I'm actually glad to hear that, because I was the unofficial OS/2 support person in my graduate department. I installed it, ported our Fortran codes, and demonstrated to people why multitasking was a great thing. I'm still miffed that it lost to Windows-???.
Over the years, I've enthused over VMS, OS/2, and SGI. I'd better not get attached to Linux.
You don't remember OS/2 Warp, do you? A multi-million dollar ad campaign, and people still didn't know what it was. It was competing with the obviously (technically, artistically, usably) inferior Windows 95, which had to be purchased and installed by the end-user as well, and still never built any significant mind-share. (except amongst those of us who had also championed VMS and similar big-iron OSes.) If you can find the old Dave Barry column, he joked about it reducing his somewhat functioning machine to a gibbering pile of junk that had to be shot. (Living in Miami, helpful passerby did this for him)
Consistency and usability are where it's at. You'd stand a better chance if the new IBM were to port Workplace Shell, slap it on top of a distro-of-choice, and then 'Macize' the overall experience. Unfortunately, that would also involve cleaning up a wide variety of hardware drivers (my recent foray into RHEL with a Radeon 9000 being case in point; the OpenGL graphics were very fast until the card locked solid), and being willing to tell some competing development projects, "that's nice, but not helpful or consistent, so we're not doing it your way". When people who abandoned Macs during the Gil Amelio days see my OS-X system, the comments are, "it's so clean". This is very similar to the people who saw me running OS/2 in the lab back in the early 90s, "it's so much more professionally done". Consistency and elegance matter, and lack of them makes people think that it looks like a research project, or their cousin Harold's home-built couch. Intellectually, it's all fun, but if you really care about being able to walk into BestBuy are find software, or having family members volunteer to have Linux installed on their own machines (rather than imposed in order to save your sanity during the holidays), then some of the ecosystem is going to have to be pruned in favor of a consistent and elegant experience. I have no personal stake in any particular desktop manager, but just realize from supporting end-users that nobody cares about the difference between desktop environments, they just want it to work, and work consistently.
You should really be asking yourself, "why do I care?" It's on servers everywhere, the back-end is where the money is, professionally, and there's an active community of specialists who do run it on the desktop. What if someone pulls the curtain off "Longhorn" or "Singularity" or "Mojo Rising", and you find an OpenBSD subsystem running a desktop environment that looks like the Windows that we've come to know and... tolerate..., with a classic-mode emulator for those apps that just don't get ported. Would you still push Linux on the desktop, or would you argue, "it's Unix-based, and runs on a variety of cheaply availble commodity hardware. Good enough."
Spielberg "saved" parts of it? The Spielberg parts are the maudlin, emotional, go out of the theatre with a warm glow, your brain be damned parts. Kubrick was fascinated by alienation, whether caused by technology, training, or personal obliviousness (try "Barry Lyndon" some time). Beautifully shot, if somewhat distant, movies.
His interpretation of "The Shining" left you the ambiguity whether Jack is having a break down, or whether there are really ghosts. He went for the long, slow, unsettlement of the audience, rather than the cheap and quick gross-out horror.
IMHO, the real problem is that our societies would stagnate. You may think that you're hip and forward thinking, just as I'm sure Francis Bacon did back in 1584 (though he probably was, "hippe and forwarde thinking"), but how will you be, comparitavely in 2385? Death doesn't just get the physically old out of the way, it also gets the old ideas. Sir Francis lived in a world of absolute monarchs where torturing and killing people over interpretation of the same scriptures was considered acceptable and desirable. Picture a world in which a man who amassed billions using 1970s technology in the 1990s, still had enough money to control computing in 2200. Picture Richard Nixon at 324, still with a secret plan to end the war on Proxima Centauri, the 600 year old Fuggers understanding money only in the sense of little disks of gold, or Benedict the 16th, worried in the 24th century that the church will have surrendered its soul if it accepts the social mores of 1974.
Personally, I'd be much happier if these people would work on a way for people to be healthy and capable (physically and mentally) up until 70 or 80, then still die. The 1000 year life span, besides being a nice round number, is about 10x our maximum current age. What's the point of outliving Methusaleh if the last several centuries are spent in a wheelchair (even if it is Luke Garner's flying one) ogling the cute, young, 300 year olds who sashay by and pat you on the head occasionally?
I cannot find the author, so insert someone suitably cynical, "all wish to live long, but none wishes to grow old". I think we should be working on improving the quality of our life, rather than extending it out of some fear of dying. Let's get cancer and heart-disease first, and deteriorating immune systems first, then worry about outliving Methuselah.
If you've checked your rural population and employment trends lately, you'd discover that in many of these places, you can't find a milk(maid/man/dude) to hire. Frankly, even amongst the underemployed, not many will work the hours a dairy farm runs at given what a farmer can afford to pay. In my parents area, northern Pennsylvania, there's actually an uptick of hispanic immigrants who come to work in the dairies, as it's otherwise almost impossible to find someone willing to get up at 4:00 a.m., milk cows, then repeat the process again at 4:00 p.m. Farmers with large enough herds to milk on the three-shift schedule have it even harder. The advantage of the machines are that they are reliable, consistent, and always on the job. If the cows are happier as a result, then that's a genuine added bonus.
