Politely, this is why I read the major papers, various periodicals, and use the news archives, as well as their campaign literature (if they're the challenger and don't have a public record) as well as the league of women voters sheets, before elections. I did this even when living in a one-party state (Chicago), because I needed to make at least an informed (if pointless) opposition. This is how to get information about the candidate, which while spun, is rarely spun by the candidate for best effect.
Having listened to telemarketed pitches in the past, I fail to see how forcing people to sit through deliberate obfuscation on the part of candidates, such as the misleading polls and false rumors spread about McCain in SC during the 2000 primaries, benefits the democratic process. Those calls are (at best) to raise money for the national party, when they aren't deliberately sliming the opposition. I get mailings from the Republicans, which have polls (heh) which are skewed to make you believe you're either a war-mongering, authoritarian, supply-sider or a spaced-out hippie communist, and from the Democrats which imply you're either a left-over McGovernite Great-Society socialist, or an iron-age warlord. The calls are more of the same, and neither is useful to the democratic process, except to excite their "base" (i.e. the extremist 10% on each end of the spectrum). (and if you want to get on both party's mailing lists, try subscribing simultaneously to the Wall Street Journal, and The Nation)
The country is ruled by extremists for a variety of reasons, of which congressional-district gerrymandering, and closed-primaries (which energize the moon-bat wing) are two of the most egregious. Add to that a general problem that no matter who you vote for, the government gets elected (CF Chicago above), and the recipe for apathy is about perfect. If you want a functioning democracy, then work on Open primaries (people regardless of party may vote for the candidates of either party, to ensure bipartisan support), draw congressional districts explicitly to exclude party affiliation, and use term-limits, age-limits, or basic-competency-limits to ensure some turnover now and then.
Don't fear protest candidates, try to force run-offs, and gather your own information; don't rely on politicians to paint their own image.
If disturbed because I'm waiting for someone real to call, I have started off with, "Listen, you goat-buggering parasite....", with the volume rising rapidly. I have suggested that telemarketers get some job that requires skill, demonstrates they have self-respect, and produces a benefit to society, such as squeegee-man at an interstate on-ramp. This level of effort and invective is rare, as it takes a particularly annoying caller to be worth the time. The goal is to leave them with their nerves sufficiently rattled that they'll flub the next couple of calls.
"I'm sorry, but I am morally opposed to everything your party stands for. Do not call here again, and best wishes to your candidate's opponent" (click)
Unfortunately, I haven't felt it worth the time, since the do not call list went into effect, to play the William Tell overture on the phone keypad, or other more creative responses.
Do it yourself. If you're annoying people, DIY Hendrix is better than recorded. Be sure to wear your American Flag suit when you step out on your back deck to serenade him. An old Heathkit amp with some bad solder joints for extra distortion will help as well.
I was feeling lazy, and didn't want to bother typing, "The Alpha of the 21st Century". I ran Itanium-2s, on HP workstations, and there really was nothing like them. The first generation (the one derided as a failure out of the box), running Beta compilers from Intel, ran circles around our ultrasparc-III systems (not cheap midrange servers). The second generation did no worse than the Xeons of the era on integer-bound code, and walked away on floating point. It's an incredible chip.
However, Intel priced them to recoup their entire development cost on the first batch, and hired IBM's marketing team from OS/2, and here we are.
My lab used to have a cartoon taped on the wall, from the days of the hot new 386, "we finally figured out that what consumers want is not more speed or more memory, but just really great tailfins."
Has anyone tried putting tailfins and chrome from a '59 caddy on a Dell and see if it sells better.
Intel had a true, high-performance, 64-bit product out years ahead of AMD, and all you people out in Desktop-land went "EWWWWWW!!!". "it's too hot, too expensive, too hard to code, and it won't run Quake e^(pi)!". Intel's fault was believing the old IBM saying, "sometimes you have to drag the customer, kicking and screaming, into the future". Instead, AMD took what they already knew how to do (improve IA-32), bolted some reasonably-well thought out 64-bit extensions onto it, and sold it as a future-proofed Xeon. Intel hemmed and hawed, eventually gave up and did the same in a manner compatible with AMD, and this is where we are now; stuck with the x86 until the Sun grows cold.
