That's incorrect, at least in part. Modern MacOS is based on CMU's Mach, which has had lightweight threading support since long before Apple got into the picture. The OS was completely designed for multiple CPUs, down to the very core.
If modern MacOS apps are not heavily multithreaded (I have no idea, I don't run priorietary hardware anymore, regardless of the OS), that's the fault of programmers not advancing past the days of MacOS 9... it has nothing whatsoever to do with the OS.
Multithreading is a system-level thing, not a language level thing.
Sure, there have been languages that make threading ubiquitous, but they've never caught on, and it's hardly necessary.
You'll notice that internet, graphics, and many other programming necessities are not built into C/C++ either. They are higher level functions, and thousands of programmers have no problem understanding C's role here. People have been writing multithreading code in C/C++ for decades... I've personally done in from the 80s until now, under a dozen or so OSs.
Don't use your chosen language as a crutch for sicking to the level of programming practiced when that langauge debuted. The whole point of C was not to define much of anything in C itself.. in truth, the language proper doesn't even do I/O... that's handled via a library. So is threading, so is graphics, etc.
The idea of an OS and/or suppoet tools handling the SMP problem is nothing more than a crutch for bad programming.
In fact, anyone who grew up with a real multitheaded, multitasking OS is already writing code that will scale just dandy to 8 cores and beyond. When you accept that a thread is nothing more or less than a typical programming construct, you simply write better code. This is no more or less an amazing thing than when regular programmers embraced subroutines or structures.
This was S.O.P. back in the late 80s under the AmigaOS, and enhanced in the early/mid 90s under BeOS. This in not new, and not even remotely tied to the advent of multicore CPUs.
The problem here is simple: UNIX and Windows. Windows had fake multitasking for so long, Windows programmers barely knew what you could do when you had "thread" in the same toolkit as "subroutine", rather than it being something exotic. UNIX, as a whole, didn't even have lightweight preemptive threads until fairly recently, and UNIX programmers are only slowly catching up.
However, neither of these is even slightly an OS problem... it's an application-level problem. If programmers continue to code as if they had a 70s-vintage OS, they're going to think in single threads and suck on 8-core CPUs. If programmers update themselves to state-of-the-1980s thinking, they'll scale to 8-cores and well beyond.
I wrote a version of LISP back in the early 80s that included both a NIL and an ERROR atom. Any undefined or illegal operation on valid arguments would result in the ERROR atom; subsequent operations on the error atom were all defined to not break, but also return the ERROR atom. A list of illegal operations would grow, attached to ERROR.
This worked fairly well, since error handling could be done a useful points, it didn't have to be done everywhere. Not as powerful as true exceptions, of course, but much easier than some exception models (they weren't as well known or defined in '83 as they are today, either).
While Tony Hoare may not have lifted NULL from LISP, LISP 1.5 had its own version, NIL, in 1960 or thereabouts... NIL was even represented as 0, internally, at least in early versions of LISP.
While less likely to cause problems than in a language like Algol or C, I do believe this is the earliest use of "null" in a programming language (at least until some cheeky joker out there finds an earlier one I didn't know about).
Well, funny thing... before I bought my HP dv9500, I did precisely this... I compared it, feature by feature, to the 17" Powerbook. Both were direct from manufacturer based on customizable web site builds. Of the things I cared about, the HP exceeded the Mac in every area except one: the Mac has both Fw800 and Fw400 ports, while the HP just the Fw400 (on 4-pin, sure, but it's for my camcorders, so that's fine).
The result: the Powerbook was $2999, the HP was $1280. Not even a contest -- Apple goes down in flames. This was done in December of 2007... I can't honestly claim that Apple hasn't halved their prices since then, but this "spec for spec" claim rarely if ever works outside the bounds of a few individuals' reality distortion fields.
Of course Windows 7 is in beta. They have changed the code once again, so that code needs to be tested, and by Microsoft's rekoning, they're in the beta phase of testing. Marketing it as Windows 7 rather than Windows Vista SP2 or SP3 or SE is immaterial... you don't really think they're not changing anything, do you? If it's new code, it needs to be tested... if it's new Microsoft code, it needs a good year of testing or more, apparently.
Plus, I could argue that under the original definition of these terms, Vista actually was released to the public as an Alpha grade product. When I first started in the computer industry, an early alpha release did not imply full testing or feature lockdown, and it was for internal or very specialized developer testing only. When you went to Beta, however, the release was at full feature lockdown, and there were few if any bugs known internally -- the point of external testing was to find the bugs you can't find within your organization.
So, we know damn well that Microsoft knew there were bugs remaining in Vista upon release. And we know that some features were left out, and some of these were added in the last service pack. So, that pings "Alpha Release" by at least the two most critical criteria. But of course, Microsoft's use of these terms is always on their terms, anyway.
And while I agree with the path of GNU/Linux evolution over that of Windows as a whole, it's silly to make claims about how a tiny bit of code such as the Linux kernel compares to the huge insanity that is all of Windows. Of course it's better in many ways, but that's a meaningless comparison... if you're pointing to the Linux kernel, compare it to the NT kernel in Vista.
The practical thing to claim is that GNU/Linux wins by being fairly shallow in interdependencies... in general, you have complex things built on simpler things, and each simple things well isolated via APIs, etc.... the way an OS should exist, where applications never become part of the OS, etc. Windows was never well designed, but they apparently pushed the crazy button on Vista, with huge layers of interdependencies and weeks of lag between new modules being added and other things dependent on them being testable. Well, there are various articles on these things... if you know anything about large project development, you can read those as example of how NOT to do these things. It's not a suprise that Vista had issues, it's a wonder it worked at all. And unless they have a new development methodology (oh, I dunno... how about something like the way GNU/Linux is developed... the necessity of doing Open Source so distributed also leads to the right design and development decisions, for the most part), don't expect Windows 7 to be different... regardless of its degree of Vista heritage.
ATSC channels all broadcast a logical channel assignment, and that's what your PVR is going to use. So if your local "NBC-4" affiliate switches from UHF 48 down to some other assignment (could go back to VHF 4, but most are not returning to the lower VHF band), you'll still see it in the tuner as logical 4.