(completely off topic, but I remember one cow we had that had injured herself, so that she had to be hosed off and have antibiotic spray applied a couple of times a day. After a couple of weeks of this routine, she became tame, and would sneak up behind you, put her head under your arm, and just about demand to be scratched behind the ears. So, I've seen the difference between a merely contented cow, and a genuinely happy one.)
Not to ruin anyone's agrarian fantasies, but my grandfather, who farmed all his life except for brief stints in the coal mines and on highway crew in the summer, said that he would have never gone into farming if it hadn't become mechanized. The work was bad enough that a summer spent shoveling (by hand) hot asphalt into holes in the highway for 8-10 hours a day was a vacation by comparison. He made sure that his kids got an education, kept their act together, and had a chance to get out.
Frankly, mechanisation of this type is what will allow small farmers (or at least smallish) to continue to exist. The problem with the OP's idea that those jobs should be left, so that people with lower IQ's have something to do, is that unmechanized farming will require far more than just their labor. Think carefully; do you want to go back to a system where fewer sysadmins are employed, but there are great opportunities in bovine lactate extraction? We did an informal poll at one of my previous jobs, and it turned out that at least 1/2 the sysadmins had some sort of agrarian upbringing. Maybe we were just all malcontents who didn't know how good we had it when we were cowherds, but somehow I doubt it.
Get cargo pants, then you can tell your boss you don't need a desk any more either, thereby saving the company money, getting yourself a raise, and still having room for all of your gadgets.
Personally, I'd start committing gadgetcide before I worried about carrying another bag around.
I'm afraid that I fail to see the issue. They advertised an X, you decided that you could afford/use an X, and when it gets shipped, you might get an X+.2, but at the minimum, you're guaranteed to get an X. It's not like they have a problem where they have either 1.5GHz procs or 800MHz procs, so that you could get either an X+.2 or an X/2 when you order an X. The phrase here is "meets or exceeds", as opposed to "averages out to...".
This is like the old days with underclocked processors, or cars not made on monday morning or friday afternoon, or doughnuts bought late in the day; you may have heard that you might get something better if you time your order right, but you're guaranteed at least what you ordered in the first place. The rest is just sour grapes and an overly-developed sense of entitlement.
Personally, I like Rocks, as I ran three parallel architectures (i386/AMD64/IA64), on the same based distribution, just with each tuned to their particular processor. Comes with SGE and Myrinet support out of the box, and there are Rolls, i.e. custom software assemblages, for OpenPBS, for those who prefer it, as well as PVFS. It's easy to set up, and easy to administer, as the nodes are presumed to be interchangeable and disposable. When you reboot a node, it's obliterated and a fresh OS and supplementary package repository are laid down on a clean disk. No questions about version skew.
They now have a custom roll to help you build a visualization wall, but I never had a chance to try that one. (try convincing your boss that you want 4 digital projectors and a big room to play with)
The downside to the above distributions are that they presume batch-queue environments, which is appropriate for most of my work, but less so for many people trying to simulate owning an SMP, without paying SMP prices.
Other people assure me that the current version of OSCAR is solid as well, but they seem to lag in the multiple architecture support area (Itanium is always behind), and don't current support AMD64 natively. On the other hand, they build on top of several RedHatish linuces, as opposed to Rocks where you get Centos (RHEL), period.
Quasi-rationally, leakage of PCBs from old transformers would be the only risk. It has been suggested previously that what little correlation there is to high-voltage lines and cancer has more to do with living in an industrial dump than the powerlines.
Look up the earth's electrical and magnetic field, add in the fields from the powerline, represent this as a percentage, then go find something else to worry about. You're more in risk of your computer falling on you.
Actually, I agree with you. I've moved my home, office, and research computing to Macs, and am amazed at the unobtrusiveness of the platform. My family may be next. For a while, I ran a medium-sized Linux shop (40 desktops, 3 clusters), with a Win2K3 Server controlling the XP boxes, but soon found that I was spending all of my tube-time on a G4-400 running 10.2. At that point, I admitted I'd already switched, and bought a modern laptop. I will not be going back, unless something radically better comes down the pipe.
However, pointing that out in a new discussion seemed to be an invitation to raving fanboy diatribes, and I just somehow didn't feel like starting another, "GNOME/XP is just as easy and convenient if you're smart enough" or "It's too expensive and the mouse is button-impaired" thread. Somehow, the obvious never sinks in around here.
The issue with the personal computer is that the current paradigm expects everyone to be a sysadmin. While similar to the Marines' "every man a rifleman" ethos, it works less well in the average Home/Office setting. Frankly, it leads to a lot of shot feet. "all right Bob, now flash the Bios... *BANG*"
When people say they're sick of the their PC, what they actually mean (from talking to a few of them), is that they're sick of having to worry about the balky innards. They just want to turn it on, write their letters, check out CNN, and play Hearts against the Novosibirsk Hearts League. However, if you ask them if they'd trade the speed, immediacy, and appearance of control that having their own PC versus a running a web-service on a dedicated, limited, device offers, they'll immediately say, "No". They also, as a rule, don't want eight devices each of which only does one job. So, we're back with PCs.