I did the PhD right out of undergrad, but had spent over a year as a Co-op at a major chemical company. In the sciences, at least, it seems that my colleagues who'd been in industry (myself included) had the hardest time adjusting to being grad students. It wasn't just the paycheck (though I did think that I should have had a profit-sharing plan from the univeristy like my company did, given what they charged), but also the sense of purpose, and frequently the facilities. It was hard to get used to having instrumentation (touted as state of the art, world class in brochures), that my company would have dumpstered, and hard to go back to sitting in class, taking exams, etc. It's better once you get back in the lab and they leave you alone, but two years of classes, preliminaries, cumes, etc., is an aggravating process to go through. The other issue was, of course, the financial one. People with families tend to have an issue with being paid grad-student stipends, and working 12 to 14 hour days (because, sometimes you use the instrument when it's both available and running correctly, not when you feel like it).
Grad school is kind of like the military; it's best to join when you're too young to know better.
IBM demo'd an image searching system about 10 years ago, where you'd sketch the outlines of what you were looking for. Probably hasn't been refined enough yet.
As for music feeling "sticky", I'm baffled on that one. I've owned various media over 30++ years, (sheet, vinyl album, tape, CD, banjo), and sticky was never a desirable state of affairs. However, as long as there is tactile feedback available, then classical should feel like polished wood, modern orchestral (Bartok), like sheet-metal sculpture, Rock disordered with edges everywhere, and Barry Manilow like you're sinking in lukewarm jello while barely heated tofu is poured over your head.
I'll admit, I've never tried this because I couldn't get anyone to spring for the displays, or the space to stack several monitors, but just follow their instructions, and you're off and running.
Thank you for the cut-and-paste reminder. While seemingly efficient (highlight then middle click), make one bump while moving the mouse and you have to do it over. I'd forgotten that one.
It's amazing how quickly you get used to not having to deal with such issues. Now let's hope His Steveness can keep it all on track a few more years.
No, seriously, if you're setting up a cluster where your work can be batch-queued, or intend to run MPI, then Rocks http://www.rocksclusters.org/ is the way to go. It also comes with tools such as SGE (Sun Grid Engine) or OpenPBS pre-configured, Intel compilers and libraries ready for you to drop a license onto (but of course the entire GNU suite is there as well, including Ada), has more monitoring tools (plus some nice web-based interfaces) than you can shake a stick at, and runs on IA-32/AMD-64/IA-64 (Itanium). It also has a Roll to help build a tiled display wall, which would be a really cool use of a small cluster.
They're also really great guys.
On the other hand, Oscar is supposed to be good, and if you're not into the whole batch-mode thing, you can get OpenMosix up and running using http://clusterknoppix.sw.be// ClusterKnoppix, and just fire jobs off into space and let them find their own unburdened node.
But still, Rocks is really an elegant and clean way to go, plus it will scale up in case you're going to deploy a huge one of these for real after you get your feet wet.
Actually, having gone the other way (Linux On The Desktop, (LOTD) 10 years to OSX), I'd be happy if the GUI/App people in the Linux community took a page from OSX, and actually spent some time worrying about polish, internal consistency, and interoperability. I've run (recently) RHEL 4.1, Ubuntu (which is a long way to making this rant irrelevant), SuSE (If you knew SuSE, like I knew SuSE...), and while they're much better than they were when I started, a little polish could still be applied. Surely some of you have graphic-arts oriented friends you could hand a six-pack of something decent to and say, "make these icons look pretty". Surely someone could agree, "you know, a hidden save dialogue activated with a right-mouse click, in addition to the identical one at the top of the window is just silly". An agreement that little buttons with pop-up menus that look like Win95/2K isn't the best interface to copy.
As for the rest, I write script files on my OSX system, run IBM's Fortran and C compilers to create command-line applications, and end up hand-twiddling my rc files to set things like SHMMAX which is just too small out of the box for real work. The entire difference to me is in the user-land experience. The Mac is just Clean, in a way that's hard to appreciate unless you come to it from too many years of CDE and CDE clones. The beauty is that I can do all that manual work if I want to (find mixed with grep is still often faster for me than Spotlight), but I don't have to. Voluntary complexity is a wonderful thing.
