The user experience might have suffered a little with the switch to BOINC, but think of what has been gained: a generic distributed computing system, where projects can fairly easily package up their computational problem for SETI@home-like processing, without having to go through all the work of setting up the distributed computing infrastructure! And with BOINC, you can specify what percentage of your cpu resources you want to go to which processes... it's like the United Way for CPU cycles. (-:
I work on the LIGO project, which is searching for gravitational waves using several huge interferometers (one out in the desert of Eastern Washington at the old Hanford Works, were Plutonium was made for the Manhattan Project; the other currently being belted by the hurricane down in Louisiana, 3002 km away). I was really impressed by the Einstein@home talk at the most recent meeting. The computation by the Einstein@home project is really very valuable to the LIGO project. If you want to run Einstein@home, it will really help LIGO.
Exactly.. MRAM will actually encourage better programming, since users will have no reason to reboot other than because of crashes; rebooting will be even more of an onus.
Nonetheless, I don't know whether that particular aspect of MRAM will make any difference. I can't remember the last time I had to reboot Linux due to a software crash! Virtual/protected memory systems are very good about isolating applications from each other already.
The real benefits of MRAM are far more exciting. Unlike conventional DRAM, MRAM does not need to be refreshed (it's nonvolatile), yet its fast enough and could be cheap enough to replace DRAM. The result is a huge POWER SAVINGS since you wouldn't have to use power to run the DRAM refresh cycles. Moreover, MRAM is simpler, so it could have higher integration densities, and thus would be cheaper.
MRAM falls into the general domain of "spintronics" (which is the name given to technologies which exploit the spin of electrons in addition to their charge). One of the most exciting applications of spintronics is in reconfigurable computing. We could make "real" reconfigurable logic -- cheap nonvolatile FPGA's. Your processor could quite literally rewire itself on the fly, adapting to the task at hand. Very exciting.
It wasn't inherently obvious, or accidentally discovered, that you could burn stuff with a giant fresnel lens? And you knew what a fresnel lens was in the first place?
I have no idea what a 'Gimp' is but computer users of a UNIX persuasion will recognise
What kind of an attitude is that? Anyway, GIMP is an acronym that stood for General Image Manipulation Program. Now the 'G' has changed to GNU.
that it as the name of an open source (read 'free') graphics editor that is often compared to PhotoShop for its range of features and abilities.
Gimp is open source AND free, but these are independent qualities. Open source does not mean free, and free does not mean open source.
'Open source' means that the source code is available at no cost to anybody that wants to download it, use it, modify it, use it to fill
yes on the "available," no on the "no cost"
empty hard drives - whatever. The point is that it is not a ready to roll program, it has to be 'compiled' before it can be used - think of it as 'ready to cook' rather than 'ready to eat'. For people who use UNIX, this is par for the course. For a lifelong Mac zealot like me, it's like being served up a dead chicken on a plate instead of a Chicken Madras.
oh come on. Are you using OS X? It is unix. You can download pre-compiled Gimp, but if you do want to compile it, it's usually a matter of two commands: "./configure; make". That's a dead chicken?
When I saw that somebody had released The Gimp in a 'just add boiling water and stir' package, I though I'd have a look.
Before you can use MacGimp, you have to install X11, a UNIX windowing system - so that you don't have to do everything from the command line. That is a free download from Apple and I figured out how to get it up and running without too much difficulty. Running OS X soon knocks all notions of 'The computer for the rest of us' out of you!
whatever you say...
Opening MacGimp for the first time was like stepping out onto the surface of an alien planet - and the gravitational pull was different too. Menus were attached to windows instead of being in the menu bar.
oh no!
In fact, there was one menu bar on the tools palette and another on the canvas window. Weird!
We're not in kansas any more...
The tools palette has recognisable icons, and tooltips to explain them when you don't.
how friendly!
The trouble is that they are in no particular order.