One suspects that what Zander is really offering is everyone having a SunRay on their desk, with massive Sun systems in the background pushing everything through the network pipe. I, as the de-facto sysadmin for the family, think this is a great idea, but I as my geekish self, don't. Personally, I think the first company/organization that comes up with a machine that includes the modern connectivity with the single-user OS experience of circa 1996 Mac/Windows is going to have a hit. It's finding someone to work out the iPod experience for the PC; connected, yet truly yours. Clean, unobtrusive, and dedicated to its function. Maybe everything that makes a PC yours kept on an iPod-Nanoish device, which is docked to a PC, and allows it to run. Without your card, it doesn't run, and with your card, it only runs your programs, and only stores your data, so other users can't infect you. Every tub on its own bottom computing.
On the other hand, maybe we'll finally get fibre to the curb, high-speed, redundant links to the network, so you'll always be on, and there's enough bandwidth so that remote content appears like local content. Then Zander, Gates, et al., will be proved right, but until then, I think the general-purpose PC is here to stay.
I think that makes them the elected leaders of anarchistic, independent, communes. You see, watery tarts lying around in shrink-wrap handing out device drivers are no basis for a system of goverment.
Can this be used for information compression in any way? After all, it was discovered about 20 years ago that simple fractal equations gave shapes very much like ferns. This could give you a shorthand way of compressing the genome of an organism, then making comparisons.
It would also, of course, be interesting if you could use this to work backwards through the genome to a set point, and (hypothetically) bring back the Auroch.
Personally, I want to see how this deals with metal incorporation at the active site, and whether their selection rules work for that as well.
Since nobody else has brought this up (maybe his readership has dropped to zero since the 80s), does this remind anyone else of Larry Niven's early Know Space stories where transplantation is easy, so the sources of 'donors' have become easy as well?
I'd worry less about seeing your late wife's face on someone else than I would, say, Jeffrey Dahmer's. After all, it's not like he wasn't a danger to society, committed unspeakable crimes, and should now 'give' something back (voluntarily or not). The chinese are already rumored to be using vital organs from executed prisoners, so this makes one wonder how much longer before we start down the same path.
No, sometimes what matters is binary compatibility. It's the ability to compile once in a controlled environment, and continue to use the package until you *need* to recompile it. It's the ability to compile something on you 4GB core machine, and turn all of the optimizations on, then run the resulting binary in some smaller environment that would never support such an effort. It's the ability to use one specfic compiler with a known set of bugs, without worrying about whether that compiler will behave properly on everyone else's systems as well.
I've had systems where some critical library wasn't updated to the latest GNU compliance, and therefore I couldn't build the app on the new system until that was resolved. Thankfully, RH 8.0 builds ran on RHEL 3.0 and SuSE 8.2. This allowed me to move the other portions of the system forward (better desktop support), while continuing to use the application, without taking a month of my life to update a library that wouldn't build any later than RH 8.0, or the app that depended on it. The app in question was well-tested with a certain combination of libraries and compilers, and introducing potential instabilities by upgrading all of the components wasn't worth it at the time.
Binary compatibility prevents you from ending up on an unending cycle of having to upgrade, downgrade, or maintain multiple versions of packages, just to ensure the apps you use keep working. I have some carefully built, statically linked, programs from 7 years ago that I can still fire up with some care. Users of Suns and AIX take this ability for granted, as do MIPS-based SGI. Only PC-based users seem to think that, "oops, I bought a new machine; better buy again/recompile all my apps!" is an acceptable practice.
Early 1990s, killer micros were everywhere, the Jollix BSD port was being detailed in Dr. Dobbs, and some of the geekier (defined as: people who could hose their machine for days on end without worrying about work not getting done) techs were playing with some new creation called, "Linux". It was unstable, wonky, with wierd command-line tools named after the original authors and a moded 1970s text editor.
On the other hand, Windows was breaking out, Macs had gone color, and VMS was ported to the new, blazing, Alpha processor. As friends put it, "Unix has had 20 years, hasn't caught on, so it's time to let it die in peace."
FF to 2005, and I'm sitting here on a friendly Unix (MacOS X, a far sight from the 3b2-300 I first met it on), typing into a system running on a semi-civil, but needing another cup of coffee and some etiquette training Unix, while variants of the underpinnings of both run everything from stock exchanges to toasters.
If BeOS does *something* notably better than the alternatives, there will be early adopters, who will embed it in some specialized niche. That niche will become vital, people will take it home, show it to friends, and it will spread.
Given that it's BeOS, which I ran briefly around 1997, I have very little hope of that happening. I'd be willing to put more money on Plan9 or Hurd catching on first, but then I've backed the wrong digital horse before. Maybe it's what people are looking for; a locked-down single user OS that's very fast for them, but doesn't allow anyone else in. Maybe it's time to put the Personal back into Personal computing, and stop trying to make home users run a multi-user mainframe os that's been stuffed into a single-user box. But I'll bet that experience comes out of Cupertino, UIUC, or even Redmond before it comes from the wreckage of Be.
Don't forget disk access issues as well. You now have file locking, non-local disk-access, and race state issues to contend with.