As one of my colleagues in High-Energy Physics put it, "A Mac is a Unix box that runs iTunes."
Given Sun's current financial state, 64-bit on AMD/Intel Solaris-x86, and technologies such as Java, it could easily be eaten by apple. Paying the premium to ship SnApples without having to deal with the screaming hordes of Linux geeks may be worth it for a company as tightly-wound and controlled as Apple. I doubt they want any confusion in people's minds over whether Aqua is their own, or just a shell on top of Gnome.
In the end, does it matter, since they entire point of the Apple experience is in the UI and higher-level libs, which aren't open, and aren't about to become so.
On the other hand, I, for one, welcome our new Efficient, Poly-cored, Multi-threaded, Preemptively Multitasking Overlords.
iPod Pico: 4GB storage, doubles as an eyebrow ring. This will be closely followed by the iPod Femto; same storage, but it's applied at the store with a roller as a temporary tattoo. It washes off when the warranty expires.
So, idle question: why don't they just license, pre-install, and use Tripwire on Windows boxes? (or reimplement it in-house, to avoid dealing with an outside company)
Fair enough, and I'll admit that I have one code that I often built using G77 for stability reasons. I am (or was) actually rather fond of G77, having run it originally on OS/2 using the EMX suite, and then used it as my primary development platform for years. However, given that Fortran has been a bit of a red-headed stepchild in the GCC suite, and the current F90 compiler is choking on a major F77 code I use on a daily basis, I have a hard time recommending it any more. I hope that GFortran improves enough that I can go back to using it, as it was nice having a consistent Fortran across various platforms for debugging.
Sadly , I miss G77; I used to use it to build older codes that my colleagues would bring to me, that modern compilers such as IFort choked hard on. I probably just need to be patient with GFortran, since if it gets within a few percent of the Absoft compiler, then it's the most cost-effective 64-bit compiler for what is soon to become my legacy cluster. (Mac OS-X) I had purchased them on the mistaken assumption that XLF had a future, and for 32-bit code, I really could not be happier, unless some of the underlying algorithms in the packages were improved. Actually, I could be happier; the programs could also colllate the data and write up the results for me, but that's an AI problem, not a Fortran problem.
When running on PPC hardware or Power3 hardware (Mac G5, AIX, respectively), for quantum chemistry packages, G77 was capable of being up to 50% slower than XLF. At 30% slower, that's equilvalent to running your machines for 8 months of the year, versus 12. Since I run simulations that routinely take two months, that 30% is a big difference.
Because if done right, you'll use Windows as the host, because the Windows functionality you need will run sufficiently better, and Microsoft sells another license. Or, you run Linux as a the host, but think warm thoughts about Microsoft because they put money into Xen, buy another license, and Microsoft gets your money as well.
Basically, it's to capture mindshare, prevent datacenters from migrating their Windows systems to Linux hosts when they start running virtualization, and ensuring that, even if you are primarily a Linux-type, you're still paying Microsoft. When you're that big, every extra dollar counts towards proving to Wall Street that you're still growing. There's a lot of money to be made in Virtualization, by ensuring that your OS (Vista/Windows 2007-and-Counting Server) is the host everywhere, rather than the client. If you also sell client licenses for older versions of the OS, for people with apps that can't migrate, so much the better.
To the poster with their "there is only one C"; Horsefeathers. That there is Only One C explains why every C program I have is a twisty little maze of #ifdefs, often for minor differences in machine/compiler combinations, and as likely to break after a compiler upgrade as continue to build. It continues to exist only because it's suitable for assembly-language-type functions, such as OS kernels, which means it ships with every unix out there, and is taught to unsuspecting comp-sci students because it came free with the OS. It's amazing that after the problems of the past few years, including buffer-overrun exploits, memory leaks, etc, caused by C's unsafe, seat-of-the-OS memory handling, that people continue to defends its use as a general purpose language.
