The Bush administration is in a mess here. Their real problem is that if they'd asked the FISA court for the authority to do what they're doing, they'd have been turned down. If they'd asked Congress for it, some tough questions would have been asked by members of Congress in a position to demand answers. Remember, the conservative right, "Bush's Base", isn't comfortable with wiretapping.
Bush can go to Congress for more wiretapping authority, but right now, he probably wouldn't get it. Hence the desperate legal moves.
And they are desperate. Notice what happened here. The Administration tried to use a secrecy order to prevent this issue from going to trial. That's because they can't win on the merits.
But since the Administration had already admitted enough in public to establish that such wiretapping was going on, that didn't prevent the court from addressing the issue.
At the appeals level, the facts of the case aren't reviewed, just the law. Because, as the district judge pointed out, it is not controverted that such wiretapping occured, that's not a issue. So the secrecy issue isn't really an issue on appeal. This leaves the Administration with only its weak arguments.
Incidentally, this is a criminal statute. See 50 USC 1811. If you work for NSA, or a telephone company and are involved in illegal wiretapping, you could go to jail for five years. That could happen years in the future, under a future administration.
On the Apple side, the classic price of getting it right was having to "rebuild the desktop" occasionally, thereby rebuilding all the derived information. That was annoying, but certainly beat having to reinstall software.
Microsoft had a chance to get it right when they added the Registry, but they botched it. There should have been a clear distinction in the Registry between application-owned info, derived info, and user-preference info, with the first being static, the second being rebuildable, and the third having a user interface for examining it. But Microsoft mixed them all up.
"In this case, the President has acted, undisputedly, as FISA forbids. FISA is the expressed
statutory policy of our Congress. The presidential power, therefore, was exercised at its lowest ebb
and cannot be sustained."
"We must first note that the Office of the Chief Executive has itself been created, with its
powers, by the Constitution. There are no hereditary Kings in America and no powers not created
by the Constitution. So all "inherent powers" must derive from that Constitution."
"For all of the reasons outlined above, this court is constrained to grant to Plaintiffs the Partial
Summary Judgment requested, and holds that the TSP violates the APA; the Separation of Powers
doctrine; the First and Fourth Amendments of the United States Constitution; and the statutory law."
"The Permanent Injunction of the TSP requested by Plaintiffs is granted inasmuch as each of
the factors required to be met to sustain such an injunction have undisputedly been met. The
irreparable injury necessary to warrant injunctive relief is clear, as the First and Fourth Amendment
rights of Plaintiffs are violated by the TSP. See Dombrowski v. Pfister, 380 U.S. 479 (1965). The
irreparable injury conversely sustained by Defendants under this injunction may be rectified by
compliance with our Constitution and/or statutory law, as amended if necessary. Plaintiffs have
prevailed, and the public interest is clear, in this matter. It is the upholding of our Constitution.
As Justice Warren wrote in U.S. v. Robel, 389 U.S. 258 (1967):
Implicit in the term 'national defense' is the notion of defending
those values and ideas which set this Nation apart. . . . It would
indeed be ironic if, in the name of national defense, we would
sanction the subversion of . . . those liberties....Id. at 264.
IT IS SO ORDERED.
ANNA DIGGS TAYLOR
UNITED STATES DISTRICT JUDGE"
Actually, the US should abandon all copyright treaties.
That's probably out of reach politically, and it's not even a good idea, but scaling back to the 50 years of TRIPS might be within reach. Few works have significant commercial value after 50 years. Maybe put in a renewal provision that allows renewal after 50 years for $10,000/year, to keep Disney happy. This would allow a smooth flow of old content into the public domain.
I don't think that this ("Make copy protection illegal for uncopyrightable material") would be constitutional given the First Amendment.
To the extent that such copy protection required collusion between the content distributor and the player manufacturer, it could be made illegal under the commerce clause.
Free spectrum, free content If it goes out over the free airwaves, like TV channels for which broadcasters do not pay, it can't be copy protected.