Example from my work is that we tend to write several hundred meg to several gig scratch files, and then perform RW operations on them continually during a calculation. If the disk isn't local to the process, then you end up flooding the network, and bringing everything to a screeching halt.
In a Mosixish/Condor type environment, you then have to deal with which processes, because of this disk limitation, can be migrated to other CPUs, or can allow a second job to start on their own because of insufficient utilization, from those which have to have exclusive access to the CPU, and near-exclusive access to the disk, in order to prevent the calc from bogging down.
Then, as the parent mentioned, you have the CPU-CPU communication issues, the network overhead, and memory access patterns, all of which are hard. In theory, had you written your code correctly in the first place, this would only be moderately annoying, but since most people's applications are single-threaded, most programming is taught in serial mode, and the tools for MPar work are still expensive and exotic, then you get a situation where it's easy to run a compute farm (massive numbers of single-processor jobs), but hard to run a parallel cluster (one job aggregating resources)
I miss the Real film. Agfapan 25, Panatomic-X, Kodachrome 25, and Ektalure G. (ok, so the last was a paper). For speed and grain, there was 2475 recording film in B&W, or 3M 640T in slide. Just wild stuff; looked like you'd made a picture with colored sand.
Of course, Delta-100/TMax-100 are probably as fine-grained as any of the above, and the Kodachrome was always a pain to print, but I have some nice negatives from those days, and to mis-quote Tom "Son of MonkeyBoy" Cruise, "I feel... a need... for No Speed!".
The HP, because with RPN it made it practically borrow-proof. Just watching people try to use it, and get expressions like they'd blown a neuron was worth the experience.
While I may not miss having to use them, there was a certain romance to the old DEC TU-78 tape drive and LP-26 line-printers. The whirring sounds of the tape loading, then the soft "pfffff" as it vented at the end, and the hammering and flapping paper sounds of the line-printer made it seem like you were actually doing something. Too bad it was frequently creating Line-printer art, or printing Life patterns.
I also kind of miss analogue mass-spectrometers, because it was easier to see the metastable peaks, but that's more of a specialized taste. I do not miss Disco, Bell-bottoms, or Donny-Osmond hair.
The great value of this news is that Intel's First-rate Fortran and Math libraries will be available for the Mactels as soon as they ship. This will ease the transition for scientists, as their codes can be rebuilt and tested on the early Macs, before the replacements for the PowerMacs and XServes ship in '07.
No, it's not XLF, which I would have preferred, but IFort is capable of producing screamingly fast code that generates correct results (generally; there are always the first couple releases of a new version that seem to have issues). It's also already a standard, so the porting from Linux-x86 to Mac-x86 will go more quickly.
Does anyone know if Pathscale plans to ports its compiler suite to OS-X or not? At the moment, I believe that they lead in the X86-64 world, having started with the SGI Open64 compilers, which were based on Cray technology. It would be nice to see them as an alternative as well.
Of course, the G5 Mac could have its OS compiled with IBM's XLC, and the race would be on.
That's the depressing thing about this change. ICC/IFORT are capable compilers and all, but IBM's compilers ported to OS-X were in a class by themselves. Extremely fast, extremely fast, extremely stable, and compatible with the flags from the AIX versions.
That's what cross-platform means, right? Runs on everything from Macs to AIX-based mainframes.
To paraphrase a Calvin and Hobbes cartoon, "does exposure to violence desensitize us, sure, does it increase our tolerance and lower expectations, sure, but does it *cause* it, well, that's not clear." it's all in the questions you ask.
I knew people in grad school who played Doom, and nobody shot their advisor. Before that, Moria in undergrad, and the number of my classmates who decided it was ok to whack people over the head and go through their pockets doesn't seem to have increased over baseline noise. On the other hand, nobody is asking, "today in Miami there were two carjackings, 29 muggings, and a random murder. Do you still care?"
Ideally, you'd pick a reasonably sized sample of individuals who don't currently play GTA, make half of them spend an hour or two a day at it, and take samples (i.e. ask questions and do blood tests) of how many started having more violent thoughts afterwards, or engaged in antisocial behaviour. A study done a few years back indicated that minor social slights raised adrenaline levels for southerners more than new englanders. This was done with Med students, and the ones from the northeast would barely be perturbed by being bumped, while the southern ones would have a more aggressive response. Something like that should be done in this case, so that we have some hard evidence, and a baseline to compare to.
Btw, in answer to the Bill O'Reilly question, a strictly non-scientific sample of Fox-viewing coworkers would indicate, "yes".
I've spent years (not full time) worrying about this possibility. If it's cheap and easy to obtain, imagine the problem of 'clotheslining' with ultra-high tensile monofilament across a highway or jogging path. Garrotting becomes trivial, and various sorts of passive terrorism as well.
Frankly, I hope that when it's sold, it's either in lengths too short to be used nefariously, or in a wound rope/twine/string that can't be unraveled. Remember, it's all fun and games until somebody loses a torso.
Slashdot Mirror
I'm actually glad to hear that, because I was the unofficial OS/2 support person in my graduate department. I installed it, ported our Fortran codes, and demonstrated to people why multitasking was a great thing. I'm still miffed that it lost to Windows-???.