They are not grouped or arranged into logical sequences, just thrown down like tools on a garage floor. Trying to select any of them proved to be equally haphazard. It took two or more clicks on
It's only a five-by-five grid, so I should hope that you could find the "recognizable icons." Moreoever, they are in some kind of order. The top row is selection tools, towards the bottom are drawing tools, and weirder things are in between! Okay, perhaps not the best scheme, but definitely not defective.
any of them to activate them, then a contextual menu appeared underneath giving a plethora of tool options.
Where would you prefer this "plethora of options" to be?
Exploring the rest of the features, it was clear that a lot of work had gone into the program, it had everything you would expect from a serious graphics editor but then many of the same features turned up in different places which is rather
So, you're saying that it has all the features of photoshop, but some of them appear in different places. Note: IT'S NOT PHOTOSHOP. Of course it's going to be different. Different is not bad.
disorientating. That, coupled with icons everywhere, even in the menus, made the program look over-complicated, obviously to appeal to people who fail to see the benefits of elegant simplicity.
When I got used to the fact that the 'open' dialog wouldn't show me any of my 'ordinary' Mac folders or anything in my 'files' hard drive I started thinking 'UNIX' and moved some photos into folders where they could be acc
I've seen many instances of arguments between research students and faculty about open-sourcing code.
The greater problem, in my experience, has been a general lack of motivation or interest in publishing software at all. Often time software is just written as an in-house hack. Very little effort goes into generality, and even less frequently does someone go to the effort to package something up for distribution. Scientists are specialists in their particular field, and are usually not really aware of the open source community, or at least not aware that they potentially have a place within it.
Government agencies actually can't copyright the IP they produce. Check out the license on Foremost, for example:
* This is a work of the US Government. In accordance with 17 USC 105,
* copyright protection is not available for any work of the US Government.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Unfortunately, this lack of copyright doesn't mean that Govt software will ever be released. Moreover, software created with government funding but not directly by the government is exempt from this rule. I'm not sure what special rules exist, but I think that the government contractors / grantees get to license the software they develop with federal money as they see fit.
For example, Traceroute was written at the Lawrence Berkeley National Laboratory, which is operated under contract from the U.S. Department of Energy by the University of California. Thus UC gets copyright. The copyright notice on traceroute is:
* Copyright (c) 1988, 1989, 1991, 1994, 1995, 1996, 1997, 1998, 1999, 2000
* The Regents of the University of California. All rights reserved.
*
Unfortunately, I think that the vast majority (excuse the use of a countable noun for an uncountable quantity) of software developed with government money is never released to the public in any way, let alone under any free / open source terms.
You know all those "Segmentation Fault" errors that ROOT gives you?
A real tool doesn't do that.
It's really very cute how enamored particle physicists are by C++. It's very fitting it turns every software construct into something they're familiar with, a particle! er, object. Too bad they can't ditch the FORTRAN habits.
"You can write bad FORTRAN in any language." - my advisor
When I first read about the double slit experiment, I said to myself, "That can't be!"
I cut two slits into a piece of cardboard and directed a flashlight beam through the slits at a wall.
And I observed exactly what one would expect, two diffuse bright spots. I said, "Hmph."
Of course, when I learned a little more, it was obvious why this didn't work. In order to see the interference pattern, your light must be coherent and columnated (as from a laser), and your slits must be very close together, and narrow (with dimensions similar to the wavelength of light). You pretty much need to use a laser as your light source, and rather than a "board" with slits, a sheet of metal with two very thin slits cut into it, very close together.
Something I found very fascinating is that the diffraction pattern you get is the fourier transform of the pattern of slits the system of interference exactly implements the fourier transform integral!
All they need to do is trace the code in the actual Linux kernels. If they can show that all of that code comes from non-copyrighted or properly released sources, then it doesn't matter what code SCO has.
They were actually looking to use "inane.com" (an adjective selected to describe attempts by amazon.com to gain market share). however, the english spelling was already taken. fortunately the market-droids had some exposure to k-rad 31337-speak and, seeking to appeal to that audience, so elected to adopt "A9" as the public designation of this project.
The fact of the matter is that a machine with 100% CPU utilization uses a lot more electricity than one with low utilization. The extra cycles aren't free.