If you're already digital, the only real effect here is that, for OTA stuff, you'll potentially have to re-scan more than once. There's a 99% chance that your local station already has their switch-off/switch-over plans online.
And of course, all those upper UHF channels (700MHz band, channels above UHF 51) are going away, at least once analog is fully switched off, if not before then. They're the channels most folks are having trouble with. It's not simply the greater propagation loss due to higher frequency, but the simple fact that most consumer antenna and amplifier gear that claims "50-1000MHz" or whatever is usually crap beyond 600MHz or so anyway.
I'd love to see an Amiga at 100MHz... the fastest units from Commodore, ever, were 25MHz. You had some add-on boards with 68040s or 68060s, but those topped out at around 60MHz. There were also some crazy PowerPC add-ins at higher speeds, but they were never supported by Lightwave.
Most of the early Toaster houses kept their Amiga/Toaster rigs around for awhile. But since Lightwave had been ported to Windows NT, it was pretty common to use Alpha-based NT machines for rendering (Lightwave 3D 5.0 was supported on Intel and Alpha NT systems as a stand-alone application in 1995)... I think that's what they used after the first or second season on B5, for example.
You could certainly mix and match screens at different horizontal and vertical resolutions. The key to understand this was realizing that Amiga Amiga video was always based on a ~35ns pixel clock, and could deliver pixels of one, two, or four clocks per pixel.. thus your (nominally) 1280, 640, or 320 pixels per scanline (a bit more using the overscan region of the screen).
Vertical resolution was always the same, 480 lines (NTSC, no overscan... more in PAL, naturally). However, you could set up a 240 line mode that just automatically scan-doubled.
Those limits, and the fact you had the Copper (video coprocessor) available to automatically change all those display registers at any scanline, made the Amiga's screens concept easy.
It wouldn't work on a modern computer graphics card, though. On modern systems, any given screen setup has independent control of the pixel clock time... that's ultimately how you can set refresh rate independently of resolution. The pixel clock is generated using a phase-locked loop, which will typically take a short period of time to lock to a new rate.... possible a whole video frame's worth of time (that's why modern PC screens "jump" when you change resolution). You monitor may or may not need its own resync, but the graphics card is the first problem here.
The Advanced Amiga Architecture (never finished) has some crazy bits to allow sliding screens at different resolutions. It allowed four different pixel clocks to be present at the same time (alas, it didn't generate these, they were left as an exercise for the systems engineer), and there was a double line-buffer as well, which could do pixel duplication and line multiplying. So, along with clever software, it would have been reasonable to support multiple sliding screens at multiple resolutions.
Another form of data rot is the data file. While you probably notice that every audio, video, photo, wordprocessor, etc. format accepts a plethora of file formats, there are some that are virtually certain to be around in 30 years, and others not-so-much.
First of all... any format with an open source reader can be potentially usable in the far, far future... but to be really pedantic, including the source code in your archives every do often.
Next, go to consumer or archival formats. JPEG is a pretty easy win for photographs -- it's a web standard, it's been used in virtually every digital camera since the dawn of digital cameras, etc. On the other hand, you may get in trouble with RAW formats, which are camera-specific... I know a number of RAW readers I have don't support the RAW format from my old Canon Pro90IS. You're better off converting to 16-bit TIFF or perhaps Adobe's DNG format (no reason not to ALSO store your original RAW files if you want, other than storage efficiency).
For video, similar things: consumer formats like DV, MPEG-2, AVC, etc. are likely to be around for a long time, simply because they have such a high level of support. Something Open Source like XViD is another one you can probably count on being usable in some way or another in the distant future... and if it vanishes, you can hack up a new version yourself (well, I know I could) if you keep that source code handy.
For documents, you're probably safe with PDF, particularly now that it's been certified as a standard... but to date, older versions have always been supported in Adobe's reader, and there are many open source readers. ODF is another one that's likely to be safe simply because it's open source, so someone's likely to be able to deal with it in the future. Microsoft formats tend to change every week or two (ok, I guess they're slowing down a bit nowadays)... they're common, which means they're LIKELY to be supported, but at some point, they become uncommon and may well get dropped by current software, with no recourse on such closed formats.
For audio, open sourced formats like OGG or huge standards like the MPEG formats (Layer 2, Layer 3, AAC) are likely to be around pretty much forever. WAV and other very simple computer formats too, well described and ultra common. Odd formats ought to be avoided, particularly if they're proprietary.
And ultimately, you need to be smart about it. Check into your backups every so often, read the release notes on your new software. If the new version of Photoshop or The GIMP drops support for Amiga IFF/ILBM images, I might want to think about updating my archive files to something more modern, while that's still a pretty easy thing to do...
Basically, you have two fears -- media rot and format rot. Using reliable media, common interfaces, multiple copies, and regular attention to the issue, you avoid any single point of failure.
To deal with media rot, you ought to deal with the idea that no single piece of media is guaranteed, so backup often and in multiple places. Archival quality DVDs are a good bet. I shoot lots of digital photos, and I do a new backup when I have roughly 1/3 new material on my HDD. Thus, I have a pretty good chance of dealing with single point failures on all my archival backups.
I also keep pretty much everything online in a RAID (actually a Drobo... same idea, more flexibility) -- that gives me access to the work at any time. This changes over time, of course... and with technological improvements every few years, the storage medium never goes out of style, and never runs out of space (I'm currently on a 2TB Drobo for this stuff... all my digital photos, all my Dad's digital photos, many scanned transparencies, etc. all in one place, as well as living on DVD backups). In fact, in the case of the Drobo, getting close to full every now and then means swapping out a drive or two for more room, thus automatically cycling in new drives for old, and extending the life. This unit can deal with the total failure of one out of four drives, so it's a fairly painless way to get past the single-point-of-failure issue, even without backups (but still, do the backups, at least of data that's important going forward).