The OP may be trying to be funny, but in fact a modern Fortran (F90/95) is easy to do string manipulation in, has dynamic memory, is type-safe, and is arguably a much higher-level language than C. (on the other hand, F90/95 bears a curious resemblence to Ada, but we'll let that slide). Memory aliasing isn't nearly the issue it is for C programs, which is why Fortran still out-performs C/C++ on any modern architecture, as long as you're using a competent compiler. (i.e. I don't want to hear about how slow your G77/GFortran programs are; they're great tools, but they're primarily Fortran Lints (flints)). Where Fortran fails is that is has the same name as a particularly onerous language from the mid-60s, known variously as FORTRAN-IV or FORTRAN66. If it was named differently, such as Cerulean, for instance, you'd complain about its verbose, Python-esque syntax, with Begin and End, rather than { and }, but you'd grudgingly learn it, write interfaces to libraries, start new projects in it, and treat it as an appropriate tool. Flame wars would break out here as the Cerulean programmers made fun of the Java-weenies, because their language is byte-code interpreted, and doesn't have matrix operations built in as primitives (how can you do quick 3d graphics without matrix ops). Because of the antique name, you picture a black and white room full of guys in white shirts with pocket-protectors and crewcuts, smoking while they load pounds of manila cards into the keypunch.
Back when we were in college we used to comment that C was a step backwards, as in terms of being a high-level language it ranked right up there with VAX assembler. Let it go. Use C++/Objective-C/Java if you're obsessed with line-noise syntax, or move on to Eiffel http://eiffelsoftware.origo.ethz.ch/index.php/Down loads//Ada or even SmallTalk http://www.squeak.org/ for new work.
>>uppose that I replace this 1.25GHz G4 Mac Mini....According to you, it wouldn't be a computer.
It's not a computer. It's an iTunes Appliance that happens to also run Word. And in case you're thinking about it, no you can't have mine. I will defend that desk-saving, hearing-sparing, piece of productivity with a pointed stick if necessary.
I've done the music instead of coffee as a wake-up move before, and it seems to work well, without the acid stomach of too much coffee. Just be picky; Wagner/Bartok works better than Brahms/Chopin. New-age white-noise is right out.
If you drill through the FA (the PDF version), you'll see that aP2 is involved in lipid metabolism and transport, and probably synthesis of several derivatives. Unfortunately, when over-expressed (or over supplied with raw material), it may be linked to some auto-immune diseases. It's probably similar to the mechanism by which your body conserves calories and stores them as fat; a great idea in pre-modern or even pre-20th century societies, but with unfortunate side-effects in our modern, calorie-rich, environments.
Well, we're a school (a mid-sized university physical science department), and I can say that despite discussions concerning the options, we're certainly not upgrading anything to Linux. Win98 machines are being repurposed as Lab machines, with a parent Win98 OS image distributed (probably by Norton ghost, though I stay away from those details lest I get sucked in to support them) either once a week, or when a machine is suspected of being borked. Slightly newer machines which can handle XP are generally improved by stealing the memory from another comparable one to get it up as high as possible, then invoking our academic discount to buy an XP-Pro license. Machines too old for either of these solutions are simply being sent to recycling.
The linux machines exist in CompSci's cluster, running back-end services (DNS, Web, etc), and one in my lab for portability testing. The issue, as always, is that we have software that expects Windows-(something), and dedicating the resources to make it behave seamlessly under Wine is not cost effective versus continuing to run Windows. I suspect other educational institutions (based on having worked at them, and having seen the same policies) are doing the same thing.
The article should read, "Win98 EOL causes mass upgrades at school/govt/business, dumpster-divers scavenge abandoned machines and put desktop linux on them". It's a wonderful dream, but there is just too much Windows-centric software out there to see it happen.
So, did the superchickens have more dark meat on them, or light? Parts that get exercised (thighs) are darker than parts that don't (breast on a non-flying chicken). I can see this at Wegman's now, "boneless thighs from free-range, hormone-free, pre-centrifuged, SuperChickens".