I measured this in 1997 on some kind of AMD K6 machine. IIRC, running dnetc doubled the power consumption of the machine.
That's true... I worked at a DoD-funded research facility where we were compelled to deliver a product that ran on Windows NT simply because that was the Navy requirement. The in-house development was done in Linux and we deployed on cygwin... sigh.
They are switching to GSM, but their GSM coverage is among the worst. Moreover, they do not support GSM data calls (i.e., say you want to call a BBS or a traditional ISP), while Cingular does.
I don't understand this.. modern operating systems already mark pages with individual write and execute permissions. Buffer overruns are exploitable when they occur in pages which are marked both writeable and executable, such as the stack... ?? I don't see why this requires new hardware to fix.
It's hard to call orkut 'innovative', when it appears to be an exact clone of Friendster with Crushlink-sans-spam thrown in for good measure (and without Friendster's poor stability and abyssmal performance). I wouldn't be surprised, though, if orkut got some blogging capabilities integrated.. that could be kind of interesting
Slashdot Mirror
Bizarrely enough, it's 34 deg F here right now!
And it's supposed to be 56 degrees tomorrow.
Here in Rochester, we appreciate global warming!
The user experience might have suffered a little with the switch to BOINC, but think of what has been gained: a generic distributed computing system, where projects can fairly easily package up their computational problem for SETI@home-like processing, without having to go through all the work of setting up the distributed computing infrastructure! And with BOINC, you can specify what percentage of your cpu resources you want to go to which processes... it's like the United Way for CPU cycles. (-:
I work on the LIGO project, which is searching for gravitational waves using several huge interferometers (one out in the desert of Eastern Washington at the old Hanford Works, were Plutonium was made for the Manhattan Project; the other currently being belted by the hurricane down in Louisiana, 3002 km away). I was really impressed by the Einstein@home talk at the most recent meeting. The computation by the Einstein@home project is really very valuable to the LIGO project. If you want to run Einstein@home, it will really help LIGO.
Some pictures from LIGO: http://www.livejournal.com/users/nibot/tag/ligo
Wikipedia entry: http://en.wikipedia.org/wiki/LIGO
No: paying lawyers will not decrease your revenue. Revenue is total funds taken in, without subtracting expenses.
Exactly.. MRAM will actually encourage better programming, since users will have no reason to reboot other than because of crashes; rebooting will be even more of an onus.
Nonetheless, I don't know whether that particular aspect of MRAM will make any difference. I can't remember the last time I had to reboot Linux due to a software crash! Virtual/protected memory systems are very good about isolating applications from each other already.
The real benefits of MRAM are far more exciting. Unlike conventional DRAM, MRAM does not need to be refreshed (it's nonvolatile), yet its fast enough and could be cheap enough to replace DRAM. The result is a huge POWER SAVINGS since you wouldn't have to use power to run the DRAM refresh cycles. Moreover, MRAM is simpler, so it could have higher integration densities, and thus would be cheaper.
MRAM falls into the general domain of "spintronics" (which is the name given to technologies which exploit the spin of electrons in addition to their charge). One of the most exciting applications of spintronics is in reconfigurable computing. We could make "real" reconfigurable logic -- cheap nonvolatile FPGA's. Your processor could quite literally rewire itself on the fly, adapting to the task at hand. Very exciting.
Now you know?
It wasn't inherently obvious, or accidentally discovered, that you could burn stuff with a giant fresnel lens? And you knew what a fresnel lens was in the first place?
You can also build your own giant spiral fresnel reflector at home.
Powerful = fastest computation
Mod parent up! Now let's get out the amp meters.. let's compare these computers kilowatt to kilowatt.
Then we can be quite literal about which computer has "more horsepower," as one kilowatt is about 1.34 horsepower. (-:
I have no idea what a 'Gimp' is but computer users of a UNIX persuasion will recognise
What kind of an attitude is that? Anyway, GIMP is an acronym that stood for General Image Manipulation Program. Now the 'G' has changed to GNU.
that it as the name of an open source (read 'free') graphics editor that is often compared to PhotoShop for its range of features and abilities.