USB drives are cheap, but not ideal... check the expected storage live of your particular unit. Some multi-level flash drives are expected to last only around 10 years without refresh.
Moving to formats, think "consumer product". CD, DVDs, and now BDs are consumer products... they have a pretty long life in the marketplace, and it's currently fashionable to support compatible upgrades. USB is similar... you're far more likely to find USB 3.0 ports on a computer 10-20 years from now than a SATA connector. But as well, these things don't change overnight... you can react to the changes as they happen. That's part of the routine, just like backing things up. I used to do backups on CD, now it's all DVD, and it may eventually be Blu-Ray, or something else that shows signs of a long market life in the consumer world.
Media is also critical... a gold archival DVD is designed to last much longer than the $0.20 cheapies. You're also going to actually see that long life keeping the discs in the dark, under the proper humidity and temperature conditions, etc.
Yup.. my dv9500 failed on graphics last week... after about 10 months. One would hope their repair/replacement strategy delivers a solution that won't tank in another ten months.
There's no telling if the laptop graphics chip failures are the extent of the problem, or only the tip of the iceberg. We have a few folks here with failed desktop graphics -- no idea if it's this or another problem. If it's really a general RoHS problem, system chips will be subject to the same dynamics.
At least one line of thinking is that the failure is hastened by thermal stress, which is far more severe on a laptop than a desktop -- if so, the laptops would fail first. The BIOS hack, to speed up the fans, seem to suggest they believe this... whether or not the real intent is just to push the failures out beyond the warrenty, or if they really think running cooler will avoid some failures (the one guarantee -- your battery life is lessened).
Well, you know, once this kind of problem gets out, every problem that occurs is automatically this problem. There will probably always be some infant mortality among graphics cards, motherboards, whatever... a few failures doesn't mean your chip is necessarily affected.
Of course, I'm saying this with my HP dv9500 on the way back to HP with a dead graphics chip (nVidia 8600M, one of the originals on the list), and I'm typing this on my main desktop, which is sporting an 8600GTS card -- so I do have some concerns. Obviously, it's easier to fix the desktop, and I'm also worried about HP... do they simply swap in a new nVidia-based replacement, which will fail in another 9-11 months? Have the problems really been resolved at nVidia?
The rumor mill varies. One rumor could be fairly tragic -- that this problem affects every nVidia chip based on RoHS (lead-free) technology. If that's the case, the mobile parts are just the tip of the iceberg, and nVidia will be lucky to survive. At best, even limited to the chips everyone "knows" are affected, it's a huge mess, and just the fear of failure will drive buyers to AMD/ATi.
You have to be at SULEV/ULEV-II to be, on the average, as clean or cleaner than an electric car (based on grid averages). The SULEV qualification is 90% cleaner than the ULEV qualification.
Series hybrids are an interesting idea, but only practical if you can offset the natural loss of efficiency. Going from generator to motor, you're probably "only" about 95% efficient... if you're storing charge, this may drop to 85% or so.
Now, of course, you have something like 40% maximum efficiency for any ICE. That's a theoretical efficiency for a highly efficient Atkinson/Miller cycle engine: http://www.mhi.co.jp/technology/review/pdf/e383/e383146.pdf... most modern ICE are more like 32-36% maximum theoretical efficiency depending on design specifics. A modern diesel engine is the low 40%'s. An adiabatic ICE could hit 60%+, but we're still waiting (I first read about the concept back in the 70s).
In practice it's usually quite a bit less (the tank to wheel efficiency on a normal ICE car is around 15%, a Prius-like hybrid more like 24%), so there's room for efficiency here. And an all-electric drivetrain should pick up a few more points. Of course, some of big efficiency losses are aerodynamic drag, which applies to electrics and ICE alike.
But it's not likely to be amazing... GM's only quoting 50mpg for their Volt... Prius-like, but but not better. Sure, no turbine either (I have read of theoretical efficiencies as high as 65%), though there are plenty of issues with turbines (noise, longevity, etc).
The SULEV -- ULEV-II specifications for emissions are based on electric car equivalents. SULEV (the California version, to which '01-'03 Prius was the first conforming vehicle) requires emissions no more than an electric car, based on an average power supplier's pollution (presumably based on late-90s/early-00's averages). That was the original premise of the spec -- a reaction to California's "10% electric" mandate in the early 90s... back in the days before NiMh technology had been perfected, and Li-ion was still a pipe dream (they were working on Lithium cells, but was a 20-something-year project).
That certainly suggests that electric cars powered from plants polluting at above average levels would pollute more than the '01-'03 Prius. And the PZEV standard, first on the '04 Prius, is cleaner still.
I won't argue that stationary plants could be cleaner -- that's a specious argument, when the observable fact is that many if not most are not cleaner. And it's likely only a political act will mandate a proper cleanup or shutdown policy.
Of course, at least in some locales, consumers have choice. Here in South Jersey, we can select our power supplier just as you can pick your long distance phone carrier. I'm on a plan that's entirely wind, small hydroelectric, and solar. Yes, it's more expensive. For a bit more, I can get 100% wind from the new Atlantic City wind farm, but I'm waiting for that to come down a bit.
I'm also very anxious to replace my '03 Prius with a reasonable plug-in hybrid... I can get 100% clean power where I live, which is already better than gasoline or non-choice grid power. But if the USA really does make the effort go energy independent, all electric will tend toward cleaner (at least in the carbon sense), whether that's clean coal, nuclear, solar, or something else. That's enough reason to invest in electric car technology ASAP, IMHO.
It's silly to claim the best scientists and engineers either were or were not born here. What this guy is reacting to is pretty clearly the fairly obvious thing: many of the best scientists in the world came here to do their works. They got the high profile exposure; the peak was certainly around WWII, when we had an influx of German scientists coming over en masse.
So sure, not every scientist or engineer in this country is "world class"... but most imports... yeah, they are, pretty much by definiton. They come here to work, in a free environment, with native-grown examples cut from the same cloth.