Slashdot Mirror
Politely, this is why I read the major papers, various periodicals, and use the news archives, as well as their campaign literature (if they're the challenger and don't have a public record) as well as the league of women voters sheets, before elections. I did this even when living in a one-party state (Chicago), because I needed to make at least an informed (if pointless) opposition. This is how to get information about the candidate, which while spun, is rarely spun by the candidate for best effect.
Having listened to telemarketed pitches in the past, I fail to see how forcing people to sit through deliberate obfuscation on the part of candidates, such as the misleading polls and false rumors spread about McCain in SC during the 2000 primaries, benefits the democratic process. Those calls are (at best) to raise money for the national party, when they aren't deliberately sliming the opposition. I get mailings from the Republicans, which have polls (heh) which are skewed to make you believe you're either a war-mongering, authoritarian, supply-sider or a spaced-out hippie communist, and from the Democrats which imply you're either a left-over McGovernite Great-Society socialist, or an iron-age warlord. The calls are more of the same, and neither is useful to the democratic process, except to excite their "base" (i.e. the extremist 10% on each end of the spectrum). (and if you want to get on both party's mailing lists, try subscribing simultaneously to the Wall Street Journal, and The Nation)
The country is ruled by extremists for a variety of reasons, of which congressional-district gerrymandering, and closed-primaries (which energize the moon-bat wing) are two of the most egregious. Add to that a general problem that no matter who you vote for, the government gets elected (CF Chicago above), and the recipe for apathy is about perfect. If you want a functioning democracy, then work on Open primaries (people regardless of party may vote for the candidates of either party, to ensure bipartisan support), draw congressional districts explicitly to exclude party affiliation, and use term-limits, age-limits, or basic-competency-limits to ensure some turnover now and then.
Don't fear protest candidates, try to force run-offs, and gather your own information; don't rely on politicians to paint their own image.
I've chosen three basic responses:
"Good day." (click) (90% of the time)
If disturbed because I'm waiting for someone real to call, I have started off with, "Listen, you goat-buggering parasite....", with the volume rising rapidly. I have suggested that telemarketers get some job that requires skill, demonstrates they have self-respect, and produces a benefit to society, such as squeegee-man at an interstate on-ramp. This level of effort and invective is rare, as it takes a particularly annoying caller to be worth the time. The goal is to leave them with their nerves sufficiently rattled that they'll flub the next couple of calls.
"I'm sorry, but I am morally opposed to everything your party stands for. Do not call here again, and best wishes to your candidate's opponent" (click)
Unfortunately, I haven't felt it worth the time, since the do not call list went into effect, to play the William Tell overture on the phone keypad, or other more creative responses.
Do it yourself. If you're annoying people, DIY Hendrix is better than recorded. Be sure to wear your American Flag suit when you step out on your back deck to serenade him. An old Heathkit amp with some bad solder joints for extra distortion will help as well.
Because of the real-world equivalent of the one good moment in Attack of the Clones, "If it isn't on Slashdot, it doesn't Exist."
I was feeling lazy, and didn't want to bother typing, "The Alpha of the 21st Century". I ran Itanium-2s, on HP workstations, and there really was nothing like them. The first generation (the one derided as a failure out of the box), running Beta compilers from Intel, ran circles around our ultrasparc-III systems (not cheap midrange servers). The second generation did no worse than the Xeons of the era on integer-bound code, and walked away on floating point. It's an incredible chip.
However, Intel priced them to recoup their entire development cost on the first batch, and hired IBM's marketing team from OS/2, and here we are.
My lab used to have a cartoon taped on the wall, from the days of the hot new 386, "we finally figured out that what consumers want is not more speed or more memory, but just really great tailfins."
Has anyone tried putting tailfins and chrome from a '59 caddy on a Dell and see if it sells better.
Intel had a true, high-performance, 64-bit product out years ahead of AMD, and all you people out in Desktop-land went "EWWWWWW!!!". "it's too hot, too expensive, too hard to code, and it won't run Quake e^(pi)!". Intel's fault was believing the old IBM saying, "sometimes you have to drag the customer, kicking and screaming, into the future". Instead, AMD took what they already knew how to do (improve IA-32), bolted some reasonably-well thought out 64-bit extensions onto it, and sold it as a future-proofed Xeon. Intel hemmed and hawed, eventually gave up and did the same in a manner compatible with AMD, and this is where we are now; stuck with the x86 until the Sun grows cold.