See above re: copy protection generally resulting in a forfeiting of copyright.
The format of broadcasts is controlled in great detail by the FCC; certainly they can forbid copy protection. In fact, they forbade over the air pay TV decades ago.
And we'll also need shorter and more granular terms, to re-strengthen formalities (registration, deposit, notice, frequent renewal).
The TRIPS agreement forbids "formalities", so that's out without WTO negotiation.
Just keep referring to those dweebs with wires hanging out of their ears as "pod people". Laugh at them behind their back. They can't hear you. They can't even hear the bus that's about to run them over. Don't they look stupid? Too cool for school, too dumb for the real world.
Lessig is starting to sound like Stallman. Stallman is more effective, though. What we need is some serious lobbying, along the following lines:
Copyright harmonization The US should not go beyond the 50 years of the TRIPS agreement. 50 years from first publication, copyright expires. That's it. Free Elvis! (The US can do that unilaterally. Less than 50 years requires international negotiation.)
Make copy protection illegal for uncopyrightable material If you can't copyright it, you can't use technical means to protect it.
Enforce the Audio Home Recording Act Any arrangement between manufacturers and/or content distributors to restrict rights guaranteed to consumers is illegal restraint of trade.
Free spectrum, free content If it goes out over the free airwaves, like TV channels for which broadcasters do not pay, it can't be copy protected.
Apple, in its early days, had a good sense of what was important in a user interface, and that was expressed in the "Apple Human Interface Guidelines". Much of that knowledge has been lost.
One of the original Apple rules was "You should never have to tell the computer something it already knows". Consistently applying this rule requires a clear separation between infomration about the host environment and individual user preferences, something most programs don't do well. Apple was reasonably faithful to that rule in their early days, but over time, got sloppy.
Microsoft never did as well, and it was an alien concept in the UNIX world.
It's common, but wrong, to bind environment decisions at program install time, which means that a change in the environment breaks applications in mysterious ways. The whole concept of "installers" is flawed, when you think about it. You should just put the application somewhere, and when launched, it adapts to the environment, hopefully not taking very long if nothing has changed. That was the original MacOS concept.
Much of the trouble comes from failing to distinguish between primary and derived sources for information. "This program understands.odf format" is primary information, and should be permanently associated with the program itself and readable by other programs. ".odf documents can be opened with any of the following programs" is derived information, and should be cached and invalidated based on the primary information. "I would prefer to open.odf documents with OpenOffice" is a user preference. None of the mainstream operating systems quite get this right. That kind of thing is the frontier in user interfaces, not eye candy.
What Heinlein was describing, though, was more like a keyboard-controlled drafting machine. Keuffel & Esser actually built something like that, just as CAD was starting to eat into their business. You could either draw manually, or you could get it to plot by keyboard entry. Great for lettering. Dead end idea.
What Heinlein didn't get was that the big win for CAD is drawing revision. With CAD, you can edit an existing drawing. With a paper-based system, you have to redraw. Which is Not Fun. The pre-CAD high point in revision tools was to draw in pencil on heavy paper and use an electric-powered eraser, then run the paper drawing through a blueline copier to make working copies.
Every OS project on the Alto. Smalltalk-80 ate the world. It also hosted an influential Lisp environment.
Actually, Interlisp didn't run on the Alto. No way would it have fit. It ran on DEC PDP-10 type machines, or, at PARC, on MAXC, Xerox PARC's in-house PDP-10 clone. DWIM (Do What I Mean) was a dud; we used to call it the "Warren Teitelbaum typing error corrector", because it was so tuned to the typing errors Warren made, like pressing the shift key at the wrong time. Despite the claims that it "never did anything disasterously wrong", I once typed "EDIT" while in the wrong mode, and DWIM typed back "=EXIT", and threw me out of Interlisp without saving. Never used DWIM again.
The Alto didn't really have much of an OS. What it had was roughly at the DOS level. When Smalltalk was running, it took over the machine, like a DOS app.