And I think, given the more recent state of governmental support, this flux of great engineers and scientists is waning. As an engineer(I'm not claiming the "great" title), I want to get my work done. If I have a great idea that won't play at my current employer, I'll move to someone who would embrace it... no problem. Under the right circumstances, same goes with the country... a regieme restricting free thought could make plenty of projects fail.
And sadly, we've been under just such a regieme these last 7+ years... politics should NEVER have veto power over science, and yet, the Bush Administration has routinely had political drioids editing scientific papers. So there's at least some validity to the claim that we simply GAVE UP our former crown as "the world's center of technology".
I don't think "Government Work" has always meant Military. But think about the realities of today. First of all, this was DARPA (Defense Advanced Research Projects Administration), not NASA. "Defense" has long been code-word for "Military", but under the Bush Adminstration, it's clear that "Defense" really means "offense", "imperalism", "American Empire", "PAX Americana", and any other PNAC buzzword you're like to employ. The world has changed.
But it wasn't all THAT different pre-Bush. I was lured... and that's is the correct word.. into working for General Electric when I graduated CMU in 1983. They sold me on the Space Shuttle... as a kid who used to stay up really late to see every Apollo launch or moon EVA that was televised in the '60s and '70s, what higher goal could I shoot for.. money was the farthest thing from my mind... every good Engineer makes a fine living.
But when I arrived in Philly, I was assigned to a ground within GE called "Computational Design".. their first effort to bring in folks allowed to work in both HW and SW. I got assigned to fix their RTL modeling system... and ultimately, found this was only being used to simulate nuclear missiles re-entering the earth, and then detonating.
I was there a total of four months. And it was all deception on their part. I wonder how many new hires were dragged into such immoral work but didn't have the guts to leave. Clearly, if this was a respected occupation, they wouldn't have tricked me into it.
There are plenty of answers. But the bottom line is simple: most of "the best and the brightest" simply do not want to work on technology designed to kill people. It really is that simple.
You cannot compare this to the Einstein era. In World War II, there was a reasonable belief that the Nazi regieme might actually be successful in their effort to take over the world. That level of desparation will get the best of us involved in things we might otherwise not support.
Despire the radical right's efforts to paint the activities of thousands of equally radical Islamists, as a similar threat... well, the best and the brightest among us are not that stupid. Duh. We are substantially smarter than the people using the fear of terrorism to maintain political power in the USA.
Personally, I was in the direct line of this kind of decision tree back in the early 80s. My first job out of college (CMU '83), I was hired at General Electric, lured in based on the prospect of working on tiny bits of the Space Shuttle.
In reality, it was all about Nuclear Weapons... specifically, GE then was dealing with the things ICBMs had to do on the way back to earth. It took me about two months (no security clearance yet) to really grok all they were doing, and a month or so to get the balls to quit a company after such a short time out of school (but I went on the work at Commodore computer, helping to deliver your Commodore 128, Amiga 2000, 3000, and 4000).
The people who make things, who really want to create something new, who have that vision and talent, will not waste their lives building the tools of death, unless our fundamental way of life is seriously threatened. So DARPA will have to subsist, basically, on the people who are rejected from other ventures, and the occasional Evil Genius.
The problem is, patents were NEVER intended to be applied to pure ideas, as they now often are. They were intended to apply only to implementations of ideas: creative expressions of an idea, perhaps, but a real thing, a creative work: an invention, not a discovery. Not something we've had for 60,000 year, "but now on the Web". Not something obvious to any High School student just learning the art.
And the reason for this was not primarily to help the patentee... the main reason for granting patents (a short monopoly on this invention) was to remove the fear of public disclosure, to ADVANCE the art. Others could study the invention, learn from it, improve upon it, find other ways to implement the same concept. In short, it was primarily in the public interest, with a small bit of support for the inventor to make it worthwhile to the inventor to fully document said invention in a publically accessible document.
Your initial argument has merit, but largely based on what's wrong with the US IP system, not what's wrong with IP in general. Unfortunately, you mire that in anti-propetarian rhetoric which will have many readers branding you a kook and not noting that you make a stellar point here.
If I create something -- a purely creative act, whether literary, musical, or engineering, it's far more than "planting a flag". This thing is a product of my mind and hard work; without me, it never would have existed. I firmly believe this sort of intellectual property has merit.
What's wrong with the system (and explicitly excluded in the original definition of a patent) is the idea that one can patent an idea -- there's you flag planting. When I worked at Commodore, we ran into this... a company called CADTrac had a patent on using XOR for cursors... basically, they had somehow patented a fundamental operation of boolean algebra.
You see this over and over again these days... take something humanity been doing for 50,000 years, tack an "on the web" to it, and all of a sudden, it's considered "new". I don't think for a minute any of this stuff, which is a combination of bad laws and companies learning to hack the system, should exist. If it's obvious, it's a general principle of physics, mathematics, etc. and you're just "planting a flag", it should not be given protection as IP. In fact, I think that's a pretty good test.
My 17" HP dv9500, at $1285, actually has more features than the $2995 17" Apple PowerBook, at least last November, which I placed the order. Same 2.4GHz Intel Core2 Duo CPU (Apple, being the only Intel-only house left, was able to offer a 2.6GHz as a pricey option), same memory (2GB), similar arrangement of ports (not identical), etc.
Sure, the HP came with 64-bit Vista rather than MacOS, but even if MacOS were the best OS human history, it wouldn't have been much value... all my apps are on Windows. Some simply don't exist on MacOS, others would have run the tab up thousands more for the same functionality, but less satisfaction (eg, I'm not about to trade Vegas for FCP or Acid for GarageBand)...
This isn't news... Apple has ALWAYS managed to seel their stuff for about 2x-3x the price of similarly powerful hardware from other companies. That's why they have all that cash in the bank. A great gig if you can get it. Just don't expect everyone to fall for it.
His name is "Chim-chim", thank-you-very-much. And he is critical element of the whole Speed Racer gestalt. If you don't get Chim-chim, you're missing the who point...
Slashdot Mirror
That's incorrect, at least in part. Modern MacOS is based on CMU's Mach, which has had lightweight threading support since long before Apple got into the picture. The OS was completely designed for multiple CPUs, down to the very core.