Gimp is open source AND free, but these are independent qualities. Open source does not mean free, and free does not mean open source.
'Open source' means that the source code is available at no cost to anybody that wants to download it, use it, modify it, use it to fill
yes on the "available," no on the "no cost"
empty hard drives - whatever. The point is that it is not a ready to roll program, it has to be 'compiled' before it can be used - think of it as 'ready to cook' rather than 'ready to eat'. For people who use UNIX, this is par for the course. For a lifelong Mac zealot like me, it's like being served up a dead chicken on a plate instead of a Chicken Madras.
oh come on. Are you using OS X? It is unix. You can download pre-compiled Gimp, but if you do want to compile it, it's usually a matter of two commands: "./configure; make". That's a dead chicken?
When I saw that somebody had released The Gimp in a 'just add boiling water and stir' package, I though I'd have a look.
Before you can use MacGimp, you have to install X11, a UNIX windowing system - so that you don't have to do everything from the command line. That is a free download from Apple and I figured out how to get it up and running without too much difficulty. Running OS X soon knocks all notions of 'The computer for the rest of us' out of you!
whatever you say...
Opening MacGimp for the first time was like stepping out onto the surface of an alien planet - and the gravitational pull was different too. Menus were attached to windows instead of being in the menu bar.
oh no!
In fact, there was one menu bar on the tools palette and another on the canvas window. Weird!
We're not in kansas any more...
The tools palette has recognisable icons, and tooltips to explain them when you don't.
how friendly!
The trouble is that they are in no particular order.
They are not grouped or arranged into logical sequences, just thrown down like tools on a garage floor. Trying to select any of them proved to be equally haphazard. It took two or more clicks on
It's only a five-by-five grid, so I should hope that you could find the "recognizable icons." Moreoever, they are in some kind of order. The top row is selection tools, towards the bottom are drawing tools, and weirder things are in between! Okay, perhaps not the best scheme, but definitely not defective.
any of them to activate them, then a contextual menu appeared underneath giving a plethora of tool options.
Where would you prefer this "plethora of options" to be?
Exploring the rest of the features, it was clear that a lot of work had gone into the program, it
had everything you would expect from a serious graphics editor but then many of the same features turned up in different places which is rather
So, you're saying that it has all the features of photoshop, but some of them appear in different places. Note: IT'S NOT PHOTOSHOP. Of course it's going to be different. Different is not bad.
disorientating. That, coupled with icons everywhere, even in the menus, made the program look over-complicated, obviously to appeal to people who fail to see the benefits of elegant simplicity.
When I got used to the fact that the 'open' dialog wouldn't show me any of my 'ordinary' Mac folders or anything in my 'files' hard drive I started thinking 'UNIX' and moved some photos into folders where they could be acc
I've seen many instances of arguments between research students and faculty about open-sourcing code.
The greater problem, in my experience, has been a general lack of motivation or interest in publishing software at all. Often time software is just written as an in-house hack. Very little effort goes into generality, and even less frequently does someone go to the effort to package something up for distribution. Scientists are specialists in their particular field, and are usually not really aware of the open source community, or at least not aware that they potentially have a place within it.
Seriously. Mod the grandparent message down. Windows can do those things just fine.
Government agencies actually can't copyright the IP they produce. Check out the license on Foremost, for example:
* This is a work of the US Government. In accordance with 17 USC 105,
* copyright protection is not available for any work of the US Government.
*
* This program is distributed in the hope that it will be useful, but
* WITHOUT ANY WARRANTY; without even the implied warranty of
* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.
Unfortunately, this lack of copyright doesn't mean that Govt software will ever be released. Moreover, software created with government funding but not directly by the government is exempt from this rule. I'm not sure what special rules exist, but I think that the government contractors / grantees get to license the software they develop with federal money as they see fit.
