sequence? No, by any chance. The rules are simple (and the Fibonacci sequence is old (~1200), so I would not be surprised if the Islamic mathematicians were aware of it, so they "ported" it in 2D (the Penrose tiling is the 2D version of the F. sequence).
By the way the story goes even back in time further: the ratio between the number of L and S for a significantly large sequence, is tau, the golden mean (again the same is true for the Penrose tiling). The golden mean was a key number (sqrt5+1)/2~1.6... in the greek world, where it was used as a proportion standard to build building and temples. It's also a key element in fractal growth, in key dimensions of our body, etc.
So the Islamic artists (scientists?) of the time were a bit like today's scientists. they gathered previous studies and assembled together using some new insights.
Why don't give them credits for it, instead of stupidly saying: "well they just got lucky?".
You take the tiles and you randomly try to tile your space, and I garantee you (actually you can demonstrate it) that you can't cover all the space, you will have blanks. There are specific "combining rules" to be followed to be able to form those nice and ordered patterns. Having a pretty geometry has nothing to do with it.
sudo apt-get build-essential
Better than download a full load of useless software in 4 CDs.
Is the guy insane?
You don't eat them though...
See the answer below. Our addiction to oil is so huge that even using all the fertile land to produce ethanol will make up to only about 40% of our needs, without having any land left for food crops. THe problem is not the land itself. It's the water supply. After oil we are addicted to water, which has been used most of the time not in th optimal way. Chemical contamination, overuse for electrical production, reduced the amount usable, for agriculture. In addition, let's not forget the desertification of the North American continent, and of Europe. Once the land becomes desert, there is usually no way back. For this reason I strongly opposed ethanol and the huge investments regarding its use instead of crop for food. Let's spend research money to improve solar and nuclear. With a solid alternative source of energy, we could extract potable water from the sea. It's a dream, I know, but poper use of nuclear energy would allow that. So I am in favor of real alternative sources of energy, which do not rely on current primary needs, and ethanol is not one of those.
Very true. Ethanol is viable only if produced from non-food-related products. Simple math, we don't have enough usable and fertile land to grow food crops AND fuel crops.
Much of Physics doesn't run on Macs, but Windows. A vast majority of Electron microscopes, probe microscopes runs on Windows too (Omicron, Philips, Veeco). It's funny that some of the best electron microscopy software actually runs on Macs, but only for PPC, so big centers (like the NCEM at the Lawrence Berkeley Lab) runs on old G4. Raman probes runs on Windows for the most part. A lot of custom built equipment uses National Instruments LabView, which runs bot Windows and Mac, but it's commolny installed in PCs, being usually cheaper. So unless you are doing advanced imaging world, most of the equipment runs on PCs. So you are probably right about biology, linguistics and possibly others. But the physics and material science world is mainly a Windows world. P.S. It's funny because sometimes you see in labs old Macs (from the late 80s and 90s) running very old equipment. Macs were the standard in Physics labs. At some point that changed....