Nature was a great journal once, but they've gone way downhill. The biology articles are still excellent (or at least seem to be; I can't judge), but they've published some terrible computer science articles.
We can do better than that. In no particular order,
IBM's VM operating system. (1972) OS/360 was a nightmare, and not even the best OS of its era. Burroughs and UNIVAC were way ahead in operating systems in the late 1960s. VM, though, had paging, good security, hypervisor capability, and good performance. In the 1970s. And it's still in use.
Backus' FORTRAN compiler (1957) The first good compiler. Optimizing, even. Better code generation than anything running on UNIX prior to the mid-1980s.
QNX (1980) The first really good microkernel OS. Still in use, deep inside railroad signalling systems, machine tools, and nuclear reactor controls, where it has to work.
NLS (1967) The first system with a mouse, windows, and a GUI. It took a mainframe to make it go in 1967, but all the key ideas were there.
AutoCAD (1982) This is the program that replaced the drafting board. Huge increase in productivity. Ever ink in a drawing by hand? Redraw a drawing to make changes? Engineering companies used to have acres of people doing that stuff. No more.
Bravo (1974) The first what-you-see-is-what-you-get text editor. Multiple fonts. Ran on the Xerox Alto. The ancestor of all modern word processors.
Those are older examples, each a major advance over previous technology. As the technology becomes more mature, the advances become smaller, but more widely deployed.
There are lots of large-scale games you can play if you have the time, dedicated people, and money. Paintball, for example. Foxhunting. Geocaching. Push balls. Big mazes. Thunderdome at Burning Man.
The logistics tend to get you, and the cost per player-hour tends to be high. If you cut costs too much, you don't get the feeling of the exotic needed to make it work. If you screw up the game design, people can be hurt. This stuff tends not to scale well.
Then there's the problem that if you do anything wierd in an urban setting today, the lower level anti-terrorism people have a cow.
What these guys really seem to be doing is running an ad agency that does sponsored public stunts.
Here are some of the real innovators. The links given are all to their earliest pages, from 1996.
Fedex.com FedEx had the first major web site that did something - you could track packages and get an immediate response.
Viaweb.com The first web site that supported page creation via the web. The first general-purpose shopping cart. Eventually became Yahoo Store. Implemented in LISP.
Due to the efforts of Willam Kahan at U.C. Berkeley, IEEE 754 floating point, which is what we have today on almost everything, is far, far better than earlier implementations.
Just for starters, IEEE floating point guarantees that, for integer values that fit in the mantissa, addition, subtraction, and multiplication will give the correct integer result. Some earlier FPUs would give results like 2+2 = 3.99999. IEEE 754 also guarantees exact equality for integer results; you're guaranteed that 6*9 == 9*6. Fixing that made spreadsheets acceptable to people who haven't studied numerical analysis.
The "not a number" feature of IEEE floating point handles annoying cases, like division by zero. Underflow is handled well. Overflow works. 80-bit floating point is supported (except on PowerPC, which broke many engineering apps when Apple went to PowerPC.)
Those of us who do serious number crunching have to deal with this all the time. It's a big deal for game physics engines, many of which have to run on the somewhat lame FPUs of game consoles.
Installation d'une fonction offensive C dans la macro
DicOOo.
La fonction C est exécutée à l'installation de DicOOo.
"DicOOo" is an installer for dictionaries into OpenOffice. Unfortunately, it seems to have too much power, and can be replaced or induced to install other things. This is an add-on to OpenOffice, and apparently an unsafe one.
The end user controlled Windows 2000. Windows XP is remote-controlled from Redmond, with automatic updates, remote DRM rule changes, and similar intrusiveness. Microsoft can reboot your XP machine by remote control. (Remember last year, when they did, even for users who thought they had that turned off?) They can't do that to Windows 2000 through SP2. Many large companies don't like that, and have stayed with Windows 2000. Yes, supposedly you can turn off all the remote controls in XP, if you can find them and if your end users don't accidentally do something that reactivates them.