I did the PhD right out of undergrad, but had spent over a year as a Co-op at a major chemical company. In the sciences, at least, it seems that my colleagues who'd been in industry (myself included) had the hardest time adjusting to being grad students. It wasn't just the paycheck (though I did think that I should have had a profit-sharing plan from the univeristy like my company did, given what they charged), but also the sense of purpose, and frequently the facilities. It was hard to get used to having instrumentation (touted as state of the art, world class in brochures), that my company would have dumpstered, and hard to go back to sitting in class, taking exams, etc. It's better once you get back in the lab and they leave you alone, but two years of classes, preliminaries, cumes, etc., is an aggravating process to go through. The other issue was, of course, the financial one. People with families tend to have an issue with being paid grad-student stipends, and working 12 to 14 hour days (because, sometimes you use the instrument when it's both available and running correctly, not when you feel like it).
Grad school is kind of like the military; it's best to join when you're too young to know better.
IBM demo'd an image searching system about 10 years ago, where you'd sketch the outlines of what you were looking for. Probably hasn't been refined enough yet.
As for music feeling "sticky", I'm baffled on that one. I've owned various media over 30++ years, (sheet, vinyl album, tape, CD, banjo), and sticky was never a desirable state of affairs. However, as long as there is tactile feedback available, then classical should feel like polished wood, modern orchestral (Bartok), like sheet-metal sculpture, Rock disordered with edges everywhere, and Barry Manilow like you're sinking in lukewarm jello while barely heated tofu is poured over your head.
http://www.rocksclusters.org/roll-documentation/vi z/4.1/
I'll admit, I've never tried this because I couldn't get anyone to spring for the displays, or the space to stack several monitors, but just follow their instructions, and you're off and running.
Thank you for the cut-and-paste reminder. While seemingly efficient (highlight then middle click), make one bump while moving the mouse and you have to do it over. I'd forgotten that one.
It's amazing how quickly you get used to not having to deal with such issues. Now let's hope His Steveness can keep it all on track a few more years.
No, seriously, if you're setting up a cluster where your work can be batch-queued, or intend to run MPI, then Rocks http://www.rocksclusters.org/ is the way to go. It also comes with tools such as SGE (Sun Grid Engine) or OpenPBS pre-configured, Intel compilers and libraries ready for you to drop a license onto (but of course the entire GNU suite is there as well, including Ada), has more monitoring tools (plus some nice web-based interfaces) than you can shake a stick at, and runs on IA-32/AMD-64/IA-64 (Itanium). It also has a Roll to help build a tiled display wall, which would be a really cool use of a small cluster.
They're also really great guys.
On the other hand, Oscar is supposed to be good, and if you're not into the whole batch-mode thing, you can get OpenMosix up and running using http://clusterknoppix.sw.be// ClusterKnoppix, and just fire jobs off into space and let them find their own unburdened node.
But still, Rocks is really an elegant and clean way to go, plus it will scale up in case you're going to deploy a huge one of these for real after you get your feet wet.
Actually, having gone the other way (Linux On The Desktop, (LOTD) 10 years to OSX), I'd be happy if the GUI/App people in the Linux community took a page from OSX, and actually spent some time worrying about polish, internal consistency, and interoperability. I've run (recently) RHEL 4.1, Ubuntu (which is a long way to making this rant irrelevant), SuSE (If you knew SuSE, like I knew SuSE...), and while they're much better than they were when I started, a little polish could still be applied. Surely some of you have graphic-arts oriented friends you could hand a six-pack of something decent to and say, "make these icons look pretty". Surely someone could agree, "you know, a hidden save dialogue activated with a right-mouse click, in addition to the identical one at the top of the window is just silly". An agreement that little buttons with pop-up menus that look like Win95/2K isn't the best interface to copy.