Over the years, I've enthused over VMS, OS/2, and SGI. I'd better not get attached to Linux.
You don't remember OS/2 Warp, do you? A multi-million dollar ad campaign, and people still didn't know what it was. It was competing with the obviously (technically, artistically, usably) inferior Windows 95, which had to be purchased and installed by the end-user as well, and still never built any significant mind-share. (except amongst those of us who had also championed VMS and similar big-iron OSes.) If you can find the old Dave Barry column, he joked about it reducing his somewhat functioning machine to a gibbering pile of junk that had to be shot. (Living in Miami, helpful passerby did this for him)
... tolerate..., with a classic-mode emulator for those apps that just don't get ported. Would you still push Linux on the desktop, or would you argue, "it's Unix-based, and runs on a variety of cheaply availble commodity hardware. Good enough."
Consistency and usability are where it's at. You'd stand a better chance if the new IBM were to port Workplace Shell, slap it on top of a distro-of-choice, and then 'Macize' the overall experience. Unfortunately, that would also involve cleaning up a wide variety of hardware drivers (my recent foray into RHEL with a Radeon 9000 being case in point; the OpenGL graphics were very fast until the card locked solid), and being willing to tell some competing development projects, "that's nice, but not helpful or consistent, so we're not doing it your way". When people who abandoned Macs during the Gil Amelio days see my OS-X system, the comments are, "it's so clean". This is very similar to the people who saw me running OS/2 in the lab back in the early 90s, "it's so much more professionally done". Consistency and elegance matter, and lack of them makes people think that it looks like a research project, or their cousin Harold's home-built couch. Intellectually, it's all fun, but if you really care about being able to walk into BestBuy are find software, or having family members volunteer to have Linux installed on their own machines (rather than imposed in order to save your sanity during the holidays), then some of the ecosystem is going to have to be pruned in favor of a consistent and elegant experience. I have no personal stake in any particular desktop manager, but just realize from supporting end-users that nobody cares about the difference between desktop environments, they just want it to work, and work consistently.
You should really be asking yourself, "why do I care?" It's on servers everywhere, the back-end is where the money is, professionally, and there's an active community of specialists who do run it on the desktop. What if someone pulls the curtain off "Longhorn" or "Singularity" or "Mojo Rising", and you find an OpenBSD subsystem running a desktop environment that looks like the Windows that we've come to know and
Just wait until you see the iNavy and iMarines posters next year. One button mouse, 24 iCBMs, It Just Works(tm)!
Spielberg "saved" parts of it? The Spielberg parts are the maudlin, emotional, go out of the theatre with a warm glow, your brain be damned parts. Kubrick was fascinated by alienation, whether caused by technology, training, or personal obliviousness (try "Barry Lyndon" some time). Beautifully shot, if somewhat distant, movies.
His interpretation of "The Shining" left you the ambiguity whether Jack is having a break down, or whether there are really ghosts. He went for the long, slow, unsettlement of the audience, rather than the cheap and quick gross-out horror.
IMHO, the real problem is that our societies would stagnate. You may think that you're hip and forward thinking, just as I'm sure Francis Bacon did back in 1584 (though he probably was, "hippe and forwarde thinking"), but how will you be, comparitavely in 2385? Death doesn't just get the physically old out of the way, it also gets the old ideas. Sir Francis lived in a world of absolute monarchs where torturing and killing people over interpretation of the same scriptures was considered acceptable and desirable. Picture a world in which a man who amassed billions using 1970s technology in the 1990s, still had enough money to control computing in 2200. Picture Richard Nixon at 324, still with a secret plan to end the war on Proxima Centauri, the 600 year old Fuggers understanding money only in the sense of little disks of gold, or Benedict the 16th, worried in the 24th century that the church will have surrendered its soul if it accepts the social mores of 1974.
Personally, I'd be much happier if these people would work on a way for people to be healthy and capable (physically and mentally) up until 70 or 80, then still die. The 1000 year life span, besides being a nice round number, is about 10x our maximum current age. What's the point of outliving Methusaleh if the last several centuries are spent in a wheelchair (even if it is Luke Garner's flying one) ogling the cute, young, 300 year olds who sashay by and pat you on the head occasionally?
I cannot find the author, so insert someone suitably cynical, "all wish to live long, but none wishes to grow old". I think we should be working on improving the quality of our life, rather than extending it out of some fear of dying. Let's get cancer and heart-disease first, and deteriorating immune systems first, then worry about outliving Methuselah.
If you've checked your rural population and employment trends lately, you'd discover that in many of these places, you can't find a milk(maid/man/dude) to hire. Frankly, even amongst the underemployed, not many will work the hours a dairy farm runs at given what a farmer can afford to pay. In my parents area, northern Pennsylvania, there's actually an uptick of hispanic immigrants who come to work in the dairies, as it's otherwise almost impossible to find someone willing to get up at 4:00 a.m., milk cows, then repeat the process again at 4:00 p.m. Farmers with large enough herds to milk on the three-shift schedule have it even harder. The advantage of the machines are that they are reliable, consistent, and always on the job. If the cows are happier as a result, then that's a genuine added bonus.