Remote control actually started, though, with Windows 2000 SP3, not with with XP. That's what the EULA change reflects. It's when Microsoft added "The Software features described below are enabled by
default to connect via the Internet to Microsoft
computer systems automatically, without separate
notice to you. You consent to the operation
of these features, unless you choose to switch
them off or not use them. " to the Windows 2000 EULA. That was when Microsoft first officially put a backdoor into your computer.
If you're a law firm, or a bank, or a medical service provider, or a government agency, or a Microsoft competitor, this creates some compliance problems. If you work for any of those, show that EULA to your lawyers and see what they have to say about it.
There are quite a few big companies still running Windows 2000 SP2 because of this. That's why you can still buy a new Windows 2000 system from Dell if you're a corporate customer.
The halting problem is not an issue for program verification.
This claim is raised repeatedly by the clueless, and it just isn't an issue.
Yes, you can construct a program that's formally undecideable. It's a hard way to write a bad program. It takes some work, and the resulting program is unlikely to be useful.
Most crash-type and security-hole problems in programs are entirely decidable. This is because almost all subscript calculations are composed from addition, multiplication by constants, and logic operations. Those are totally decideable, and there are good decision algorithms for that problem. Only when multiplication of two variables (both non-constant) is introduced can formal undecidability appear. See Presburger arithmetic.
In fact, halting is decidable for all deterministic machines with finite memory. Either you repeat a previous state, or halt within a finite number of cycles. The decision process may be made arbitrarily hard, but that's not undecidability. True undecidability in the Turing sense requires infinite memory.
Most of the practical problems with program verification come from dealing with interactions between various parts of the program. Containing those interactions well enough that you can localize problems is constraining on the programmer. "Design by contract" languages like Eiffel try to do that, but they're not popular. Retrofitting design by contract into C and C++ has been discussed, but the proposed schemes all have holes you could drive a truck through. A big truck.
It is not possible for a program to analyze another program and find all the bugs; see halting problem.
Wrong. It is quite possible to analyze a program and find all the bugs that violate the language constraints (null pointers, buffer overflows, etc.). That's what program verification is for. For some programs, you can't tell whether a bug condition will occur, so you treat that as a bug.
Automated program verification is a good idea that went away because C and C++ have such ambiguous semantics. It's hopeless for those languages. The "pointer equals array" concept alone makes it very tough, because the language has no idea how big an array is. Worst idea in the language, and the root cause of buffer overflows.
You can put HD content on a DVD, just not as much of it. This is useful if you create animation work. Is there anything consumer-grade that can play that out into a 1080p monitor?
You can put about 20 minutes of DVD-formatted content on a CD-ROM, and most PCs will happily play it. Putting one of those disks in a consumer DVD player, though, generally has disappointing results. Some players will play such things, winding the disk up to an unusually high speed. Most will fail to play it. Some will crash. One crashed so badly that it had to be power cycled twice. It's the cheap players with PC-type drives that usually work.
Apple's "Core Animation" looks like a scheme for developing apps with a Flash-based GUI. Video games have been doing that for years; in many games, the user interface is authored in Flash, but displayed with a third-party Flash player built into the game. For that matter, developers have been able to put Flash-based GUIs into Photon applications for QNX since about 2001. So this isn't exactly a new idea.
Did Apple provide any useful guidance for developers on what a GUI developed this way should look like, or is every app going to have some different approach?
Palo Alto has a place called Neotte. This is a tea bar with WiFi and a power strip at every table. Everybody there is on a laptop. The tea is about $4, and they have a modest selection of bakery items. It's the next notch up from Starbucks.
Perhaps a upscale gamer cafe where you bring your own laptop...
That's a nice system. There are some annoying limitations, though. It's tough to change the orientation of the robot, although you can handle that if you have a rotational joint further up so you can rotate the torso. Small diameter balls have the same problems as small diameter wheels - it's easy to get stuck in small depressions. That's why the Segway has such big wheels. And driving a sphere is always a tough problem mechanically. Most of the solutions have trouble with dirt accumulating on the drive wheels, which is why optical mice have replaced ball mice. It's possible to build a spherical electric motor, and that might be the way to go if this concept turns out to be useful.