As for the rest, I write script files on my OSX system, run IBM's Fortran and C compilers to create command-line applications, and end up hand-twiddling my rc files to set things like SHMMAX which is just too small out of the box for real work. The entire difference to me is in the user-land experience. The Mac is just Clean, in a way that's hard to appreciate unless you come to it from too many years of CDE and CDE clones. The beauty is that I can do all that manual work if I want to (find mixed with grep is still often faster for me than Spotlight), but I don't have to. Voluntary complexity is a wonderful thing.
As one of my colleagues in High-Energy Physics put it, "A Mac is a Unix box that runs iTunes."
Given Sun's current financial state, 64-bit on AMD/Intel Solaris-x86, and technologies such as Java, it could easily be eaten by apple. Paying the premium to ship SnApples without having to deal with the screaming hordes of Linux geeks may be worth it for a company as tightly-wound and controlled as Apple. I doubt they want any confusion in people's minds over whether Aqua is their own, or just a shell on top of Gnome. In the end, does it matter, since they entire point of the Apple experience is in the UI and higher-level libs, which aren't open, and aren't about to become so. On the other hand, I, for one, welcome our new Efficient, Poly-cored, Multi-threaded, Preemptively Multitasking Overlords.
iPod Pico: 4GB storage, doubles as an eyebrow ring. This will be closely followed by the iPod Femto; same storage, but it's applied at the store with a roller as a temporary tattoo. It washes off when the warranty expires.
So, idle question: why don't they just license, pre-install, and use Tripwire on Windows boxes? (or reimplement it in-house, to avoid dealing with an outside company)
Fair enough, and I'll admit that I have one code that I often built using G77 for stability reasons. I am (or was) actually rather fond of G77, having run it originally on OS/2 using the EMX suite, and then used it as my primary development platform for years. However, given that Fortran has been a bit of a red-headed stepchild in the GCC suite, and the current F90 compiler is choking on a major F77 code I use on a daily basis, I have a hard time recommending it any more. I hope that GFortran improves enough that I can go back to using it, as it was nice having a consistent Fortran across various platforms for debugging. Sadly , I miss G77; I used to use it to build older codes that my colleagues would bring to me, that modern compilers such as IFort choked hard on. I probably just need to be patient with GFortran, since if it gets within a few percent of the Absoft compiler, then it's the most cost-effective 64-bit compiler for what is soon to become my legacy cluster. (Mac OS-X) I had purchased them on the mistaken assumption that XLF had a future, and for 32-bit code, I really could not be happier, unless some of the underlying algorithms in the packages were improved. Actually, I could be happier; the programs could also colllate the data and write up the results for me, but that's an AI problem, not a Fortran problem.
When running on PPC hardware or Power3 hardware (Mac G5, AIX, respectively), for quantum chemistry packages, G77 was capable of being up to 50% slower than XLF. At 30% slower, that's equilvalent to running your machines for 8 months of the year, versus 12. Since I run simulations that routinely take two months, that 30% is a big difference.
Because if done right, you'll use Windows as the host, because the Windows functionality you need will run sufficiently better, and Microsoft sells another license. Or, you run Linux as a the host, but think warm thoughts about Microsoft because they put money into Xen, buy another license, and Microsoft gets your money as well.
Basically, it's to capture mindshare, prevent datacenters from migrating their Windows systems to Linux hosts when they start running virtualization, and ensuring that, even if you are primarily a Linux-type, you're still paying Microsoft. When you're that big, every extra dollar counts towards proving to Wall Street that you're still growing. There's a lot of money to be made in Virtualization, by ensuring that your OS (Vista/Windows 2007-and-Counting Server) is the host everywhere, rather than the client. If you also sell client licenses for older versions of the OS, for people with apps that can't migrate, so much the better.
To the poster with their "there is only one C"; Horsefeathers. That there is Only One C explains why every C program I have is a twisty little maze of #ifdefs, often for minor differences in machine/compiler combinations, and as likely to break after a compiler upgrade as continue to build. It continues to exist only because it's suitable for assembly-language-type functions, such as OS kernels, which means it ships with every unix out there, and is taught to unsuspecting comp-sci students because it came free with the OS. It's amazing that after the problems of the past few years, including buffer-overrun exploits, memory leaks, etc, caused by C's unsafe, seat-of-the-OS memory handling, that people continue to defends its use as a general purpose language.
n loads/ /Ada or even SmallTalk http://www.squeak.org/ for new work.