(completely off topic, but I remember one cow we had that had injured herself, so that she had to be hosed off and have antibiotic spray applied a couple of times a day. After a couple of weeks of this routine, she became tame, and would sneak up behind you, put her head under your arm, and just about demand to be scratched behind the ears. So, I've seen the difference between a merely contented cow, and a genuinely happy one.)
Not to ruin anyone's agrarian fantasies, but my grandfather, who farmed all his life except for brief stints in the coal mines and on highway crew in the summer, said that he would have never gone into farming if it hadn't become mechanized. The work was bad enough that a summer spent shoveling (by hand) hot asphalt into holes in the highway for 8-10 hours a day was a vacation by comparison. He made sure that his kids got an education, kept their act together, and had a chance to get out.
Frankly, mechanisation of this type is what will allow small farmers (or at least smallish) to continue to exist. The problem with the OP's idea that those jobs should be left, so that people with lower IQ's have something to do, is that unmechanized farming will require far more than just their labor. Think carefully; do you want to go back to a system where fewer sysadmins are employed, but there are great opportunities in bovine lactate extraction? We did an informal poll at one of my previous jobs, and it turned out that at least 1/2 the sysadmins had some sort of agrarian upbringing. Maybe we were just all malcontents who didn't know how good we had it when we were cowherds, but somehow I doubt it.
Machines should Work, People should Think.
Get cargo pants, then you can tell your boss you don't need a desk any more either, thereby saving the company money, getting yourself a raise, and still having room for all of your gadgets.
Personally, I'd start committing gadgetcide before I worried about carrying another bag around.
I'm afraid that I fail to see the issue. They advertised an X, you decided that you could afford/use an X, and when it gets shipped, you might get an X+.2, but at the minimum, you're guaranteed to get an X. It's not like they have a problem where they have either 1.5GHz procs or 800MHz procs, so that you could get either an X+.2 or an X/2 when you order an X. The phrase here is "meets or exceeds", as opposed to "averages out to...".
This is like the old days with underclocked processors, or cars not made on monday morning or friday afternoon, or doughnuts bought late in the day; you may have heard that you might get something better if you time your order right, but you're guaranteed at least what you ordered in the first place. The rest is just sour grapes and an overly-developed sense of entitlement.
Which is good, but then the users appear, which isn't so good.
But their links could at least have mentioned OSCAR http://oscar.openclustergroup.org/ or my personal favorite, ROCKS http://www.rocksclusters.org/, as these are more prevalent than xCat systems.
Personally, I like Rocks, as I ran three parallel architectures (i386/AMD64/IA64), on the same based distribution, just with each tuned to their particular processor. Comes with SGE and Myrinet support out of the box, and there are Rolls, i.e. custom software assemblages, for OpenPBS, for those who prefer it, as well as PVFS. It's easy to set up, and easy to administer, as the nodes are presumed to be interchangeable and disposable. When you reboot a node, it's obliterated and a fresh OS and supplementary package repository are laid down on a clean disk. No questions about version skew.
They now have a custom roll to help you build a visualization wall, but I never had a chance to try that one. (try convincing your boss that you want 4 digital projectors and a big room to play with)
The downside to the above distributions are that they presume batch-queue environments, which is appropriate for most of my work, but less so for many people trying to simulate owning an SMP, without paying SMP prices.
Other people assure me that the current version of OSCAR is solid as well, but they seem to lag in the multiple architecture support area (Itanium is always behind), and don't current support AMD64 natively. On the other hand, they build on top of several RedHatish linuces, as opposed to Rocks where you get Centos (RHEL), period.
Quasi-rationally, leakage of PCBs from old transformers would be the only risk. It has been suggested previously that what little correlation there is to high-voltage lines and cancer has more to do with living in an industrial dump than the powerlines.
Look up the earth's electrical and magnetic field, add in the fields from the powerline, represent this as a percentage, then go find something else to worry about. You're more in risk of your computer falling on you.
Actually, I agree with you. I've moved my home, office, and research computing to Macs, and am amazed at the unobtrusiveness of the platform. My family may be next. For a while, I ran a medium-sized Linux shop (40 desktops, 3 clusters), with a Win2K3 Server controlling the XP boxes, but soon found that I was spending all of my tube-time on a G4-400 running 10.2. At that point, I admitted I'd already switched, and bought a modern laptop. I will not be going back, unless something radically better comes down the pipe.
However, pointing that out in a new discussion seemed to be an invitation to raving fanboy diatribes, and I just somehow didn't feel like starting another, "GNOME/XP is just as easy and convenient if you're smart enough" or "It's too expensive and the mouse is button-impaired" thread. Somehow, the obvious never sinks in around here.
So, Mac-on, dude. They'll catch up someday.
The issue with the personal computer is that the current paradigm expects everyone to be a sysadmin. While similar to the Marines' "every man a rifleman" ethos, it works less well in the average Home/Office setting. Frankly, it leads to a lot of shot feet. "all right Bob, now flash the Bios... *BANG*"
When people say they're sick of the their PC, what they actually mean (from talking to a few of them), is that they're sick of having to worry about the balky innards. They just want to turn it on, write their letters, check out CNN, and play Hearts against the Novosibirsk Hearts League. However, if you ask them if they'd trade the speed, immediacy, and appearance of control that having their own PC versus a running a web-service on a dedicated, limited, device offers, they'll immediately say, "No". They also, as a rule, don't want eight devices each of which only does one job. So, we're back with PCs.