It's good to see all this activity in self-balancing systems again, having worked on this around 1994-1995, and seen others working on it in the 1980s. Today, you can buy so much of what you need off the shelf, like good INS units. We used to waste too much time building custom stuff.
Slashdot Mirror
The Bush administration is in a mess here. Their real problem is that if they'd asked the FISA court for the authority to do what they're doing, they'd have been turned down. If they'd asked Congress for it, some tough questions would have been asked by members of Congress in a position to demand answers. Remember, the conservative right, "Bush's Base", isn't comfortable with wiretapping. Bush can go to Congress for more wiretapping authority, but right now, he probably wouldn't get it. Hence the desperate legal moves.
And they are desperate. Notice what happened here. The Administration tried to use a secrecy order to prevent this issue from going to trial. That's because they can't win on the merits. But since the Administration had already admitted enough in public to establish that such wiretapping was going on, that didn't prevent the court from addressing the issue.
At the appeals level, the facts of the case aren't reviewed, just the law. Because, as the district judge pointed out, it is not controverted that such wiretapping occured, that's not a issue. So the secrecy issue isn't really an issue on appeal. This leaves the Administration with only its weak arguments.
Incidentally, this is a criminal statute. See 50 USC 1811. If you work for NSA, or a telephone company and are involved in illegal wiretapping, you could go to jail for five years. That could happen years in the future, under a future administration.
On the Apple side, the classic price of getting it right was having to "rebuild the desktop" occasionally, thereby rebuilding all the derived information. That was annoying, but certainly beat having to reinstall software.
Microsoft had a chance to get it right when they added the Registry, but they botched it. There should have been a clear distinction in the Registry between application-owned info, derived info, and user-preference info, with the first being static, the second being rebuildable, and the third having a user interface for examining it. But Microsoft mixed them all up.
The actual decision by the court is worth reading. Some quotes:
IT IS SO ORDERED.
ANNA DIGGS TAYLOR
UNITED STATES DISTRICT JUDGE"
Actually, the US should abandon all copyright treaties.
That's probably out of reach politically, and it's not even a good idea, but scaling back to the 50 years of TRIPS might be within reach. Few works have significant commercial value after 50 years. Maybe put in a renewal provision that allows renewal after 50 years for $10,000/year, to keep Disney happy. This would allow a smooth flow of old content into the public domain.
I don't think that this ("Make copy protection illegal for uncopyrightable material") would be constitutional given the First Amendment.
To the extent that such copy protection required collusion between the content distributor and the player manufacturer, it could be made illegal under the commerce clause.
Free spectrum, free content If it goes out over the free airwaves, like TV channels for which broadcasters do not pay, it can't be copy protected.
See above re: copy protection generally resulting in a forfeiting of copyright.
The format of broadcasts is controlled in great detail by the FCC; certainly they can forbid copy protection. In fact, they forbade over the air pay TV decades ago.
And we'll also need shorter and more granular terms, to re-strengthen formalities (registration, deposit, notice, frequent renewal).
The TRIPS agreement forbids "formalities", so that's out without WTO negotiation.
Just keep referring to those dweebs with wires hanging out of their ears as "pod people". Laugh at them behind their back. They can't hear you. They can't even hear the bus that's about to run them over. Don't they look stupid? Too cool for school, too dumb for the real world.
Live by the hype, die by the hype
Lessig is starting to sound like Stallman. Stallman is more effective, though. What we need is some serious lobbying, along the following lines:
-
Copyright harmonization The US should not go beyond the 50 years of the TRIPS agreement. 50 years from first publication, copyright expires. That's it. Free Elvis! (The US can do that unilaterally. Less than 50 years requires international negotiation.)
-
Make copy protection illegal for uncopyrightable material If you can't copyright it, you can't use technical means to protect it.