If modern MacOS apps are not heavily multithreaded (I have no idea, I don't run priorietary hardware anymore, regardless of the OS), that's the fault of programmers not advancing past the days of MacOS 9... it has nothing whatsoever to do with the OS.
Multithreading is a system-level thing, not a language level thing.
Sure, there have been languages that make threading ubiquitous, but they've never caught on, and it's hardly necessary.
You'll notice that internet, graphics, and many other programming necessities are not built into C/C++ either. They are higher level functions, and thousands of programmers have no problem understanding C's role here. People have been writing multithreading code in C/C++ for decades... I've personally done in from the 80s until now, under a dozen or so OSs.
Don't use your chosen language as a crutch for sicking to the level of programming practiced when that langauge debuted. The whole point of C was not to define much of anything in C itself.. in truth, the language proper doesn't even do I/O... that's handled via a library. So is threading, so is graphics, etc.
The idea of an OS and/or suppoet tools handling the SMP problem is nothing more than a crutch for bad programming.
In fact, anyone who grew up with a real multitheaded, multitasking OS is already writing code that will scale just dandy to 8 cores and beyond. When you accept that a thread is nothing more or less than a typical programming construct, you simply write better code. This is no more or less an amazing thing than when regular programmers embraced subroutines or structures.
This was S.O.P. back in the late 80s under the AmigaOS, and enhanced in the early/mid 90s under BeOS. This in not new, and not even remotely tied to the advent of multicore CPUs.
The problem here is simple: UNIX and Windows. Windows had fake multitasking for so long, Windows programmers barely knew what you could do when you had "thread" in the same toolkit as "subroutine", rather than it being something exotic. UNIX, as a whole, didn't even have lightweight preemptive threads until fairly recently, and UNIX programmers are only slowly catching up.
However, neither of these is even slightly an OS problem... it's an application-level problem. If programmers continue to code as if they had a 70s-vintage OS, they're going to think in single threads and suck on 8-core CPUs. If programmers update themselves to state-of-the-1980s thinking, they'll scale to 8-cores and well beyond.
I wrote a version of LISP back in the early 80s that included both a NIL and an ERROR atom. Any undefined or illegal operation on valid arguments would result in the ERROR atom; subsequent operations on the error atom were all defined to not break, but also return the ERROR atom. A list of illegal operations would grow, attached to ERROR.
This worked fairly well, since error handling could be done a useful points, it didn't have to be done everywhere. Not as powerful as true exceptions, of course, but much easier than some exception models (they weren't as well known or defined in '83 as they are today, either).
While Tony Hoare may not have lifted NULL from LISP, LISP 1.5 had its own version, NIL, in 1960 or thereabouts... NIL was even represented as 0, internally, at least in early versions of LISP.
While less likely to cause problems than in a language like Algol or C, I do believe this is the earliest use of "null" in a programming language (at least until some cheeky joker out there finds an earlier one I didn't know about).
Well, funny thing... before I bought my HP dv9500, I did precisely this... I compared it, feature by feature, to the 17" Powerbook. Both were direct from manufacturer based on customizable web site builds. Of the things I cared about, the HP exceeded the Mac in every area except one: the Mac has both Fw800 and Fw400 ports, while the HP just the Fw400 (on 4-pin, sure, but it's for my camcorders, so that's fine).
The result: the Powerbook was $2999, the HP was $1280. Not even a contest -- Apple goes down in flames. This was done in December of 2007... I can't honestly claim that Apple hasn't halved their prices since then, but this "spec for spec" claim rarely if ever works outside the bounds of a few individuals' reality distortion fields.
Of course Windows 7 is in beta. They have changed the code once again, so that code needs to be tested, and by Microsoft's rekoning, they're in the beta phase of testing. Marketing it as Windows 7 rather than Windows Vista SP2 or SP3 or SE is immaterial... you don't really think they're not changing anything, do you? If it's new code, it needs to be tested... if it's new Microsoft code, it needs a good year of testing or more, apparently.
Plus, I could argue that under the original definition of these terms, Vista actually was released to the public as an Alpha grade product. When I first started in the computer industry, an early alpha release did not imply full testing or feature lockdown, and it was for internal or very specialized developer testing only. When you went to Beta, however, the release was at full feature lockdown, and there were few if any bugs known internally -- the point of external testing was to find the bugs you can't find within your organization.
So, we know damn well that Microsoft knew there were bugs remaining in Vista upon release. And we know that some features were left out, and some of these were added in the last service pack. So, that pings "Alpha Release" by at least the two most critical criteria. But of course, Microsoft's use of these terms is always on their terms, anyway.
And while I agree with the path of GNU/Linux evolution over that of Windows as a whole, it's silly to make claims about how a tiny bit of code such as the Linux kernel compares to the huge insanity that is all of Windows. Of course it's better in many ways, but that's a meaningless comparison... if you're pointing to the Linux kernel, compare it to the NT kernel in Vista.
The practical thing to claim is that GNU/Linux wins by being fairly shallow in interdependencies... in general, you have complex things built on simpler things, and each simple things well isolated via APIs, etc.... the way an OS should exist, where applications never become part of the OS, etc. Windows was never well designed, but they apparently pushed the crazy button on Vista, with huge layers of interdependencies and weeks of lag between new modules being added and other things dependent on them being testable. Well, there are various articles on these things... if you know anything about large project development, you can read those as example of how NOT to do these things. It's not a suprise that Vista had issues, it's a wonder it worked at all. And unless they have a new development methodology (oh, I dunno... how about something like the way GNU/Linux is developed... the necessity of doing Open Source so distributed also leads to the right design and development decisions, for the most part), don't expect Windows 7 to be different... regardless of its degree of Vista heritage.
ATSC channels all broadcast a logical channel assignment, and that's what your PVR is going to use. So if your local "NBC-4" affiliate switches from UHF 48 down to some other assignment (could go back to VHF 4, but most are not returning to the lower VHF band), you'll still see it in the tuner as logical 4.