One suspects that what Zander is really offering is everyone having a SunRay on their desk, with massive Sun systems in the background pushing everything through the network pipe. I, as the de-facto sysadmin for the family, think this is a great idea, but I as my geekish self, don't. Personally, I think the first company/organization that comes up with a machine that includes the modern connectivity with the single-user OS experience of circa 1996 Mac/Windows is going to have a hit. It's finding someone to work out the iPod experience for the PC; connected, yet truly yours. Clean, unobtrusive, and dedicated to its function. Maybe everything that makes a PC yours kept on an iPod-Nanoish device, which is docked to a PC, and allows it to run. Without your card, it doesn't run, and with your card, it only runs your programs, and only stores your data, so other users can't infect you. Every tub on its own bottom computing.
On the other hand, maybe we'll finally get fibre to the curb, high-speed, redundant links to the network, so you'll always be on, and there's enough bandwidth so that remote content appears like local content. Then Zander, Gates, et al., will be proved right, but until then, I think the general-purpose PC is here to stay.
I think that makes them the elected leaders of anarchistic, independent, communes. You see, watery tarts lying around in shrink-wrap handing out device drivers are no basis for a system of goverment.
Can this be used for information compression in any way? After all, it was discovered about 20 years ago that simple fractal equations gave shapes very much like ferns. This could give you a shorthand way of compressing the genome of an organism, then making comparisons.
It would also, of course, be interesting if you could use this to work backwards through the genome to a set point, and (hypothetically) bring back the Auroch.
Personally, I want to see how this deals with metal incorporation at the active site, and whether their selection rules work for that as well.
Since nobody else has brought this up (maybe his readership has dropped to zero since the 80s), does this remind anyone else of Larry Niven's early Know Space stories where transplantation is easy, so the sources of 'donors' have become easy as well?
I'd worry less about seeing your late wife's face on someone else than I would, say, Jeffrey Dahmer's. After all, it's not like he wasn't a danger to society, committed unspeakable crimes, and should now 'give' something back (voluntarily or not). The chinese are already rumored to be using vital organs from executed prisoners, so this makes one wonder how much longer before we start down the same path.
No, sometimes what matters is binary compatibility. It's the ability to compile once in a controlled environment, and continue to use the package until you *need* to recompile it. It's the ability to compile something on you 4GB core machine, and turn all of the optimizations on, then run the resulting binary in some smaller environment that would never support such an effort. It's the ability to use one specfic compiler with a known set of bugs, without worrying about whether that compiler will behave properly on everyone else's systems as well.
I've had systems where some critical library wasn't updated to the latest GNU compliance, and therefore I couldn't build the app on the new system until that was resolved. Thankfully, RH 8.0 builds ran on RHEL 3.0 and SuSE 8.2. This allowed me to move the other portions of the system forward (better desktop support), while continuing to use the application, without taking a month of my life to update a library that wouldn't build any later than RH 8.0, or the app that depended on it. The app in question was well-tested with a certain combination of libraries and compilers, and introducing potential instabilities by upgrading all of the components wasn't worth it at the time.
Binary compatibility prevents you from ending up on an unending cycle of having to upgrade, downgrade, or maintain multiple versions of packages, just to ensure the apps you use keep working. I have some carefully built, statically linked, programs from 7 years ago that I can still fire up with some care. Users of Suns and AIX take this ability for granted, as do MIPS-based SGI. Only PC-based users seem to think that, "oops, I bought a new machine; better buy again/recompile all my apps!" is an acceptable practice.
Sure; Unix.
Early 1990s, killer micros were everywhere, the Jollix BSD port was being detailed in Dr. Dobbs, and some of the geekier (defined as: people who could hose their machine for days on end without worrying about work not getting done) techs were playing with some new creation called, "Linux". It was unstable, wonky, with wierd command-line tools named after the original authors and a moded 1970s text editor.
On the other hand, Windows was breaking out, Macs had gone color, and VMS was ported to the new, blazing, Alpha processor. As friends put it, "Unix has had 20 years, hasn't caught on, so it's time to let it die in peace."
FF to 2005, and I'm sitting here on a friendly Unix (MacOS X, a far sight from the 3b2-300 I first met it on), typing into a system running on a semi-civil, but needing another cup of coffee and some etiquette training Unix, while variants of the underpinnings of both run everything from stock exchanges to toasters.
If BeOS does *something* notably better than the alternatives, there will be early adopters, who will embed it in some specialized niche. That niche will become vital, people will take it home, show it to friends, and it will spread.
Given that it's BeOS, which I ran briefly around 1997, I have very little hope of that happening. I'd be willing to put more money on Plan9 or Hurd catching on first, but then I've backed the wrong digital horse before. Maybe it's what people are looking for; a locked-down single user OS that's very fast for them, but doesn't allow anyone else in. Maybe it's time to put the Personal back into Personal computing, and stop trying to make home users run a multi-user mainframe os that's been stuffed into a single-user box. But I'll bet that experience comes out of Cupertino, UIUC, or even Redmond before it comes from the wreckage of Be.