For example, Traceroute was written at the Lawrence Berkeley National Laboratory, which is operated under contract from the U.S. Department of Energy by the University of California. Thus UC gets copyright. The copyright notice on traceroute is:
* Copyright (c) 1988, 1989, 1991, 1994, 1995, 1996, 1997, 1998, 1999, 2000
* The Regents of the University of California. All rights reserved.
*
Unfortunately, I think that the vast majority (excuse the use of a countable noun for an uncountable quantity) of software developed with government money is never released to the public in any way, let alone under any free / open source terms.
You know all those "Segmentation Fault" errors that ROOT gives you?
A real tool doesn't do that.
It's really very cute how enamored particle physicists are by C++. It's very fitting it turns every software construct into something they're familiar with, a particle! er, object. Too bad they can't ditch the FORTRAN habits.
"You can write bad FORTRAN in any language." - my advisor
1. Mozilla
2. Mindterm SSH (er, and the JRE I guess)
3. Exceed
3 1/2. Adobe Acrobat Reader
Note that the above turn Windows into an excellent dumb terminal. (-:
5. TextPad
6. LaTeX
7. Matlab
8. vmware
9. MSVC6
10. OpenOffice
Have you actually tried it?
When I first read about the double slit experiment, I said to myself, "That can't be!"
I cut two slits into a piece of cardboard and directed a flashlight beam through the slits at a wall.
And I observed exactly what one would expect, two diffuse bright spots. I said, "Hmph."
Of course, when I learned a little more, it was obvious why this didn't work. In order to see the interference pattern, your light must be coherent and columnated (as from a laser), and your slits must be very close together, and narrow (with dimensions similar to the wavelength of light). You pretty much need to use a laser as your light source, and rather than a "board" with slits, a sheet of metal with two very thin slits cut into it, very close together.
Something I found very fascinating is that the diffraction pattern you get is the fourier transform of the pattern of slits the system of interference exactly implements the fourier transform integral!
All they need to do is trace the code in the actual Linux kernels. If they can show that all of that code comes from non-copyrighted or properly released sources, then it doesn't matter what code SCO has.
You haven't looked at the Linux source, have you?
Do you have any idea how much of it there is?
They were actually looking to use "inane.com" (an adjective selected to describe attempts by amazon.com to gain market share). however, the english spelling was already taken. fortunately the market-droids had some exposure to k-rad 31337-speak and, seeking to appeal to that audience, so elected to adopt "A9" as the public designation of this project.
In the site info, it lists an ASIN number for each page!
what's next? referral fees? it's funny that all web pages are suddenly inducted into the same numbering system as *products for sale*.
The fact of the matter is that a machine with 100% CPU utilization uses a lot more electricity than one with low utilization. The extra cycles aren't free.
I measured this in 1997 on some kind of AMD K6 machine. IIRC, running dnetc doubled the power consumption of the machine.
That's true... I worked at a DoD-funded research facility where we were compelled to deliver a product that ran on Windows NT simply because that was the Navy requirement. The in-house development was done in Linux and we deployed on cygwin... sigh.
Their customer service is atrocious.
They are switching to GSM, but their GSM coverage is among the worst. Moreover, they do not support GSM data calls (i.e., say you want to call a BBS or a traditional ISP), while Cingular does.
I like how the "Property of US Postal Service" label on a mail crate is blurred out in this photo:
V 2% 5CIMG/IMG_0019.JPG
http://www.cowanalexander.com/Events/CA031004/P
I don't understand this.. modern operating systems already mark pages with individual write and execute permissions. Buffer overruns are exploitable when they occur in pages which are marked both writeable and executable, such as the stack... ?? I don't see why this requires new hardware to fix.
Oh yeah, it's cute that Orkut has the little "beta" in the corner, just like Friendster. And of course, Frienster took that from Napster...
It's hard to call orkut 'innovative', when it appears to be an exact clone of Friendster with Crushlink-sans-spam thrown in for good measure (and without Friendster's poor stability and abyssmal performance). I wouldn't be surprised, though, if orkut got some blogging capabilities integrated.. that could be kind of interesting
Finally, a use for the huge budget surplus! yay!