If you're already digital, the only real effect here is that, for OTA stuff, you'll potentially have to re-scan more than once. There's a 99% chance that your local station already has their switch-off/switch-over plans online.
And of course, all those upper UHF channels (700MHz band, channels above UHF 51) are going away, at least once analog is fully switched off, if not before then. They're the channels most folks are having trouble with. It's not simply the greater propagation loss due to higher frequency, but the simple fact that most consumer antenna and amplifier gear that claims "50-1000MHz" or whatever is usually crap beyond 600MHz or so anyway.
I'd love to see an Amiga at 100MHz... the fastest units from Commodore, ever, were 25MHz. You had some add-on boards with 68040s or 68060s, but those topped out at around 60MHz. There were also some crazy PowerPC add-ins at higher speeds, but they were never supported by Lightwave.
Most of the early Toaster houses kept their Amiga/Toaster rigs around for awhile. But since Lightwave had been ported to Windows NT, it was pretty common to use Alpha-based NT machines for rendering (Lightwave 3D 5.0 was supported on Intel and Alpha NT systems as a stand-alone application in 1995)... I think that's what they used after the first or second season on B5, for example.
You could certainly mix and match screens at different horizontal and vertical resolutions. The key to understand this was realizing that Amiga Amiga video was always based on a ~35ns pixel clock, and could deliver pixels of one, two, or four clocks per pixel.. thus your (nominally) 1280, 640, or 320 pixels per scanline (a bit more using the overscan region of the screen).
Vertical resolution was always the same, 480 lines (NTSC, no overscan... more in PAL, naturally). However, you could set up a 240 line mode that just automatically scan-doubled.
Those limits, and the fact you had the Copper (video coprocessor) available to automatically change all those display registers at any scanline, made the Amiga's screens concept easy.
It wouldn't work on a modern computer graphics card, though. On modern systems, any given screen setup has independent control of the pixel clock time... that's ultimately how you can set refresh rate independently of resolution. The pixel clock is generated using a phase-locked loop, which will typically take a short period of time to lock to a new rate.... possible a whole video frame's worth of time (that's why modern PC screens "jump" when you change resolution). You monitor may or may not need its own resync, but the graphics card is the first problem here.
The Advanced Amiga Architecture (never finished) has some crazy bits to allow sliding screens at different resolutions. It allowed four different pixel clocks to be present at the same time (alas, it didn't generate these, they were left as an exercise for the systems engineer), and there was a double line-buffer as well, which could do pixel duplication and line multiplying. So, along with clever software, it would have been reasonable to support multiple sliding screens at multiple resolutions.
Another form of data rot is the data file. While you probably notice that every audio, video, photo, wordprocessor, etc. format accepts a plethora of file formats, there are some that are virtually certain to be around in 30 years, and others not-so-much.
First of all... any format with an open source reader can be potentially usable in the far, far future... but to be really pedantic, including the source code in your archives every do often.
Next, go to consumer or archival formats. JPEG is a pretty easy win for photographs -- it's a web standard, it's been used in virtually every digital camera since the dawn of digital cameras, etc. On the other hand, you may get in trouble with RAW formats, which are camera-specific... I know a number of RAW readers I have don't support the RAW format from my old Canon Pro90IS. You're better off converting to 16-bit TIFF or perhaps Adobe's DNG format (no reason not to ALSO store your original RAW files if you want, other than storage efficiency).
For video, similar things: consumer formats like DV, MPEG-2, AVC, etc. are likely to be around for a long time, simply because they have such a high level of support. Something Open Source like XViD is another one you can probably count on being usable in some way or another in the distant future... and if it vanishes, you can hack up a new version yourself (well, I know I could) if you keep that source code handy.
For documents, you're probably safe with PDF, particularly now that it's been certified as a standard... but to date, older versions have always been supported in Adobe's reader, and there are many open source readers. ODF is another one that's likely to be safe simply because it's open source, so someone's likely to be able to deal with it in the future. Microsoft formats tend to change every week or two (ok, I guess they're slowing down a bit nowadays)... they're common, which means they're LIKELY to be supported, but at some point, they become uncommon and may well get dropped by current software, with no recourse on such closed formats.
For audio, open sourced formats like OGG or huge standards like the MPEG formats (Layer 2, Layer 3, AAC) are likely to be around pretty much forever. WAV and other very simple computer formats too, well described and ultra common. Odd formats ought to be avoided, particularly if they're proprietary.
And ultimately, you need to be smart about it. Check into your backups every so often, read the release notes on your new software. If the new version of Photoshop or The GIMP drops support for Amiga IFF/ILBM images, I might want to think about updating my archive files to something more modern, while that's still a pretty easy thing to do...
Actually, glass mastered DVDs fail, but for other reasons: the layers separate, the aluminium oxidizes, etc.
You're better off with archival-quality DVD-Rs stored in the dark... no to mention avoiding a pretty hefty pre-disc charge for the glass master.
Basically, you have two fears -- media rot and format rot. Using reliable media, common interfaces, multiple copies, and regular attention to the issue, you avoid any single point of failure.
To deal with media rot, you ought to deal with the idea that no single piece of media is guaranteed, so backup often and in multiple places. Archival quality DVDs are a good bet. I shoot lots of digital photos, and I do a new backup when I have roughly 1/3 new material on my HDD. Thus, I have a pretty good chance of dealing with single point failures on all my archival backups.
I also keep pretty much everything online in a RAID (actually a Drobo... same idea, more flexibility) -- that gives me access to the work at any time. This changes over time, of course... and with technological improvements every few years, the storage medium never goes out of style, and never runs out of space (I'm currently on a 2TB Drobo for this stuff... all my digital photos, all my Dad's digital photos, many scanned transparencies, etc. all in one place, as well as living on DVD backups). In fact, in the case of the Drobo, getting close to full every now and then means swapping out a drive or two for more room, thus automatically cycling in new drives for old, and extending the life. This unit can deal with the total failure of one out of four drives, so it's a fairly painless way to get past the single-point-of-failure issue, even without backups (but still, do the backups, at least of data that's important going forward).