-
Enforce the Audio Home Recording Act Any arrangement between manufacturers and/or content distributors to restrict rights guaranteed to consumers is illegal restraint of trade.
-
Free spectrum, free content If it goes out over the free airwaves, like TV channels for which broadcasters do not pay, it can't be copy protected.
Now that's a reasonable agenda to lobby for.Apple, in its early days, had a good sense of what was important in a user interface, and that was expressed in the "Apple Human Interface Guidelines". Much of that knowledge has been lost.
One of the original Apple rules was "You should never have to tell the computer something it already knows". Consistently applying this rule requires a clear separation between infomration about the host environment and individual user preferences, something most programs don't do well. Apple was reasonably faithful to that rule in their early days, but over time, got sloppy. Microsoft never did as well, and it was an alien concept in the UNIX world.
It's common, but wrong, to bind environment decisions at program install time, which means that a change in the environment breaks applications in mysterious ways. The whole concept of "installers" is flawed, when you think about it. You should just put the application somewhere, and when launched, it adapts to the environment, hopefully not taking very long if nothing has changed. That was the original MacOS concept.
Much of the trouble comes from failing to distinguish between primary and derived sources for information. "This program understands .odf format" is primary information, and should be permanently associated with the program itself and readable by other programs. ".odf documents can be opened with any of the following programs" is derived information, and should be cached and invalidated based on the primary information. "I would prefer to open .odf documents with OpenOffice" is a user preference. None of the mainstream operating systems quite get this right. That kind of thing is the frontier in user interfaces, not eye candy.
The people who wrote AutoCAD all knew that story.
What Heinlein was describing, though, was more like a keyboard-controlled drafting machine. Keuffel & Esser actually built something like that, just as CAD was starting to eat into their business. You could either draw manually, or you could get it to plot by keyboard entry. Great for lettering. Dead end idea.
What Heinlein didn't get was that the big win for CAD is drawing revision. With CAD, you can edit an existing drawing. With a paper-based system, you have to redraw. Which is Not Fun. The pre-CAD high point in revision tools was to draw in pencil on heavy paper and use an electric-powered eraser, then run the paper drawing through a blueline copier to make working copies.
Every OS project on the Alto. Smalltalk-80 ate the world. It also hosted an influential Lisp environment.
Actually, Interlisp didn't run on the Alto. No way would it have fit. It ran on DEC PDP-10 type machines, or, at PARC, on MAXC, Xerox PARC's in-house PDP-10 clone. DWIM (Do What I Mean) was a dud; we used to call it the "Warren Teitelbaum typing error corrector", because it was so tuned to the typing errors Warren made, like pressing the shift key at the wrong time. Despite the claims that it "never did anything disasterously wrong", I once typed "EDIT" while in the wrong mode, and DWIM typed back "=EXIT", and threw me out of Interlisp without saving. Never used DWIM again.
The Alto didn't really have much of an OS. What it had was roughly at the DOS level. When Smalltalk was running, it took over the machine, like a DOS app.
Nature was a great journal once, but they've gone way downhill. The biology articles are still excellent (or at least seem to be; I can't judge), but they've published some terrible computer science articles.
-
IBM's VM operating system. (1972) OS/360 was a nightmare, and not even the best OS of its era. Burroughs and UNIVAC were way ahead in operating systems in the late 1960s. VM, though, had paging, good security, hypervisor capability, and good performance. In the 1970s. And it's still in use.
-
Backus' FORTRAN compiler (1957) The first good compiler. Optimizing, even. Better code generation than anything running on UNIX prior to the mid-1980s.
-
QNX (1980) The first really good microkernel OS. Still in use, deep inside railroad signalling systems, machine tools, and nuclear reactor controls, where it has to work.
-
NLS (1967) The first system with a mouse, windows, and a GUI. It took a mainframe to make it go in 1967, but all the key ideas were there.
-
AutoCAD (1982) This is the program that replaced the drafting board. Huge increase in productivity. Ever ink in a drawing by hand? Redraw a drawing to make changes? Engineering companies used to have acres of people doing that stuff. No more.