Don't forget disk access issues as well. You now have file locking, non-local disk-access, and race state issues to contend with.
Example from my work is that we tend to write several hundred meg to several gig scratch files, and then perform RW operations on them continually during a calculation. If the disk isn't local to the process, then you end up flooding the network, and bringing everything to a screeching halt.
In a Mosixish/Condor type environment, you then have to deal with which processes, because of this disk limitation, can be migrated to other CPUs, or can allow a second job to start on their own because of insufficient utilization, from those which have to have exclusive access to the CPU, and near-exclusive access to the disk, in order to prevent the calc from bogging down.
Then, as the parent mentioned, you have the CPU-CPU communication issues, the network overhead, and memory access patterns, all of which are hard. In theory, had you written your code correctly in the first place, this would only be moderately annoying, but since most people's applications are single-threaded, most programming is taught in serial mode, and the tools for MPar work are still expensive and exotic, then you get a situation where it's easy to run a compute farm (massive numbers of single-processor jobs), but hard to run a parallel cluster (one job aggregating resources)
I miss the Real film. Agfapan 25, Panatomic-X, Kodachrome 25, and Ektalure G. (ok, so the last was a paper). For speed and grain, there was 2475 recording film in B&W, or 3M 640T in slide. Just wild stuff; looked like you'd made a picture with colored sand.
... a need... for No Speed!".
Of course, Delta-100/TMax-100 are probably as fine-grained as any of the above, and the Kodachrome was always a pain to print, but I have some nice negatives from those days, and to mis-quote Tom "Son of MonkeyBoy" Cruise, "I feel
The HP, because with RPN it made it practically borrow-proof. Just watching people try to use it, and get expressions like they'd blown a neuron was worth the experience.
While I may not miss having to use them, there was a certain romance to the old DEC TU-78 tape drive and LP-26 line-printers. The whirring sounds of the tape loading, then the soft "pfffff" as it vented at the end, and the hammering and flapping paper sounds of the line-printer made it seem like you were actually doing something. Too bad it was frequently creating Line-printer art, or printing Life patterns.
I also kind of miss analogue mass-spectrometers, because it was easier to see the metastable peaks, but that's more of a specialized taste. I do not miss Disco, Bell-bottoms, or Donny-Osmond hair.
The great value of this news is that Intel's First-rate Fortran and Math libraries will be available for the Mactels as soon as they ship. This will ease the transition for scientists, as their codes can be rebuilt and tested on the early Macs, before the replacements for the PowerMacs and XServes ship in '07.
No, it's not XLF, which I would have preferred, but IFort is capable of producing screamingly fast code that generates correct results (generally; there are always the first couple releases of a new version that seem to have issues). It's also already a standard, so the porting from Linux-x86 to Mac-x86 will go more quickly.
Does anyone know if Pathscale plans to ports its compiler suite to OS-X or not? At the moment, I believe that they lead in the X86-64 world, having started with the SGI Open64 compilers, which were based on Cray technology. It would be nice to see them as an alternative as well.
Of course, the G5 Mac could have its OS compiled with IBM's XLC, and the race would be on.
That's the depressing thing about this change. ICC/IFORT are capable compilers and all, but IBM's compilers ported to OS-X were in a class by themselves. Extremely fast, extremely fast, extremely stable, and compatible with the flags from the AIX versions.
That's what cross-platform means, right? Runs on everything from Macs to AIX-based mainframes.
To paraphrase a Calvin and Hobbes cartoon, "does exposure to violence desensitize us, sure, does it increase our tolerance and lower expectations, sure, but does it *cause* it, well, that's not clear." it's all in the questions you ask.
I knew people in grad school who played Doom, and nobody shot their advisor. Before that, Moria in undergrad, and the number of my classmates who decided it was ok to whack people over the head and go through their pockets doesn't seem to have increased over baseline noise. On the other hand, nobody is asking, "today in Miami there were two carjackings, 29 muggings, and a random murder. Do you still care?"
Ideally, you'd pick a reasonably sized sample of individuals who don't currently play GTA, make half of them spend an hour or two a day at it, and take samples (i.e. ask questions and do blood tests) of how many started having more violent thoughts afterwards, or engaged in antisocial behaviour. A study done a few years back indicated that minor social slights raised adrenaline levels for southerners more than new englanders. This was done with Med students, and the ones from the northeast would barely be perturbed by being bumped, while the southern ones would have a more aggressive response. Something like that should be done in this case, so that we have some hard evidence, and a baseline to compare to.
Btw, in answer to the Bill O'Reilly question, a strictly non-scientific sample of Fox-viewing coworkers would indicate, "yes".
I've spent years (not full time) worrying about this possibility. If it's cheap and easy to obtain, imagine the problem of 'clotheslining' with ultra-high tensile monofilament across a highway or jogging path. Garrotting becomes trivial, and various sorts of passive terrorism as well.
Frankly, I hope that when it's sold, it's either in lengths too short to be used nefariously, or in a wound rope/twine/string that can't be unraveled. Remember, it's all fun and games until somebody loses a torso.