USB drives are cheap, but not ideal... check the expected storage live of your particular unit. Some multi-level flash drives are expected to last only around 10 years without refresh.
Moving to formats, think "consumer product". CD, DVDs, and now BDs are consumer products... they have a pretty long life in the marketplace, and it's currently fashionable to support compatible upgrades. USB is similar... you're far more likely to find USB 3.0 ports on a computer 10-20 years from now than a SATA connector. But as well, these things don't change overnight... you can react to the changes as they happen. That's part of the routine, just like backing things up. I used to do backups on CD, now it's all DVD, and it may eventually be Blu-Ray, or something else that shows signs of a long market life in the consumer world.
Media is also critical... a gold archival DVD is designed to last much longer than the $0.20 cheapies. You're also going to actually see that long life keeping the discs in the dark, under the proper humidity and temperature conditions, etc.
Yup.. my dv9500 failed on graphics last week... after about 10 months. One would hope their repair/replacement strategy delivers a solution that won't tank in another ten months.
There's no telling if the laptop graphics chip failures are the extent of the problem, or only the tip of the iceberg. We have a few folks here with failed desktop graphics -- no idea if it's this or another problem. If it's really a general RoHS problem, system chips will be subject to the same dynamics.
At least one line of thinking is that the failure is hastened by thermal stress, which is far more severe on a laptop than a desktop -- if so, the laptops would fail first. The BIOS hack, to speed up the fans, seem to suggest they believe this... whether or not the real intent is just to push the failures out beyond the warrenty, or if they really think running cooler will avoid some failures (the one guarantee -- your battery life is lessened).
Well, you know, once this kind of problem gets out, every problem that occurs is automatically this problem. There will probably always be some infant mortality among graphics cards, motherboards, whatever... a few failures doesn't mean your chip is necessarily affected.
Of course, I'm saying this with my HP dv9500 on the way back to HP with a dead graphics chip (nVidia 8600M, one of the originals on the list), and I'm typing this on my main desktop, which is sporting an 8600GTS card -- so I do have some concerns. Obviously, it's easier to fix the desktop, and I'm also worried about HP... do they simply swap in a new nVidia-based replacement, which will fail in another 9-11 months? Have the problems really been resolved at nVidia?
The rumor mill varies. One rumor could be fairly tragic -- that this problem affects every nVidia chip based on RoHS (lead-free) technology. If that's the case, the mobile parts are just the tip of the iceberg, and nVidia will be lucky to survive. At best, even limited to the chips everyone "knows" are affected, it's a huge mess, and just the fear of failure will drive buyers to AMD/ATi.
You have to be at SULEV/ULEV-II to be, on the average, as clean or cleaner than an electric car (based on grid averages). The SULEV qualification is 90% cleaner than the ULEV qualification.
Series hybrids are an interesting idea, but only practical if you can offset the natural loss of efficiency. Going from generator to motor, you're probably "only" about 95% efficient... if you're storing charge, this may drop to 85% or so.
Now, of course, you have something like 40% maximum efficiency for any ICE. That's a theoretical efficiency for a highly efficient Atkinson/Miller cycle engine: http://www.mhi.co.jp/technology/review/pdf/e383/e383146.pdf... most modern ICE are more like 32-36% maximum theoretical efficiency depending on design specifics. A modern diesel engine is the low 40%'s. An adiabatic ICE could hit 60%+, but we're still waiting (I first read about the concept back in the 70s).
In practice it's usually quite a bit less (the tank to wheel efficiency on a normal ICE car is around 15%, a Prius-like hybrid more like 24%), so there's room for efficiency here. And an all-electric drivetrain should pick up a few more points. Of course, some of big efficiency losses are aerodynamic drag, which applies to electrics and ICE alike.
But it's not likely to be amazing... GM's only quoting 50mpg for their Volt... Prius-like, but but not better. Sure, no turbine either (I have read of theoretical efficiencies as high as 65%), though there are plenty of issues with turbines (noise, longevity, etc).
The SULEV -- ULEV-II specifications for emissions are based on electric car equivalents. SULEV (the California version, to which '01-'03 Prius was the first conforming vehicle) requires emissions no more than an electric car, based on an average power supplier's pollution (presumably based on late-90s/early-00's averages). That was the original premise of the spec -- a reaction to California's "10% electric" mandate in the early 90s... back in the days before NiMh technology had been perfected, and Li-ion was still a pipe dream (they were working on Lithium cells, but was a 20-something-year project).
That certainly suggests that electric cars powered from plants polluting at above average levels would pollute more than the '01-'03 Prius. And the PZEV standard, first on the '04 Prius, is cleaner still.
I won't argue that stationary plants could be cleaner -- that's a specious argument, when the observable fact is that many if not most are not cleaner. And it's likely only a political act will mandate a proper cleanup or shutdown policy.
Of course, at least in some locales, consumers have choice. Here in South Jersey, we can select our power supplier just as you can pick your long distance phone carrier. I'm on a plan that's entirely wind, small hydroelectric, and solar. Yes, it's more expensive. For a bit more, I can get 100% wind from the new Atlantic City wind farm, but I'm waiting for that to come down a bit.
I'm also very anxious to replace my '03 Prius with a reasonable plug-in hybrid... I can get 100% clean power where I live, which is already better than gasoline or non-choice grid power. But if the USA really does make the effort go energy independent, all electric will tend toward cleaner (at least in the carbon sense), whether that's clean coal, nuclear, solar, or something else. That's enough reason to invest in electric car technology ASAP, IMHO.
It's silly to claim the best scientists and engineers either were or were not born here. What this guy is reacting to is pretty clearly the fairly obvious thing: many of the best scientists in the world came here to do their works. They got the high profile exposure; the peak was certainly around WWII, when we had an influx of German scientists coming over en masse.
So sure, not every scientist or engineer in this country is "world class"... but most imports... yeah, they are, pretty much by definiton. They come here to work, in a free environment, with native-grown examples cut from the same cloth.