-
Bravo (1974) The first what-you-see-is-what-you-get text editor. Multiple fonts. Ran on the Xerox Alto. The ancestor of all modern word processors.
Those are older examples, each a major advance over previous technology. As the technology becomes more mature, the advances become smaller, but more widely deployed.This always amazes me. I mean, all the mammals have basically the same parts, and humans have over 99% DNA similarity with chimpanzees.
One can puzzle over things like how the basic structures of animals evolved, but the human/ape thing isn't even an issue.
There are lots of large-scale games you can play if you have the time, dedicated people, and money. Paintball, for example. Foxhunting. Geocaching. Push balls. Big mazes. Thunderdome at Burning Man.
The logistics tend to get you, and the cost per player-hour tends to be high. If you cut costs too much, you don't get the feeling of the exotic needed to make it work. If you screw up the game design, people can be hurt. This stuff tends not to scale well.
Then there's the problem that if you do anything wierd in an urban setting today, the lower level anti-terrorism people have a cow.
What these guys really seem to be doing is running an ad agency that does sponsored public stunts.
Due to the efforts of Willam Kahan at U.C. Berkeley, IEEE 754 floating point, which is what we have today on almost everything, is far, far better than earlier implementations.
Just for starters, IEEE floating point guarantees that, for integer values that fit in the mantissa, addition, subtraction, and multiplication will give the correct integer result. Some earlier FPUs would give results like 2+2 = 3.99999. IEEE 754 also guarantees exact equality for integer results; you're guaranteed that 6*9 == 9*6. Fixing that made spreadsheets acceptable to people who haven't studied numerical analysis.
The "not a number" feature of IEEE floating point handles annoying cases, like division by zero. Underflow is handled well. Overflow works. 80-bit floating point is supported (except on PowerPC, which broke many engineering apps when Apple went to PowerPC.)
Those of us who do serious number crunching have to deal with this all the time. It's a big deal for game physics engines, many of which have to run on the somewhat lame FPUs of game consoles.
Installation d'une fonction offensive C dans la macro DicOOo.
La fonction C est exécutée à l'installation de DicOOo.
"DicOOo" is an installer for dictionaries into OpenOffice. Unfortunately, it seems to have too much power, and can be replaced or induced to install other things. This is an add-on to OpenOffice, and apparently an unsafe one.
Apple had a power adapter safety recall by the Consumer Product Safety Commission back in the G3 era, and a battery recall last year. Is this a new problem?
It's a real issue, but a lost cause.
The end user controlled Windows 2000. Windows XP is remote-controlled from Redmond, with automatic updates, remote DRM rule changes, and similar intrusiveness. Microsoft can reboot your XP machine by remote control. (Remember last year, when they did, even for users who thought they had that turned off?) They can't do that to Windows 2000 through SP2. Many large companies don't like that, and have stayed with Windows 2000. Yes, supposedly you can turn off all the remote controls in XP, if you can find them and if your end users don't accidentally do something that reactivates them.
Remote control actually started, though, with Windows 2000 SP3, not with with XP. That's what the EULA change reflects. It's when Microsoft added "The Software features described below are enabled by default to connect via the Internet to Microsoft computer systems automatically, without separate notice to you. You consent to the operation of these features, unless you choose to switch them off or not use them. " to the Windows 2000 EULA. That was when Microsoft first officially put a backdoor into your computer.
If you're a law firm, or a bank, or a medical service provider, or a government agency, or a Microsoft competitor, this creates some compliance problems. If you work for any of those, show that EULA to your lawyers and see what they have to say about it.
There are quite a few big companies still running Windows 2000 SP2 because of this. That's why you can still buy a new Windows 2000 system from Dell if you're a corporate customer.
Most of what's on YouTube seems to be from commercial content sources, usually movies and broadcast TV. I'm surprised the MPAA hasn't shut them down.
Most content seems to be recompressed, and badly, with huge blocky artifacts.