The OP may be trying to be funny, but in fact a modern Fortran (F90/95) is easy to do string manipulation in, has dynamic memory, is type-safe, and is arguably a much higher-level language than C. (on the other hand, F90/95 bears a curious resemblence to Ada, but we'll let that slide). Memory aliasing isn't nearly the issue it is for C programs, which is why Fortran still out-performs C/C++ on any modern architecture, as long as you're using a competent compiler. (i.e. I don't want to hear about how slow your G77/GFortran programs are; they're great tools, but they're primarily Fortran Lints (flints)). Where Fortran fails is that is has the same name as a particularly onerous language from the mid-60s, known variously as FORTRAN-IV or FORTRAN66. If it was named differently, such as Cerulean, for instance, you'd complain about its verbose, Python-esque syntax, with Begin and End, rather than { and }, but you'd grudgingly learn it, write interfaces to libraries, start new projects in it, and treat it as an appropriate tool. Flame wars would break out here as the Cerulean programmers made fun of the Java-weenies, because their language is byte-code interpreted, and doesn't have matrix operations built in as primitives (how can you do quick 3d graphics without matrix ops). Because of the antique name, you picture a black and white room full of guys in white shirts with pocket-protectors and crewcuts, smoking while they load pounds of manila cards into the keypunch.
Back when we were in college we used to comment that C was a step backwards, as in terms of being a high-level language it ranked right up there with VAX assembler. Let it go. Use C++/Objective-C/Java if you're obsessed with line-noise syntax, or move on to Eiffel http://eiffelsoftware.origo.ethz.ch/index.php/Dow
>>uppose that I replace this 1.25GHz G4 Mac Mini ....According to you, it wouldn't be a computer.
It's not a computer. It's an iTunes Appliance that happens to also run Word. And in case you're thinking about it, no you can't have mine. I will defend that desk-saving, hearing-sparing, piece of productivity with a pointed stick if necessary.
I've done the music instead of coffee as a wake-up move before, and it seems to work well, without the acid stomach of too much coffee. Just be picky; Wagner/Bartok works better than Brahms/Chopin. New-age white-noise is right out.
If you drill through the FA (the PDF version), you'll see that aP2 is involved in lipid metabolism and transport, and probably synthesis of several derivatives. Unfortunately, when over-expressed (or over supplied with raw material), it may be linked to some auto-immune diseases. It's probably similar to the mechanism by which your body conserves calories and stores them as fat; a great idea in pre-modern or even pre-20th century societies, but with unfortunate side-effects in our modern, calorie-rich, environments.
Well, we're a school (a mid-sized university physical science department), and I can say that despite discussions concerning the options, we're certainly not upgrading anything to Linux. Win98 machines are being repurposed as Lab machines, with a parent Win98 OS image distributed (probably by Norton ghost, though I stay away from those details lest I get sucked in to support them) either once a week, or when a machine is suspected of being borked. Slightly newer machines which can handle XP are generally improved by stealing the memory from another comparable one to get it up as high as possible, then invoking our academic discount to buy an XP-Pro license. Machines too old for either of these solutions are simply being sent to recycling.
The linux machines exist in CompSci's cluster, running back-end services (DNS, Web, etc), and one in my lab for portability testing. The issue, as always, is that we have software that expects Windows-(something), and dedicating the resources to make it behave seamlessly under Wine is not cost effective versus continuing to run Windows. I suspect other educational institutions (based on having worked at them, and having seen the same policies) are doing the same thing.
The article should read, "Win98 EOL causes mass upgrades at school/govt/business, dumpster-divers scavenge abandoned machines and put desktop linux on them". It's a wonderful dream, but there is just too much Windows-centric software out there to see it happen.
So, did the superchickens have more dark meat on them, or light? Parts that get exercised (thighs) are darker than parts that don't (breast on a non-flying chicken). I can see this at Wegman's now, "boneless thighs from free-range, hormone-free, pre-centrifuged, SuperChickens".