And I think, given the more recent state of governmental support, this flux of great engineers and scientists is waning. As an engineer(I'm not claiming the "great" title), I want to get my work done. If I have a great idea that won't play at my current employer, I'll move to someone who would embrace it... no problem. Under the right circumstances, same goes with the country... a regieme restricting free thought could make plenty of projects fail.
And sadly, we've been under just such a regieme these last 7+ years... politics should NEVER have veto power over science, and yet, the Bush Administration has routinely had political drioids editing scientific papers. So there's at least some validity to the claim that we simply GAVE UP our former crown as "the world's center of technology".
I don't think "Government Work" has always meant Military. But think about the realities of today. First of all, this was DARPA (Defense Advanced Research Projects Administration), not NASA. "Defense" has long been code-word for "Military", but under the Bush Adminstration, it's clear that "Defense" really means "offense", "imperalism", "American Empire", "PAX Americana", and any other PNAC buzzword you're like to employ. The world has changed.
But it wasn't all THAT different pre-Bush. I was lured... and that's is the correct word.. into working for General Electric when I graduated CMU in 1983. They sold me on the Space Shuttle... as a kid who used to stay up really late to see every Apollo launch or moon EVA that was televised in the '60s and '70s, what higher goal could I shoot for.. money was the farthest thing from my mind... every good Engineer makes a fine living.
But when I arrived in Philly, I was assigned to a ground within GE called "Computational Design".. their first effort to bring in folks allowed to work in both HW and SW. I got assigned to fix their RTL modeling system... and ultimately, found this was only being used to simulate nuclear missiles re-entering the earth, and then detonating.
I was there a total of four months. And it was all deception on their part. I wonder how many new hires were dragged into such immoral work but didn't have the guts to leave. Clearly, if this was a respected occupation, they wouldn't have tricked me into it.
There are plenty of answers. But the bottom line is simple: most of "the best and the brightest" simply do not want to work on technology designed to kill people. It really is that simple.
You cannot compare this to the Einstein era. In World War II, there was a reasonable belief that the Nazi regieme might actually be successful in their effort to take over the world. That level of desparation will get the best of us involved in things we might otherwise not support.
Despire the radical right's efforts to paint the activities of thousands of equally radical Islamists, as a similar threat... well, the best and the brightest among us are not that stupid. Duh. We are substantially smarter than the people using the fear of terrorism to maintain political power in the USA.
Personally, I was in the direct line of this kind of decision tree back in the early 80s. My first job out of college (CMU '83), I was hired at General Electric, lured in based on the prospect of working on tiny bits of the Space Shuttle.
In reality, it was all about Nuclear Weapons... specifically, GE then was dealing with the things ICBMs had to do on the way back to earth. It took me about two months (no security clearance yet) to really grok all they were doing, and a month or so to get the balls to quit a company after such a short time out of school (but I went on the work at Commodore computer, helping to deliver your Commodore 128, Amiga 2000, 3000, and 4000).
The people who make things, who really want to create something new, who have that vision and talent, will not waste their lives building the tools of death, unless our fundamental way of life is seriously threatened. So DARPA will have to subsist, basically, on the people who are rejected from other ventures, and the occasional Evil Genius.
The problem is, patents were NEVER intended to be applied to pure ideas, as they now often are. They were intended to apply only to implementations of ideas: creative expressions of an idea, perhaps, but a real thing, a creative work: an invention, not a discovery. Not something we've had for 60,000 year, "but now on the Web". Not something obvious to any High School student just learning the art.
And the reason for this was not primarily to help the patentee... the main reason for granting patents (a short monopoly on this invention) was to remove the fear of public disclosure, to ADVANCE the art. Others could study the invention, learn from it, improve upon it, find other ways to implement the same concept. In short, it was primarily in the public interest, with a small bit of support for the inventor to make it worthwhile to the inventor to fully document said invention in a publically accessible document.
Your initial argument has merit, but largely based on what's wrong with the US IP system, not what's wrong with IP in general. Unfortunately, you mire that in anti-propetarian rhetoric which will have many readers branding you a kook and not noting that you make a stellar point here.
If I create something -- a purely creative act, whether literary, musical, or engineering, it's far more than "planting a flag". This thing is a product of my mind and hard work; without me, it never would have existed. I firmly believe this sort of intellectual property has merit.
What's wrong with the system (and explicitly excluded in the original definition of a patent) is the idea that one can patent an idea -- there's you flag planting. When I worked at Commodore, we ran into this... a company called CADTrac had a patent on using XOR for cursors... basically, they had somehow patented a fundamental operation of boolean algebra.
You see this over and over again these days... take something humanity been doing for 50,000 years, tack an "on the web" to it, and all of a sudden, it's considered "new". I don't think for a minute any of this stuff, which is a combination of bad laws and companies learning to hack the system, should exist. If it's obvious, it's a general principle of physics, mathematics, etc. and you're just "planting a flag", it should not be given protection as IP. In fact, I think that's a pretty good test.
My 17" HP dv9500, at $1285, actually has more features than the $2995 17" Apple PowerBook, at least last November, which I placed the order. Same 2.4GHz Intel Core2 Duo CPU (Apple, being the only Intel-only house left, was able to offer a 2.6GHz as a pricey option), same memory (2GB), similar arrangement of ports (not identical), etc.
Sure, the HP came with 64-bit Vista rather than MacOS, but even if MacOS were the best OS human history, it wouldn't have been much value... all my apps are on Windows. Some simply don't exist on MacOS, others would have run the tab up thousands more for the same functionality, but less satisfaction (eg, I'm not about to trade Vegas for FCP or Acid for GarageBand)...
This isn't news... Apple has ALWAYS managed to seel their stuff for about 2x-3x the price of similarly powerful hardware from other companies. That's why they have all that cash in the bank. A great gig if you can get it. Just don't expect everyone to fall for it.
His name is "Chim-chim", thank-you-very-much. And he is critical element of the whole Speed Racer gestalt. If you don't get Chim-chim, you're missing the who point...