The halting problem is not an issue for program verification. This claim is raised repeatedly by the clueless, and it just isn't an issue.
Yes, you can construct a program that's formally undecideable. It's a hard way to write a bad program. It takes some work, and the resulting program is unlikely to be useful.
Most crash-type and security-hole problems in programs are entirely decidable. This is because almost all subscript calculations are composed from addition, multiplication by constants, and logic operations. Those are totally decideable, and there are good decision algorithms for that problem. Only when multiplication of two variables (both non-constant) is introduced can formal undecidability appear. See Presburger arithmetic.
In fact, halting is decidable for all deterministic machines with finite memory. Either you repeat a previous state, or halt within a finite number of cycles. The decision process may be made arbitrarily hard, but that's not undecidability. True undecidability in the Turing sense requires infinite memory.
Most of the practical problems with program verification come from dealing with interactions between various parts of the program. Containing those interactions well enough that you can localize problems is constraining on the programmer. "Design by contract" languages like Eiffel try to do that, but they're not popular. Retrofitting design by contract into C and C++ has been discussed, but the proposed schemes all have holes you could drive a truck through. A big truck.
Although software work seldom uses proof of correctness techniques, there's a whole industry doing it for hardware. There was a machine-generated formal proof of correctness for the FPU in AMD's K7 processor. AMD thus avoided the "Pentium division bug".
It is not possible for a program to analyze another program and find all the bugs; see halting problem .
Wrong. It is quite possible to analyze a program and find all the bugs that violate the language constraints (null pointers, buffer overflows, etc.). That's what program verification is for. For some programs, you can't tell whether a bug condition will occur, so you treat that as a bug.
Automated program verification is a good idea that went away because C and C++ have such ambiguous semantics. It's hopeless for those languages. The "pointer equals array" concept alone makes it very tough, because the language has no idea how big an array is. Worst idea in the language, and the root cause of buffer overflows.
Good verifiers were written for Pascal (I headed one of those projects), a good one was written for Java (at DEC, just before DEC went under), and Microsoft is working on one for C#.
You can put HD content on a DVD, just not as much of it. This is useful if you create animation work. Is there anything consumer-grade that can play that out into a 1080p monitor?
You can put about 20 minutes of DVD-formatted content on a CD-ROM, and most PCs will happily play it. Putting one of those disks in a consumer DVD player, though, generally has disappointing results. Some players will play such things, winding the disk up to an unusually high speed. Most will fail to play it. Some will crash. One crashed so badly that it had to be power cycled twice. It's the cheap players with PC-type drives that usually work.
Apple's "Core Animation" looks like a scheme for developing apps with a Flash-based GUI. Video games have been doing that for years; in many games, the user interface is authored in Flash, but displayed with a third-party Flash player built into the game. For that matter, developers have been able to put Flash-based GUIs into Photon applications for QNX since about 2001. So this isn't exactly a new idea.
Did Apple provide any useful guidance for developers on what a GUI developed this way should look like, or is every app going to have some different approach?
Palo Alto has a place called Neotte. This is a tea bar with WiFi and a power strip at every table. Everybody there is on a laptop. The tea is about $4, and they have a modest selection of bakery items. It's the next notch up from Starbucks.
Perhaps a upscale gamer cafe where you bring your own laptop...
That's a nice system. There are some annoying limitations, though. It's tough to change the orientation of the robot, although you can handle that if you have a rotational joint further up so you can rotate the torso. Small diameter balls have the same problems as small diameter wheels - it's easy to get stuck in small depressions. That's why the Segway has such big wheels. And driving a sphere is always a tough problem mechanically. Most of the solutions have trouble with dirt accumulating on the drive wheels, which is why optical mice have replaced ball mice. It's possible to build a spherical electric motor, and that might be the way to go if this concept turns out to be useful.
It's good to see all this activity in self-balancing systems again, having worked on this around 1994-1995, and seen others working on it in the 1980s. Today, you can buy so much of what you need off the shelf, like good INS units. We used to waste too much time building custom stuff.